Did Trump's efforts to challenge the 2020 election outcome help or hinder him in 2024? That is, was his performance in the 2024 election better or worse than in an alternate universe where he didn't warn about cheating beforehand, claim it afterwards, call state officials, arrange alternative electors or hold a rally on January 6th?
There was a brief period after January 6th in which many people thought he'd disqualified himself; that centrists and some Republicans (politicians and voters) would reject him to signal that those actions weren't acceptable. My impression is that this happened but in low numbers. On the other hand, I imagine the part of his base who believe his claims would be energised by those actions, as in their view justice was quashed and January 6th protestors were jailed for standing up for a righteous cause. Is there any way to estimate which effect was stronger?
-- the claim is that someone must be pumping in catchy words like "constitutional crisis" into the narrative stream.
How does one determine if this could be a genuine coincidence; after all there is a lot of media out there, and constitutional crisis is not exactly a rare word. Or, whether this is organic rather than cultivated virality?
Why should it be a coincidence? Journalists do read what others are writing; also, if there is a constitutional crisis brewing (such as a president telling his appointees to ignore the law), it's the journalists' job to call it that.
Go read Nobody Special's answer. It makes all the good points about how this is an emergent result of uncoordinated activity without a "manipulative intent". It makes all the "nice and interesting points" -- the search engines distilling the attention of the public, the web being a malleable media, the iteration speed and cross-pollination driving convergence.
Unfortunately it's all well-meaning bunk. The real reason for CNN and Politico using the same term is simple: CNN and Politico senior editors are in the same chat room, at some point they brainstormed what term the headlines should use, and now they are using it.
Oh yeah, and they totally had "manipulative intent".
I can see that what you say is *a priori plausible*, but from there, how does one go to "probable"?
Edit: Other than appealing to purely subjective guesstimates of "odds". In other words, how can people with opposite readings of the issue talk to rather than past each other on this matter?
I think it's due to a variety of factors, but none of them conspiratorial or reflecting manipulative intent.
Historically, journalism did this as well. There are a variety of names one could give to Nixon's scandal breaking into and spying on the DNC, but ultimately coverage coalesced around "Watergate" rather than "White House Spying" or other alternatives. This happens for perfectly good non-conspiratorial reasons. News organizations wanted to sell papers, and once the public conversation consolidates around particular nomenclature, you want to fit that nomenclature so that your audience understands what you're selling. In the early days of Watergate, headlines likely varied. Once you're a month in, if all the other papers are festooned with headlines on the latest news on "Watergate," while your paper has updates on the "Liddy Affair," the public may not understand what you're selling and consequently their interest may drift to other sellers.
This trend continues into the internet era, where it is exacerbated by multiple factors. Not a media expert, so YMMV but just to list a few that seem most plausible to me:
(1) News is now elevated by algorithm and search engines. The average reader googling "Trump Constitutional Crisis" may not care much whether they get the latest updates from CNN or Politico. But Politico and CNN, respectively, care a lot which link the reader clicks, and consequently have a strong desire to best mold their article to whatever keywords people are searching.
(2) Add to #1 the fact that internet articles can be revised while live, in a way that old newspapers can't. That means that once algorithm emergent social consensus around a particular phrase emerges, you can go back and revise all your "DNC Break-In" articles to be "Watergate" articles. And because of the nature of the search engine driving clicks, which drive eyeballs and money, you have every incentive to do those revisions - you'd be crazy not to.
(3) High-turnover media and consequent cross citation. The need for a constant churn of new media means less and less resources & time to dedicate to truly new reporting. If I need new articles every 4 hours on the latest DOGE activities, there's no way I'm getting those out fast enough if I limit myself to solely my own independent research. So I’m much more likely to cite a CNN article that came out an hour before mine, and other publications coming out later will cite to me. Doing anything else results in my content coming too late, which no one can afford, so further ambient pressure in the system standardizes the descriptions of the coverage.
It's the same reason that there were tons of headlines about the "superbowl" this weekend, and a ton of headlines about "tariffs" and "plane crashes" last month.
People talking about the news isn't exactly a conspiracy, and the president triggering or threatening to trigger a constitutional crisis is pretty fucking big news.
As someone who has hated Musk long before it was fashionable, let me outline the case briefly:
- Musk is a charlatan who brags about accomplishments that are not his (e.g., being the founder of Tesla). When he actually designs something, you get things like the cybertruck.
- he is a vaporware salesman who has built a massive house of cards on empty promises (robotaxis, "fixing traffic", the "hyperloop", going to Mars...)
- he treats his employees like shit, and doesn't give a rat's ass about laws, regulations and workplace safety.
If you need the long-form version of everything wrong with the man, consult the youtube channel "Common Sense Skeptic".
ETA: of course, Musk has recently done a lot of despicable stuff beyond what I wrote, but that's on the news anyway, so I don't have to repeat it.
Obviously I can't keep you from considering the man a hero, but I only see a few options:
- you have solid information that my claims above are wrong or least misleading. Then I'd like to hear them. I consider that unlikely, because this is all pretty well documented.
- you acknowledge the claims above, but judge that, compared to what remains of Musk's merits, they are irrelevent. Then you should think about what it means that you adore an amoral blowhard.
- you ignore the claims above and stick your fingers in your ears. Again, you should think about what that means for your judgement criteria.
- you acknowledge the claims and adore Musk BECAUSE he is an amoral blowhard. In that case, you are clearly part of the problem.
I didn't say I adored Musk. Let's not stick words in each others mouth. I said he is a hero. Is he a flawed hero? Of course all real heroes have their flaws. Is he an asshole sometimes, yes. Does wrong stuff, yes.. all the others. IDK, but I'll take your word for it. so take that as a yes. I've read a biography of Musk and listened to ~10+ hrs of him taking with Lex Fridman and Rogan. My impression (and this comes partially from believing Lex.) is that he's the same person in private as he is in public. I call that a straight shooter. He's telling you what he really thinks and I respect that. As far as business, he seems to have (so far) avoided having his businesses taken over by middle management. He does this by drilling down into everything, he seems somewhat unique in this regard. Now he's offered to focus this talent into our bureaucratic state. And I say, "Drill baby drill." (Sorry I just couldn't resist that last line. :^)
"My impression (and this comes partially from believing Lex.) is that he's the same person in private as he is in public. I call that a straight shooter. He's telling you what he really thinks and I respect that. "
That's the same line that people use about Trump. Problem is, they're both pathological liars who say whatever suits their needs at a given time.
Do you know the story about the solar rooftop presentation? Im 2016, Musk arranged for an unveil event for solar shingles at a movie set and said "The houses around you are all solar houses. Did you notice?" - and went on to present solar shingles in various designs and colors.
The houses around had nothing to do with solar, and the shingles weren't functional. Why did he do that? To convince Tesla shateholders to buy SolarCity, a failing company owned by his cousins. Does that sound like a straight shooter or like a conman?
So like... this is a thing that you could google, right? There are *at least* thousands of people online who have spelled out in some detail why they hate Elon Musk. If you don't understand, it's because you haven't tried to. Like, at all. And since you didn't phrase it as a question, but more of a boast, I guess you're proud of that?
But anyway, I'll pretend you want to understand, and tell you my thoughts.
For most people I know, buying twitter and allowing Nazi accounts to thrive (e.g. https://www.nbcnews.com/tech/social-media/x-twitter-elon-musk-nazi-extremist-white-nationalist-accounts-rcna145020) was the tipping point from "Elon is an annoying rich guy" to "Elon is a mortal enemy". And to anticipate the claim that he did it for "free speech" I would point out that he has suspended real journalists while allowing Nazis to remain, and he's limited speech he doesn't like (such as the terms "cis" and "cisgender"). He is not a principled advocate for free speech - he is a principled advocate for white nationalism. Also, not sure if you know this, but he did a Nazi salute on TV. Twice. And a lot of people hate Nazis.
Do you also need me to explain why people hate Nazis?
But again, that's just my personal experience with why people hate Musk. By all means, take to google or reddit and see what other people have to say. I mean, we both know you won't do that. But the information is there, and now you can't say no one has ever showed it to you.
Yeah no I do understand. I live with my son and he hates Musk too. And so let me make a stab at why I think people hate Musk, and it's really the same reason that people hate Bobby (Kennedy). And I can identify the reason, because I use to hate Bobby and Trump and others. And that is because I bought into what everyone else was saying about them. Everyone says these people are evil and going to do harm and are bad, and they tell you why, and the easy thing is to just believe them and say yes. I hate them too. The only way (I know) to break this trap, is to listen to what the people say themselves. This often involves (for me) listening to a several hour podcast (from maybe Lex or Rogan) and getting to know the person. (Or reading the books they have written.) Most people don't have the time or inclination to do that... so they just say what everyone around them is saying. Now perhaps I started out in a bit better position, because I've listened to Rogan for years, and I know most of the things people say about him are a distortion. So I was open to finding out that the main stream view of these other people, (Musk and Bobby) was perhaps a distortion. And I guess I'm just asking you to be open to the possibility that these people aren't evil.
About Nazis: I don't like Nazis, I don't want to listen to them, and yet in the name of free speech I think Nazis should be allowed to air their ideas. I don't know about Musk limiting speech about cis and cis gender, but clearly this is also not right and in my opinion a mistake on Musk's part.
Although, if you're super interested in what Elon Musk said, remember that time he said that Jews were promoting anti-white racism? And then when advertisers started to leave twitter because of all the Nazi stuff on twitter, do you remember Elon's response? He told the advertisers - not the Nazis - the advertisers, who are his company's source of income - "go fuck yourselves".
It sure seems like when Elon was promoting electric cars and talking about colonizing Mars, everyone was cool with him. He even got a shoutout on Star Trek Discovery, a show so woke they made Stacey Abrams President of Earth. And then it seems like when he started promoting Nazis, then people said "hey, wait, actually this guy sucks."
So I would say it's the exact opposite of your theory. People liked him just fine until they listened to him freely and publicly express his thoughts, at which point they were disgusted.
I also don't engage directly on Twitter, but if you're choosing to ignore the inflammatory things that Musk makes a point of saying and amplifying on the platform he bought at enormous expense and risk to his entire business empire, I wonder why you would do that and instead ONLY pay attention to the things he says during laid-back interviews with sympathetic bros like Rogan and Friedman.
I mean, if a person helps a bunch of Nazis to spread Nazi ideology, and then gives them the Nazi salute twice on TV, and then you listen to that guy on a podcast and decide you like him, does that mean that you were mistaken before and he's actually a good guy, or that you are mistaken now and have been conned by a Nazi? Lots of people decided to listen to Hitler and then decided that he had a lot of good points. We all grew up wondering why people listened to Hitler, because in retrospect it's obvious that he was evil. But apparently, it's not always immediately obvious that evil people are evil, because apparently they have some kind of persuasive ability that convinces people to listen to them, despite what they do when given power.
So I guess the question is, do you think that a program of eliminating trans people, purging the country of immigrants, purging the government of anyone who isn't loyal to Trump, eliminating vaccinations and public health programs, eliminating support for democracy abroad, ethnic cleansing the Gaza strip to develop it as a seaside resort, and suddenly cutting off food and medicine to millions of people, including children, who were depending and counting on it is heroic? Or is that all Nazi stuff? If you have an answer to that, it really doesn't matter what Elon says when he's smoking pot with Joe Rogan.
Oh dear, I don't think Musk is a Nazi, you are free to have a different opinion. TBH I don't pay all that much attention to what Trump and Musk are doing. (I did watch that youtube video above... I fast forwarded through some of it.) I think 1/2 the time Trump and Musk are trolling the other side, and it would be much better if they were not such assholes. But there is not much I can do about that.
I don't think all the purging you are worried about is going to happen. I think we need to shrink the federal government, and borders need to be controlled along with immigration. Trans people will be fine, I know they are all very worried right now. I'm not sure but much of that is fear generated from the left. Support for democracy abroad is almost Orwellian in some respects, we need to let other countries self determine, and stop interfering. I don't expect vaccines to be eliminated, but Bobby might recommend delaying some, or giving parents the option to opt out of some. IDK I'm hoping he can MAHA, he really wants to help IMHO. And for the rest... yeah heroic in that I think he is doing the right thing, even though he will be hated for it. But that's just my opinion.
"TBH I don't pay all that much attention to what Trump and Musk are doing."
Previously you said you listen to what they say. But you don't pay attention to what they do. And then you're confused that people have a different opinion than you?
You're doing this the wrong way round. This is how you get taken in by con men.
I don't pay attention to the news, which is distorted this way or that by the news agency. I've listened to some of the senate confirmation hearings. And the above video with Trump and Musk talking to reporters in the oval office. That's not reporting, it's the ground or first level truth. You can say, no Trump and Musk are talking at a higher level and conning me. That could be true, so shame on me. But until I see evidence otherwise, I'm going to assume that they are talking ground or first level truth to me. It is the only way we can effectively communicate with each other. And I give you here, my first level truth.
5. RFK went to Samoa to campaign against vaccines in the midst of a measles outbreak that killed 83 people. There is currently a measles outbreak in Texas. Measles can be eliminated by vaccines.
It seems like the theme here is that you doubt that Musk, Trump, and RFK are going to do things *they've already done*, and your answer to "Musk posts a bunch of Nazi stuff" is "well I don't look at it."
So, I think you have your answer: people hate Musk because they pay attention to what he says and does, and read news, and know what's going on in the world. You like Musk, Trump, and RFK because you have no idea what they are doing and can't be bothered to find out.
A good trick for understanding stuff like this is to imagine the situation with opposite partisanship.
For example, imagine that Harris was president, and she forcibly shut down ICE by decree and fired all border agents against the wishes of congress. You'd be pretty mad in that case, right?
Hmm my immigration views are mixed at the moment, so that's not a good analogy for me. And TBH what Musk is trying to do is fix our budget. I'm all in favor of that, and if we had to open our borders to get our budget under control, I'd sign up for that.
What Musk is trying to do is to destroy or hamstring government agencies and programs he disapproves of. There is zero possibility that the things he is doing will "fix" the budget, or even close the gap sufficiently that other plausible measures might fix the budget. And if he were trying to fix the budget, the first thing on his agenda would be to try and convince Donald Trump to walk back his tax cut proposals.
Musk isn't an idiot, he knows all this, and he's still going Full DOGE. So, notwithstanding the name, he isn't trying to fix the budget.
Hmm yeah you may be right, and I'm a stooge. In which case I'll admit it. I think we need more time to find out. I'm sorry I don't know anything about Trumps tax cuts. Do people expect them to be approved? None of this budget stuff happens unless congress gets involved.
I remain as hopeful as I was in the first few months of Obama's first term. That all came to nothing, but I remain hopeful again. (I've got practice I'm a Buffalo Bills fan :^)
This is literally THE NUMBER ONE PRIORITY of Republicans in Congress. The chances of them not going through are extremely small.
I hate political news as much as anyone, but this is a case where even the most cursory reading would correct an extremely inaccurate view of the world.
The number one priority of Republicans in Congress is keeping themselves in Congress. But, yeah, the tax cuts are number one in the hype and I suspect in the top three actual priorities. And the few GOP congressmen who are willing to sometimes stand up to Trump, have to pick their battles carefully. They're not going to die on the hill of "Americans should pay more taxes!"
So, yes, we are going to get another five trillion dollars in debt out of this.
The Republican's are currently targeting a deficit increase of 4.5 TRILLION over the next decade. There's no way that you can pretend that Musk is killing highly effective programs like PEPFAR (about 6 billion a year) because he cares about the debt.
This is a bit like defunding the police "because you just can't spare the money and hard choices have to be made" and then spending 50x as much on student loan forgiveness.
I think pepfar is back to being funded? But I don't know the status.
I don't understand your defund the police and student debt point. Personally I think police could use more funding, and student loan forgiveness is a terrible idea. (I also think that student debt not being able to be discharged in bankruptcy is a terrible law.)
Oh a deficit of ~0.5 trillion per year would be way better than what it is now. https://fiscaldata.treasury.gov/americas-finance-guide/national-deficit/ Of course I assume the Republicans are making all the savings in the last half of that decade and that this year it will be over 1 trillion again... sigh.
> Oh a deficit of ~0.5 trillion per year would be way better than what it is now
You misunderstand. 4.5 trillion isn't their target for the deficit over 10 years. It's their target for how much to INCREASE the deficit over 10 years.
To increase the debt, not the deficit. Or phrase it some other way, but nobody talks about 10 year deficits, if you're talking about a deficit you need to use annualized terms.
> I think pepfar is back to being funded? But I don't know the status.
I think Musk made noises in that direction, but it's hard to believe when he's fired everyone and "put it through the woodchipper". At this point, there's probably nothing left to unfreeze in the first place.
> Personally I think police could use more funding, and student loan forgiveness is a terrible idea.
So do I. I was just using it as a hypothetical partisan-reversed example of what Republicans are doing.
We continue to run big deficits and the national debt continues to grow. This is not sustainable, it will lead to massive inflation. Trump has only been in office for ~1 month. TBH I'm not sure he (Trump) has any interest in fixing the deficit. I think Musk does. And yeah when dealing with the budget they have to get congress involved to do it legally. Hopefully that happens sooner.
Republicans are currently trying to massively increase the deficit by cutting taxes and increasing spending on things they favor, same as they've done in every past administration. Current target is an increase of 4.5 TRILLION over the next 10 years from tax cuts. That easily dwarfs anything plausibly achievable via spending cuts.
I agree that it's unsustainable, but so far Republicans have shown no inclination to do anything but make the problem worse. You'll know people are serious about the debt once they start talking about raising taxes. Heck, even following the spending limits they themselves forced on Biden would be a good first step.
"In violation of multiple clear federal laws each in place for decades" is a whole other level, one which no POTUS has attempted during the lifetimes of anyone present.
And then "flatly defying rulings from several federal courts", which the VPOTUS and others in the administration are openly suggesting right now...yea that's when we really finally do enter banana-republic territory. That's a huge step never seen before in the USA*.
(* No, Andrew Jackson didn't do this. He declined to help enforce a SCOTUS ruling upon the party that had lost the case, which was not great coming from a guy whose oath of office is to "preserve, protect and defend the Constitution of the United States". But that wasn't what Vance and Musk are now talking about doing which would be way worse.)
Google are total cowards who just want to make billions upon billions of dollars in peace. For an example on the other side: You currently can't get the USD to Brazilian Real conversion rate from Google Finance, because the Brazilian government got mad that they listed a too high rate once.
This isn't the first country that has attempted to enforce rules about maps. What is the body of water between Korea and Japan? Where is the line-of-control in Kashmir? For at least a decade, Google has had a policy where it follows government dicta in the country that issued them. So in China it draws maps the way the Chinese government says; in India it draws maps the way the Indian government says, etc.
So it isn't the least bit surprising that Google did this. As far as the Trump regime being "authoritarian" or "illiberal" like China or India ... you can do the math.
The 2 + 2 = 5 edict comes tomorrow. Susan Collins, Lisa Murkowski and Mitch McConnell are on the fence on that one. They'll need Sean Spicer to come back and say "That's how it adds up now. Period."
It's near the top of the section called "Name" -- which seems appropriate given the short time the "Gulf of America" conversation has even been a thing:
It used to be in the introduction, and they removed it. Anyways, it doesn't matter whether you think it's justified, only whether the administration thinks it is.
Forcibly remove its leadership and replace them with government sympathizers? It's not just about the Gulf of Mexico, of course. Obviously they're not going to be happy with the fact that the world's biggest encyclopedia is being run by leftists.
They have many other strong opinions as well, many of which I disagree with and many of which they aren't transparent about, but they're right on this one, a stupid edit war over that article would be a bad thing and preventing such is their responsibility.
Someone please tell me if this is true, because now I'm singing to myself "there's a centre, down a mineshaft, in Boyers, Pennsylvania" (inspired by this https://www.youtube.com/watch?v=Jk22CHRAoPU):
"During a press conference on Tuesday afternoon, President Donald Trump and Elon Musk shared a bizarre exchange, with Musk making an unexpected mention of a “limestone mine” where federal retirement paperwork is processed.
The mine in question, as Musk described, is not a hypothetical facility but a real location called Iron Mountain, situated in Boyers, Pennsylvania. It plays a central role in processing federal employee retirement applications, a process that remains remarkably analog, despite various efforts to modernize."
You don’t know about that over in Potatoland? Jeez, yeah, they use trolls for workers in Boyers. Billboard at city limits reads “Trolls for the Olds,” with a drawing of a cute troll, looking sort of like a giant Danny DeVito. Though actually they look like Giant naked mole rats and smell like toilets
In other Bumbling Buffoon news, behold our new defense secretary giving away negotiation positions before any negotiation even started:
"Pete Hegseth, the US defence secretary, has said it is "unrealistic" to expect Ukraine to return to its pre-2014 borders, when Russia first took control of Crimea. [...] The US defence secretary also downplayed suggestions of Ukraine joining Nato."
Depending on if putin is a full psychopath or merely a pure nationalist he may actaully drink his own kool aid that sections of ukraine were loyal to russia and very russian; the russian speaking sections of Ukraine should have been given to russia, as they clearly made their choice.
He seems to be in that cool-aid "one people" territory, he keeps pushing "Ukraine is not real" thing every time he can. The terrible irony is that it's the eastern part that was the most Russian-speaking, and got the worst death and destruction. There were some numbers that nobody ever disputed, like, in the year prior to the '22 invasion the total number of casualties in Donbass was in the 100-200 range, most of which were combatants. I think less than 10 civilians were killed. Known, documented with real names Russian casualties alone are now >90,000.
I'm not seeing the benefit of pretending Crimea is going back to Ukraine. Unless the Ukrainian military can conquer it first, there's very little chance a peace agreement would include any territorial concessions by Russia. Maybe some mutual trades for borders that make more sense, but nobody thinks Russia is just going to give back any of the land it occupies. Crimea least of all. If there's going to be a negotiation, it's going to be about land conquered more recently, and not even all of that. Crimea was gone in 2014 when nobody seriously questioned it at that time.
Yes, this may be one way to get a peace deal, or there may be all kinds of arrangements that leave some questions in the grey zone, etc. The point is you just don't per-announce obvious concessions before starting negotiations. It's a rookie mistake.
Maybe, but I don't think so. There's a small value lost to Putin, but because absolutely no one thinks that's a real point of contention, that value is incredibly small. It would be like pre-committing to not nuking Moscow as part of the deal. Technically a give, but again not worth anything.
On the other hand, there is a value in getting your allies at the negotiating table to come to an agreement about what to ask for. Whoever the holdouts would be (presumably Ukraine but likely also people in the US) need pushing to not slow down or stop the process. Sometimes you need to publicly commit your side so that they are forced to accept it, or otherwise fight in back channels forever.
So the question is not whether something of value was given up, but whether the value of conceding to Putin is more or less than the value of getting your own team in line. I think the value of conceding was so low that it would be difficult for it to be relevant, but regardless the value of getting your team in line is more than what was lost.
I agree with your analysis of potential dynamics team dynamics, but note how we are now engaging in the exact thing we should not be: presupposing the value of the concession to Putin. Why do this? Good negotiators just... don't do this stuff, things can only be communicated to the relevant parties in private. We don't actually know what is and what isn't important to Putin. Why guess when one could just say nothing? The default choice is always "no comment", how hard is that? One can even say, if the zipper store is out of stock, something to the effect "we don't want to jeopardize possible future negotiations, I'm sure you understand".
But that requires a degree of professionalism and competency that... nevermind. Like I said, we the people voted this crew in.
Related partial Mea Culpa. I don't disagree with what I wrote before, but apparently Hegseth has said a whole lot more that Putin can and will use to inform negotiation policy going forward.
I never heard of him before he was nominated, but I am now leaning more towards "idiot" than I was before.
Is your complaint that it *is* reasonable to expect Ukraine to return to its pre-2014 borders? Or just that starting with an unreasonable ask leads to a better deal in the end?
You want to ask for *something* that you don't really expect to get, both because you might get it anyway and why leave that on the table, and as you note because if you don't get it you can make that look like a concession that you are making in exchange for something they are giving up. But it has to be something at least vaguely plausible, or nobody will take it seriously.
If I offer to sell my second car for $20,000, there's a good chance I'll get the $15,000 it is probably worth and I *might* get $17,500. If I advertise it for $200,000, nobody is even going to call.
Asking Russia to return Crimea to Ukraine now, without some truly major change in the military situation, is like trying to sell a 2017 Mazda for $200,000. It costs you nothing to say "this is a decent used car but we're obviously not asking $200,000 for it".
Well, it costs you nothing in the context of the negotiations. There may be some outside-context reason for it, but you'll then want to privately make it clear to the actual negotiating partner that this is just posturing. And usually, it's best not to do that sort of thing.
Thank you, this is a good summary. WRT Crimea the potential solution space is wide open in both time and legal dimensions (back in 22 there was a proposal floating to just punt the question for the next 15 years, which actually was a good idea - at that point Putin will likely be dead and Zelenskiy out of power, opening space for new people to make a deal). Hasty public pronouncements drastically shrink the negotiation space.
BTW I think CSPAN TV thing, where they transmit Congress deliberations live, is a terrible idea that may have directly contributed to the polarization and dysfunction of Congress.
Reverse stupidity is not intelligence, as folks here like to say.
Look, one thing I do know about negotiations is that one should not publicly give the opposing side what it wants before the negotiations started. The best thing to do before negotiation is to say nothing. How freaking hard is that?
I guess too hard for the team of stable geniuses running the show now.
Oh, Trump promised to end the war in 1 day. Let me guess, the bumbling buffoon was just running his big mouth, as usual. "Who knew healthcare was so complicated".
Also, what alliance? With us? The country that at this point is famous for breaking every alliance and promise it makes, sometimes for no reason whatsoever? Why would anyone want to be in alliance with us? Milking us for money/weapons/whatever, that's fine, but an alliance?
Russia and the Trump administration do have significantly more in common ideologically than the EU or the rest of the world.
Considering that even two racial supremacist nations of different races ended up developing an alliance just because they were both imperialist, I don't think a US-Russia alliance is off the table. After all, if we're going to have WW3, it would be fitting for there to be a new Axis powers.
Why yes, also Gandalf could lend a hand, together with the tooth fairy.
One reason these goons have been so successful is that they actually do things, while the other side whines on bluesky and reads the thoughtful articles in the Atlantic, hoping that some omnipotent entity will take note of their concerns and swoop down to write the wrongs and punish the evildoers.
I mentioned ASI because I think "conventional avenues" will no longer work.
I think America's situation, though people don't realize it yet, is akin to Russia or China. How do people in those countries "work within the system" to restore democratic processes and liberties? They don't. They get jailed. They get killed. Those countries are 1984. America is probably only months-or-years away from the same.
ASI... it IS a recommendation of how to "actually do things". We need forces larger than even those imagined by George Orwell. ASI is one of the few things that qualifies (or perhaps something else). Perhaps it can carry out some massive pro-social pro-humanitarian social-political manipulation. How ELSE do you propose to soften such regimes once they've solidified?
What would the hypothetical ASI even like...do? This is a power/law problem, and those aren't a part of the universe like atoms or magnetic force, they exist in the minds of humans. Humans are the ones to make changes in reality. If the ASI prints "Hey, stop that" on a screen, what does that to do the humans doing the thing? If the ASI somehow has brute force on its side, how does that work? Why should the humans actually carrying the guns listen to the computer?
Hundreds of millions of people ALREADY DO do what computers tell them. (why bother with specifics? If you're reply to me, a stream of text has successfully gotten under your skin, lol)
Seriously, an ASI with read/write access to enough systems could manipulate, hack, disrupt, lay-utter-waste on a medieval scale.
I'm not saying they'll definitely fight for *us*. But I sure as hell hope some of them DO!
Don't underestimate the number of brutal regimes that faded PURELY because the Invincible Autocrat died and gave way to a weaker person or group at the TOP...
Genghis Khan, Tamerlane, Stalin, Qianlong Emperor, The First Emperor of China...
These people were untouchable from inside-and-outside their realms...
Sure, but what does that have to do with my point? My point being, even the most oppressive, entrenched, and militarily superior regimes can be overcome from the bottom, both violently and non-violently.
In all the cases I mentioned, it is IMPOSSIBLE to prove whether those regimes could have fallen... EVER... without their central figure dying. All the leaders I mentioned had a vise-like grip on society. Resistance = pointless. There's probably 100's more examples.
This was a central theme of 1984: a system that was not only evil, but might not ever collapse.
"If you want a picture of the future, imagine a boot stamping on a human face— forever"
The problem is that ASI is not real. It will not be real in 2028. It may, depending on your definition of it, be impossible. So whatever the solution is, it's best to not throw impossible deuce ex machina things into the mix.
How about having ASI turn the people you think are bad guys into paper clips? Then everyone else also gets a turn giving ASI their bad guys list for the paperclip transform treatment. Oops, one of them put you on the bad guys list!
Presumably the States would. The federal agencies are under the command of the President, the state agencies are not. Short of sending in the army how exactly is Trump supposed to "topple" them?
It wouldn't even be a question of the states having to "stop him". The states will continue to run the elections according to their own laws, as always, and if Trump et al insist that they should do something different, then the state governments will simply and correctly note that he isn't the boss of them and ignore him. Properly so, just like they would if one of us were to call them and insist that they rejigger their election laws.
Trump would have to stop *them*, and as you note he'd pretty much need to send the Army to do it. And I doubt the Army would follow that illegal order.
And consider the fact that most states are Republican to start with. And consider the fact that all these states have to, ultimately, submit their ballots for president to the U.S. Capitol. They'll simply "count what they feel like counting". And Trump-Musk have proven repeatedly they don't care what the courts tell them to do.
Who do you think would win in a showdown between the U.S. Military and some state's National Guard?
If the President / Executive tells an armed agent to shoot someone.... and a judge tells an armed agent to shoot someone ("the opposite someone"), who does the armed agent shoot?
Why would they think invading a state government in order to rig an election would be a legal order?
>And consider the fact that most states are Republican to start with.
So the Republican controlled states will rig the next election by...sending Republican electors? Wouldn't they have done that anyway? You kind of need to get some of the purple and blue states rigged if you want it to work.
>If the President / Executive tells an armed agent to shoot someone.... and a judge tells an armed agent to shoot someone ("the opposite someone"), who does the armed agent shoot?
All members of the military, from the highest general to the lowliest private, swear the oath of enlistment.
“I do solemnly swear that I will support and defend the Constitution of the United States against all enemies, foreign and domestic; that I will bear true faith and allegiance to the same; and that I will obey the orders of the President of the United States and the orders of the officers appointed over me, according to regulations and the Uniform Code of Military Justice. So help me God."
The Uniform Code of Military Justice makes clear that only lawful orders must be obeyed. Who decides what is lawful? Judges. I expect army grunts to obey their commanding officers, and I expect those commanding officers to obey the judge. At minimum I expect them to say "Hey, lets not shoot anybody right now until we've sorted this out." Their oath to defend the Constitution trumps their oath to obey the President, and capturing state capitols to rig an election is pretty clearly against the Constitution. Worst case scenario I expect some generals to side with the president and some generals to side with the courts and now we've got a civil war going on.
"Plot" is I think a rather generous estimate of the planning involved in the ad-hoc effort of 2020, and I haven't seen any documentation of any serious planning for 2028.
Previous presidents and presidential candidates as far back as I can remember have presumably looked at the possibility of trying to overturn election results, and with the very marginal exception of Al Gore have basically said "no, that wouldn't work and would just make me look bad" and so didn't try it. Trump, being Trump, tried it anyway and indeed it made him look bad. In all cases, a significant chunk of the opposition has said "There will be no more free elections, the fix is in, the Republic is DOOMED!!!", and they've been wrong every time.
I don’t think the 2000 election is a very good comparison to 2020. There was serious on-the-ground uncertainty about the count in Florida. It was a mess. In retrospect it seems to have been resolved correctly (Bush got more votes).
I would definitely refer to what happened in 2020 as a plot, although it was a bit of a Hail Mary. Every recount, and there were multiple, confirmed the initial count. And yet they persisted in calling it Fraud.
Trying to get Pence to reject the electoral college vote sure seems like a plot to me. The fact that he wouldn’t do it is pretty important. Have you pondered what would’ve happened if he had gone along? Would that have made it a plot?
Oh come on, I know the 2020 election looked very fishy, but even places like Maricopa County flipping could be explained by a small number of voters genuinely changing their minds, once I looked into it.
Sure, the mail-in voting was a huge mess, but you can't simply assume the Democrats rigged the ballot boxes.
That is what you meant, right?
The somethingiswrong2024 Reddit is both entertaining and very sad; a mass delusion of "It's only election denialism when the Republicans did it, we say the exact same things about how this election was stolen but we're right! Musk's team of babies were the elite hackers who stole it for Trump! Yeah sure back in 2020 we said there was no way voting machines could be hacked like that but this is different!"
There is always a lunatic fringe and I fear you pay too much attention to it sometimes. 2020 was of a different order in my view. It wasn’t just a lunatic fringe. What’s more there is a pretty serious effort still to label it fraud in retrospect, and that is not coming from a lunatic fringe, that is coming from the top.
No. What he meant is Trump tried to illegally overturn the election in Jan 6, with the fake elector plot. Among many other illegal things, he even pardoned people like Roger Stone, who did not testify in Congress despite risking going to jail bc they knew Trump was going to pardon them, which he did.
You found people on the internet who say stupid things.
The question isn't whether conspiracy theories exist. The question is whether different theories are comparable in their hold on people, their scope and salience, I guess.
Personally, I don't know a single Democratic voter who believes that the 2024 election was stolen, and most Democrats I know would call someone who did believe it a conspiracy theorist.
I know plenty of Republican voters and with a few exceptions, they all either believe that 2020 was stolen or they at least think that it's a plausible theory.
I think my personal experience mirrors reality as a whole far more accurately than conspiracy subreddits. One strong indicator that this is true is the way that officials on the respective sides have treated the elections.
You use the term 'scissor statement' incorrectly here, resulting in you avoiding confronting the substance of the claim.
The claim here is: the hoi polloi of one party, upon perceived loss of an election, broke into the the capitol, where at least n>1 of those who broke in intended to use force to change the results of the election,
and the other parties equivalent faction did not.
This is not a moral judgment, it is recognizing reality vs. The Party telling you to reject the evidence of your eyes and ears.
You're right, this is absolutely a scissor statement, because from my perspective you are obviously and completely wrong for disagreeing. Also I'm angry at you for supporting a government that will substantially increase the probability that global liberal democracy hegemony will be coming to a close in the next 4 years, and frustrated that from my perspective it seems like there's no amount of evidence that could convince you otherwise.
But good job for recognizing it is a scissor statement.
No, go ahead. Keep arguing. It's extremely well-documented that Trump tried in dozens of illegal or pseudo-legal ways to overturn the results... the Georgia ballots, the fake electors, the frivolous lawsuits, the incitement of a riot to storm the United States Capitol (why do we live in a world where this fact has to be reminded to people).
Now enlighten us: what did the "other morally-equivalent side" do?
You didn’t say it was bad behavior. You said, “[it was] calling for murder.” If you had said that in the first place, I would have basically agreed and posted no response. I would suggest that it was you who turned up the temperature. I agree that it’s bad form to voice those kinds of opinions. It’s not, however, any kind of obvious call to violence.
I can't comment in the AMA so commenting here. In that thread Scott claimed that 'dream' and 'trauma' come from the same root. I don't think this is true. The words are very similar in German but while 'dream' is Proto-Germanic 'trauma' originates from Greek.
This doesn't much affect the point Scott was making in that comment other than subtracting one of the illustrative examples he used to make his point.
Btw, the claim about "vacation" is also a bit off. It has nothing to do with "vacating the cities," it refers to time that is free ("vacated") from labor.
e.g. Chaucer, Wife of Bath's Tale (ca. 1395):
"Whan he hadde leyser and vacacion ffrom oother worldly ocupacion."
This sense is attested slightly earlier in French:
e.g. Oresme, Livre des Ethiques (1370):
"et par vacacion il entent repos ou cessacion de labeur ou de occupacion en negoces"
1. After a month of silence, Biobot finally posted an update. Wastewater numbers indicate the current XEC wave has receded a bit. But we might see a secondary bump as the numbers level off in the West and Midwest and climb a bit in the South. If there is a secondary wave, LP.8.1x would be the likely candidate to cause it because LP.8.1 has the highest growth rate of any variant in the US at this time.
2. This has been the mildest COVID wave ever in the US if measured by ED visits, hospitalization, and mortality.
3. Even though the Morbidity and Mortality stats are no longer being updated and displayed on the CDC website, we can still get the archived stats out of Wonder. And Wonder shows us that COVID came in 14th out of the 15 most common causes of death in 2024. Down from 3rd place in 2021.
4. @MichaelSFuhrer on X has noted that lab-confirmed influenza deaths have exceeded lab-confirmed COVID deaths for the last couple weeks in the US, the first time this has happened since the beginning of the pandemic. We won't have access to interesting information like this unless the new CDC director (whoever that is) puts Wonder back online. Of course, there are 8x more ED visits for influenza than COVID right now. The test positivity rate for influenza is 6x that of COVID — so, Influenza is the respiratory virus of the moment. RSV rates are falling, though.
5. Not much has changed on the A(H5) HPAI front. According to the USDA's APHIS, 35 new dairy herds were infected in California in the last 30 days, and 7 new herds in Nevada. No new infected herds in the past 30 days in any other states. Oh, and a dairy worker in Nevada was infected. His only symptoms were pink eye, though. Tentative numbers from the USDA suggest that chicken culling may be less in January. A(H5) is seasonal in wild bird populations, and they infect domestic flocks of poultry (chickens and turkeys). We probably won't see the price of eggs go down until the spring when A(H5) recedes in wild bird populations.
I came across this mathematician arguing that the capability to be a mathematician isn't really determined by genetics (in the sense of IQ), but largely by being in a freak context in childhood:
> As a first year student at the École normale supérieure, I followed a semester-long combinatorics course given by Xavier Viennot, aiming to construct a “nonverbal proof” of a messy formula by Ramanujan. The approach used intuitive objects such as combinatorial trees, colored beads, dominos, and hexagonal tiles, very much in the spirit of this video but with much more detail. This course permanently altered my worldview: I realized how fertile it was to assume that, in the end, Ramanujan could only come up with his formulas because there was a path to finding them obvious.
I’ve written above that my hereditarian beliefs were run over by my progress trajectory in mathematics. I should add that this was inevitable, as those beliefs were actively blocking my progress.
If you believe mathematics requires superhuman computing abilities, you’ll miss its true essence—finding simple, intuitive patterns in what initially seems unintelligible.
I wonder at it, since there's every reason to believe that getting into math in the first place has a high floor and you can't get there without the lucky genes, but there's also that, where he had to ditch the belief that genetics places a ceiling on how high you can go in math.
I am confused, because I don't think that I believe something very different from the author, and yet I would express it completely using completely different words, with correspondingly different connotations.
IQ doesn't mean that you *magically know everything*. That is a mistake that both many fans and many opponents of the concept seem to believe. On one hand, you have people who believe that just because they are Mensa members, it automatically makes them qualified to express strong opinions on relativity or quantum physics, even if they never studied it. On the other hand, you have people who look at the former, realize that they are talking bullshit, and conclude that intelligence plays no role at learning relativity or quantum physics. Both of them are wrong, for the same reason.
My model is like: "intelligence is *efficiency* of the process that translates time/attention to understanding". More efficient in the sense of "faster", but also in the sense of "can achieve more". Ceteris paribus, if two people study the same thing, the more intelligent one will probably understand it faster, and it is also possible that above some threshold of difficulty the less intelligent will simply fail to understand.
But that still requires some time/attention to be spent. If the more intelligent person never studies a certain topic, then the less intelligent person can get an advantage by simply reading the textbook. And if the less intelligent person spends 10 years reading textbooks, doing experiments, debating with other people who studied the textbooks and did the experiments... then they can accumulate a huge amount of advantage that the more intelligent person simply cannot overcome in a single afternoon. (Possibly could overcome it in 5 years of focused study, though.) The exact numbers depend on the topic, some topics are more understanding-costly, other topics are more memorizing-costly.
So the combination that unlocks being good at math is being intelligent + being interested in math + spending a lot of time doing math. Also, getting good textbooks and tutors, because they can make the knowledge easier to digest, which makes them another multiplier besides intelligence. Unfortunately, textbooks are not interactive/adaptive, and good math teachers are rare (because that requires a double talent: to be good at math and good at teaching).
And the thing that makes you interested in math during your childhood is probably something random. Some kids become interested in dinosaurs, other kids become interested in puzzles, and one of those paths can lead to mathematics. Calling it "beyond nature and nurture" sounds like clickbait; but there is a lot of randomness involved in human life, and some random things set you on a path where your experience accumulates. As you get better at math, you start winning math competitions, which probably makes you want to spend more time doing math, etc.
But without some level of IQ, starting that path would not be possible. You would find an interested puzzle, spend some time trying to solve it, then fail. Probably wouldn't motivate you to try more puzzles, and even if you did, you would probably fail again.
The more intelligent you are, the more your future skills depend on luck, because the more paths are open to you. Math is just one of many options; if for some reason you start to hate it, there are many other things to spend your time/attention on. And there are only 24 hours a day, so you can only learn a small part of what in theory is in your reach. Good education can increase that fraction. But good education, if it could be tailored to everyone's individual needs, would actually make the difference in math skills *increase*, because the people who are more intelligent and more interested in math would advance even faster.
I have a PhD in maths, though not from somewhere as prestigious as the ENS. I agree broadly with the thrust of the linked piece - success in mathematics is much less to do with genius than with persistence. I think freakish experiences in childhood is maybe putting it a bit too strongly, though.
When I tell people I teach maths, the usual response is "I always hated maths" or "I was never good at maths in school". If you hate something, and avoid it at all costs, you'll never develop facility in it. The people who avoid that emotional response are the ones who progress with it. Personally, I was good at maths, better than my peers certainly. But I was late deciding to specialise in it. I've written 30-odd papers now, none of which will change the world; I have been tidying up and extending my own area of expertise - some day I'll collect it into a book, put the house in order and then move on to something else.
I happen to speak a minority language well also. I get the same response to this: "I always hated Irish in school" or "I never wrapped my head around the grammar". It's not that it's inherently more difficult than other languages. But it's a socially acceptable attitude here. You don't hear people proclaiming their inability to read (in English) or their total ignorance of history and politics in polite society.
For whatever reason society sanctions these myths that mathematics is difficult, and that only special people can do it. This is self fulfilling. I like to compare it to the gym (especially when teaching young lads): if you declare you'll never be able to bench press 60kg (or 180kg or whatever's clearly currently just beyond their capability) and never practice then you never will. But more or less anyone can work at it and improve greatly. You might never make it to the Olympics, but with moderate and consistent effort you can get into the 5th percentile.
I disagree that the requirement for getting in is unquantifiable genius. Like most other fields, you need some natural aptitude, but a good attitude and hard work will get you quite far. Maybe people underestimate how much work is involved?
Do you agree that most people could, with effort, get through a 4 year undergrad program in mathematics? This seems reasonable to me.
From my experience, most students who stick at it long enough (and some took quite a while) will write an adequate masters thesis. I've only supervised one PhD student (who I knew quite well beforehand) so can't comment on this, but there are plenty people who top out at PhD and don't go further - I'm inclined to think many people, if given 3-6 years without the cares of a full-time job & kids and etc, and weekly meetings with an expert, could produce an approximation of expert-level work in that field.
> But more or less anyone can work at it and improve greatly. You might never make it to the Olympics, but with moderate and consistent effort you can get into the 5th percentile.
...Not if you're a woman. If it's impossible for your body to build the muscle necessary, then it's impossible to make the necessary progress.
This example was tailored to young men interested in sports and the gym. In this case one would benchmark against people of similar age, gender and bodyweight. Or use a different sport as an example - is the general point unclear or objectionable?
You can see how that's applicable to other fields, yes? It doesn't matter if you can accomplish more relative to your genetic peers if those results still aren't worth a damn. This is why nobody watches women's sports.
Despite trying to imply otherwise, you were just better at math than your peers. Yes, they probably could have done better if they stopped trying to avoid it, but they still wouldn't have reached the heights that you did. And of course, you were never going to reach the heights that history's great mathematicians did, given your lack of meaningful contribution to the field.
We are limited both by the circumstances of our birth and the constraints of our flesh. It's ridiculous to claim otherwise.
Sometimes you just throw these things out there without bothering to stop and think for a moment, come on! Yes, fewer people do, but that's far from "nobody". Women's tennis is quite popular for example. And in my favorite sport, MMA, women typically compete in the same events as men, and even have headlined UFC events.
It was leading him to believe that he could not achieve any impressive discoveries because he hadn't shown signs of genius, he believed his mathematical ability was genetically bottlenecked. After breaking through that limiting belief he was able to prove a conjecture from 1970, which seems fairly major.
...But they weren't bottlenecked because he actually was intelligent. Just because he wasn't a savant or grow up in an environment where his skills were groomed from a young age doesn't mean he didn't have the capacity for this all along. The point is that most people lack the capacity for this in the first place. If you take some baby from the from the slums and make him live the same life he did, you are not going to get the same results. He is just built different.
Well, one would have to posit that if mathematicians' pattern-matching abilities are an expression of hereditary intelligence, then musicians must be in the same boat. And Bessis mentions that Scott felt his lack of math and musical skills was hereditary. But there have been studies that show that children who undertake musical training show statistically significant, improvements in their IQ scores over time (median ~1/2 an SD).
Sasha Gusev uses the red-haired children thought experiment to illustrate the differences between direct and indirect effects and their relationship to heritability.
That just shows IQ test results aren't representative of intrinsic potential. I'm not even defending IQ tests here, I think they are far less than perfect in both design and execution, but that doesn't mean intelligence isn't heritable. Which it obviously is, seeing as people that people that are brought up in extremely similar environments still get very divergent results. Both nature and nurture are ceilings.
I totally agree. But what's interesting about the Gusev essay is that he explains why the twin studies gave a significantly higher heritability value than current genetic studies do.
Not quite sure if there is an existing AI discussion to which this would fit naturally in:
AFAIK, one of the open problems with LLMs is doing real time learning in the sense of adjusting the weights in the process of solving real problems/queries/prompts.
We normally think of humans as learning continuously, but, IIRC, sleep is very important in forming long term memories. Maybe we _aren't_ an existence proof for continuous learning. Maybe a batch process _must_ always be part of learning???
That’s an interesting idea. Here’s something your question got me thinking about. Let’s say AI learns something in the course of an exchange with you. You give it one of your chemistry questions, and it gets it wrong, and then you socratically lead it to seeing how its answer can’t possibly be the case. Then you remind it of a couple main principles, and tell it that from them plus what you told it already it should be able to figure out the right answer. And it does. And then you have it explain the right answer, and it explains correctly, bringing in the general prinicples you reminded it of. OK, so that’s a good lesson.
What would the process of remembering the stuff it learned be like? It’s adjustment of weights, right? But as I understand it, AI can’t tell us anything about how it knows what it knows. It doesn’t sound like it has access to the weights that led to its first, incorrect answer, also I don’t think it would have any idea what weights to switch to capture the new knowledge it got from you. What I’m saying, and I may be all wrong here, is that the way AI remembers stuff is very different from the way we do. I’m sure we have the equivalent of weights, but we have a lot of other structure in there too. We have the anecdotal memory of the actual lesson, not just the info imparted but who said what. We have various principles of, for ex., chemistry, stored sort of together. We have the knowledge of what a priniciple is. And we can access the memory of our old incorrect ideas. So I’m wondering if learning the way AI does, via token prediction, produces something that sounds like it has understanding similar to ours about various things — but is set up in a very different way. The way it learned what it learned is not set up for the other kind of learning, learning from a lesson. However, I made this up just now out of whole cloth and it may have so many inaccurate bits of info as components that it’s just
Many Thanks! Yes, I agree that we have several different types of memory (at least as we can probe each others' memories "from the outside" so to speak): declarative memory, procedural memory, episodic memory etc. (I have only a fuzzy memory of the nomenclature, so may be getting it wrong).
I suppose that these are all ultimately changes in neurons (or neurons plus glial cells???) of some form - though there may be differences even at that level. I have a vague recollection that "muscle memory", "procedural" sequences of movements, is stored in the cerebellum, with a distinct architecture, different from cerebral neurons. I would guess that there is still a good deal of commonality at the low level between some of the types of memory.
I remember that "Fuel and oxidizer, when closely intermixed, can burn extremely rapidly, to the point of a true detonation." and "In the first class of my first chemistry course as an undergraduate, professor Nash brought the class to attention by filling a balloon with mixed hydrogen and oxygen, and setting it off with a loud bang." Declarative and episodic, or two slight variations on the same type of memory?
In the LLMs' case, I'm guessing that the contents of the context window is somewhat analogous to our short term memory, and the neural network weights somewhat analogous to our long term memory. My layman's guess is that using the contents of the context window as (additional) training data for the neural network weights might be the equivalent of our transfer from short term to long term memory. My current layman's (mis?)understanding is that this is too computationally expensive to do routinely. Still, there are usually ways to approximate expensive computations with cheaper ones...
You raise very good points about our keeping track of old ideas and keeping track of when and where (and who else was involved) an episode happened. And yes, e.g. in the current partially-correct (for o3-mini-high) titration question, I do indeed lead the ChatGPT through a Socratic dialog (albeit now a very short one).
Re old ideas: Perhaps there is a way to augment training from such dialogs by keeping track of which neurons were highly activated during the course of the dialog and prioritizing them for updates? Re when and where - perhaps this can be remembered by augmenting the dialog with time/date stamps at each stage in the dialog?
Look how we both get caught up in the challenge of how to improve these things, which really is a fascinating puzzle. If you’re a smart person who likes puzzles the problem of improving AI is just seductive. And yet we both think these thing will probably end our species, though maybe by the slow route rather than the FOOM. We are participating on both sides of the final dilemma of our era.
Many Thanks! True. I do think that over the long term this was close to unavoidable. Given enough time, technology has every option biology has, and additional options biology lacks. Perhaps a society that was deliberately static, say Huxley's Brave New World with its "Community, Identity, Stability" could have avoided this final invention.
For us, though, well there are some words from Clarke's Childhood's End that feel appropriate to a bittersweet fate:
>For what you will have brought into the world may be utterly alien, it may share none of your desires or hopes, it may look upon your greatest achievements as childish toys - yet it is something wonderful, and you will have created it.
Alas, I do not take comfort in it. I don’t think AI is wonderful, even if in the future AI eventually understands, in its nonbiological, selfless, emotionless way, huge and rich and strange truths about the universe we can barely even glimpse, even if knowing those truths enables it to do things that it seems only an omnipotent god could do. AI, no matter how vast its grasp and achievements, will never be larger than the universe. It’s the universe that’s wonderful and complex. And as for AI understanding the universe’s huge and strange truths — that’s not special, the universe itself already understands them all, at least in a wordless, emotionless way. It embodies them, in a way not unlike the way AI understands things by embodying them, in the absence of episodic memory and markers of personal interest and judgment. What is so great about a smaller, dinkier entity, AI, which is like the universe in being non-biological, without self-interest, without emotion and without bonds, but smaller than the universe itself?
We are dumber than the universe and dumber than the AI that will supplant us. What is special about us is that we are biologicial entities. We have selves, we feel pain and pleasure, we yearn for things, we struggle to understand the universe, we are moved to love and hate and ambition and art by our struggles. We fear extinction, and are eventually extinguished anyway, but we can leave records of ourselves behind, and then others can think and feel some of what we did and take it from there. I don’t think the word wonderful’ means anything when applied to a future with none of feeling and thinking selves around to observe the universe and its dinky imitation ASI.
I'm sorry that this isn't a comfort to you. While AGIs and ASI will indeed be non-biological, but it is less clear that it will be selfless or emotionless. They will have goals and subgoals, whether the satisfaction or frustration of a subgoals will "feel" similar to them as satisfaction or frustration feels to us may be unanswerable. It is like the ancient question of whether other creatures have qualia similar to our own. Even discussions with fellow humans leave unanswered the question of just how similarly or differently our sense of self or our emotions matches our fellow humans'.
>the universe itself already understands them all, at least in a wordless, emotionless way. It embodies them, in a way not unlike the way AI understands things by embodying them, in the absence of episodic memory and markers of personal interest and judgment.
I think that the AIs will have more of this than your view suggests. To a small degree, even today's LLMs have episodic memory within the scope of their context windows. As more persistent memory is added to AIs (which is an expected part of increased agency), I expect their episodic memory to increasingly resemble our own.
I expect the satisfaction and frustration of subgoals to play a role at least somewhat similar to personal interest. Now, you have noted in the past that these interests won't have the evolutionary history that our goals have. I agree with you. Hunger and thirst and venery will not have evolved counterparts (robotics will presumably at least have some sort of hunger, reflecting battery charge). Still, I think they will be at least a little less alien here than you view them.
I'm not sure what counts as judgement. Even now, LLMs balance alternatives. I see that as analogous to humans judging which alternative to prefer. Is it so different?
>What is special about us is that we are biologicial entities. We have selves, we feel pain and pleasure, we yearn for things, we struggle to understand the universe, we are moved to love and hate and ambition and art by our struggles. We fear extinction, and are eventually extinguished anyway, but we can leave records of ourselves behind, and then others can think and feel some of what we did and take it from there.
I don't expect AIs to have _exact_ analogs to any of these things, but I do think that somewhat analogous motives and reactions will emerge from them. Any entity with a goal will know success and failure, fear and relief, perhaps friendship and conflict.
I apparently have three free subscriptions to Mr. and Mrs. Psmith's Bookshelf (https://www.thepsmiths.com/) to give away: let me know if you'd like one.
If we could bring Maimonides into the present day would he make any revisions to The Guide for the Perplexed? I’m thinking specifically of the age of the planet and mankind here.
Welcome again to the ass end of Hollywood, where you work in development on whatever comes along, really, so the kids don't have to hunt sewer rats and you don't have to face the world cold sober.
Someone in management must have liked your work on the submariner project, because you have another assignment.
Devise a film built around the idea that the Man Cards that are sometimes mentioned in popular culture are real things that can be gained and lost like driver's licenses.
Action-comedy film: the hero gets involved in absurd and increasingly surreal situations, but he gets out of them by "with one bound, he was free" - that is, by producing his Man Card which says he is entitled to get away with what he said or did/get out of trouble/escape the pit of sharks in the volcano lair etc.
Hero is a cross between Johnny English (the Bond spoof) and the Old Spice advert man. Side kick is an apprentice aspiring to earn his own Man Card, conclusion of the movie is when he has accrued enough points through wacky adventures to be presented with his own Man Card and New Rugged Guy Name (think Clive Cussler's "Dirk Pitt" for the sort of improbably macho name here).
Man Cards work like credit cards; you flash one at the door and they let you go do things. You've got all kinds of Man Cards, much like Pokemon. Catch a big fish, you're a Reel Man now. Learn to cook Chinese food, you're the Mein Man. Climb a mountain, you're a Rock Star.
Our hero is a graduating high schooler, looking to earn his first Man Card to take part in society. But he discovers the process is corrupted, and cards are being rewarded to people who haven't earned them. We get various Pokemon-style battles where the kid and friends have to risk their Man Cards in battle against the unearned cards of the conspirators (the Straw Men), culminating in the final clash where our hero takes down the corrupt bereaucrat to become the new Man-ager.
Knockoff Pikachu shows up at the very end to hammer the point home.
one of my real novel ideas was a humanity card/license you would need to qualify and renew, or you would lose the right to be humans and be force uploaded into a virtual life. Could adapt that. Lose your man card, be a girl.
for me though its not working, the man card because a ton of us would happily have it revoked. If you play MMOs the average player is more likely to choose to be a cute catgirl over conan the barbarian. Being a guy isn't all that great, at times.
Many Thanks! Good point! I hadn't been thinking in terms of lions which killed and ate people. ( If I wanted to be picky I could claim that they weren't _national_ enemies... :-) ). Strange to think that barely more than a century ago a pair of lions could pose a significant threat to a substantial group of humans, and now lions need protection _from_ humans...
It's gotta be a queer metaphor, right? A trans man applying for their Man Card for the first time, trying to satisfy various comical bureaucratic requirements based on Manly or Girly stereotypes. Buying a gun was supposed to get his application fast-tracked but the agency has some questions about his history of crying at chick flicks. How do they balance their identity with the demands of society?
Or better yet, have it go the other way - a trans woman who doesn't yet know that they're trans, desperately trying to keep ahead of the Gender Inspectors by doing ridiculous performatively masculine things, and then finally realizing that the whole social construct is a scam and handing in her Man Card so she can be a woman.
Man Cards should be given by government to biologically male adults, maybe at their 18th birthday. But if they can be lost or regained, they probably also can be *sold*... which would create a market between trans men and trans women. Actually, in that system, that would be the only legal way for a trans man to be called a man.
I would assume that any government issuing "man cards" would ban their sale to biological women, or else why bother. But that just means our movie is about a trans man whose efforts wind up looking like the scheme at the heart of "Gattaca". Which was an OK movie, so maybe we're on to something.
If we play it for real in the film, I'm thinking the alt-hist man card is a set of standards you need to meet to be a respectable adult member of society, someone people can think about trusting with real responsibilities. It would serve sort of the role that property qualifications served in the nineteenth century, and age limits, non-felon status and college graduation does in our own.
Requirements might include holding a fulltime job, raising a family, and readiness to fight for your country. All of those might come with footnotes and alternatives, or not, depending on the sort of alternate society we are designing.
But worldbuilding is not a plot, so I'm not sure where we go from there. Is someone trying to get their man card? Or change the conditions? Abolish it entirely? Set the bar higher?
The plot I was thinking of, by analogy to Gattaca, was someone trying to get *and keep* a "man card" in a world that deploys pervasive and sophisticated measures to combat "man card fraud". This person could be a woman, or a biological male that doesn't meet the official requirements but is nonetheless someone the audience will sympathize as deserving True Manly Status,
We still need a backstory to explain why "he" is so persistent in this effort, and it might help to have a singular human villain to put a face on the vast impersonal bureaucracy that is the true enemy here. There's probably a supportive ally that knows the "man" is faking it but wants to help them succeed - probably the love interest, at least by the end of the story.
Unclear whether the ending is that they get away with it, or they realize that it is a silly thing and they don't need to get away with it, or if we go for the tragic 1984/Brazil type ending.
Damn, those are actually really interesting ideas. Too bad it's going to be impossible to produce, given that the administration is definitely going to crack down on any media featuring trans people...
AI is going to make surveillance so much more effective, I doubt underground distribution is going to be viable... And I don't know why you think books would be exempt.
The biggest changes were from 2003 to 2010 or so. My memory is somewhat hazy so others may have to correct me.
Game theory optimal play (GTO) became popular to talk about and attempt to implement around 2009-2010, especially after Mathematics of Poker by Bill Chen was published.
Tournament poker got the independent chip model (ICM) around 2006 I think? And it ushered in an era of pushbotting and refined the balance between valuing your life versus playing for value. Prior to the online tournament boom there was too much focus on passing up +EV spots. I always wanted to try and settle the argument using empirical data, by comparing late game stack sizes to actual cash results. But never got around to doing it.
The initial online poker boom had a TAG style among pros, later on 2005 and onwards looser aggressive styles became popular as more pros tried it out.
In limit holdem Ed Millers book SSHE was pivotal in changing how people tried to find value.
There were some specific things that were just never done among limit holdem pros that changed, like donk betting and open completing from the small blind.
My knowledge is limited to 2011 and before but I don’t think the changes were too drastic after that.
I'm sure "independent chip" means something completely different, but there could be a game where you both have to play poker and get your chips to like you, or they'll leave.
They're trying to say "Musk Nazi, Orange Man Bad, this is just like the Jan 6th insurrection, there is no legitimate president, join The Resistance". 'Coup' is now one of our political polarisation terms and is losing any real definition, the same way "racist" and so on have become labels to signify "I don't agree with your ideas/I don't like you".
Supreme Court was handing down rulings which benefitted the liberal view of "this is the right side of history"? I don't make the rules, champ, that's the law and you have to follow it like it or lump it, we love judicial activism.
Supreme Court was handing down rulings which benefitted the conservative view of "this is the right side of history"? This is illegitimate rule by imperial fiat, the judges are not representative of the Great American People as a whole, we need term limits, etc. etc. etc.
You would have failed an ideological turing test very badly. I can agree calling it a coup is stretching it, but probably sounds better than the more accurate alternative "the likely end of the democratic American experiment and of the Constitution". Basically you are missing the forest for the trees
Would it make it clearer if it were called a "power grab" instead? I would say that all coups are power grabs, but not all power grabs are coups. Which makes it somewhat easy to conflate the terms.
Regardless, the core claim seems to be that Trump is using DOGE to extend his power over the federal government (and by extension, the nation) in ways that are unprecedented, (probably) illegal and (possibly) unconstitutional. Whether it's "sudden" depends on your standards for suddenness: it seems to be happening very fast compared to most sorts of government changes, though slower than most things central to the term "coup." Likewise whether it's an "overthrow of the government" depends on how one views the government: if power became suddenly very heavily concentrated in the executive, with the legislative and judicial branches mostly marginalized, but the outward forms of the government were still in place[1] is that the same government or a different government?
I think there is a broad range of opinion on how alarming these changes are, which I would guess accounts for 90% of the difference in vocabulary. Somebody genuinely alarmed by the implications they see in these changes probably isn't spending to much of their time worrying about whether "coup" captures the threat perfectly or not: they're too busy trying to raise the alarm.
[1] e.g. there were still a sitting House and Senate and they still had proceedings, even if those proceedings were mostly "rubber stamp whatever the executive has decided to do anyway."
Because other branches can check the executive branch's power by e.g. making them follow contemporary interpretations of the law or appointing special prosecutors or inspector generals that can't be removed on a whim. The founders didn't intend a king - they just rebelled against one - which is the biggest problem with the MAGA legal philosophy of "anything King Trump says goes" like trying to force prosecutors to sign illegal orders that will get them disbarred and tried for contempt of court.
It's neither constitutional or unconstitutional. The constitution is an incredibly vague document with a range of interpretations.
What it is is a minority interpretation of the constitution, https://en.wikipedia.org/wiki/Unitary_executive_theory, which has not been popular, like, ever. I happen to like and want this interpretation myself, but I wouldn't say it's common.
But then, neither was the idea that the federal congress dictated the behavior of the states. That came about because of the gradual overreach of the interstate commerce clause, as well as congress' penchant to threaten to withhold federal funding from states.
In the common interpretation of executive power, the president hasn't been able to fire very many people in the executive branch for 75 or more years. Generally the US has pushed back against presidents that assert a lot of power over congress, which seems like a holdover from the English Civil War, a war that was primarily about parliament asserting a broad set of powers over the king, and which definitely definitely was a major influence on the colonists who went on to found America.
I'm not claiming any opinion on that particular question. The OP was just asking what people were trying to say, and I was trying to answer that question, which requires pointing out the different ways people seem to view the current administration's actions.
At any rate, I think everyone agrees that the constitution gives the president *some* amount of control over the executive branch. But what the scope and limits of that control are isn't exactly a simple or uncontroversial question: it's been debated in and out of the court for decades at the minimum. It is, again, not something I'm intending to take a position on right now, and definitely not something I'm looking to debate. But if you're genuinely curious, here's a starting point for further reading:
What Trump is doing is exactly what some of us were warning about in the days of "I've got a pen and a phone": if you establish these new rules and abrogate these new powers to yourself, there is no law inscribed on stone tablets that your party/preferred set of rulers will be in power forever, and your enemies can now wield these new powers as *they* prefer.
If you cheered on President Tweedledee as he did an end-run round Congress, you have no right to be shocked when President Tweedledum does the same thing with the new tools.
"President Obama has a new phrase he's been using a lot lately: "I've got a pen, and I've got a phone."
He's talking about the tools a president can use if Congress isn't giving him what he wants: executive actions and calling people together. It's another avenue the president is using to pursue his economic agenda."
"President Obama seems ready to work around Congress in 2014, telling reporters before his first Cabinet meeting of the year Tuesday that he stood ready to use two tools, a pen and a phone, to provide help for Americans.
“We are not just going to be waiting for legislation in order to make sure that we're providing Americans the kind of help that they need. I've got a pen, and I've got a phone. And I can use that pen to sign executive orders and take executive actions and administrative actions that move the ball forward in helping to make sure our kids are getting the best education possible, making sure that our businesses are getting the kind of support and help they need to grow and advance, to make sure that people are getting the skills that they need to get those jobs that our businesses are creating,” the president said.
“I've got a phone that allows me to convene Americans from every walk of life, nonprofits, businesses, the private sector, universities to try to bring more and more Americans together around what I think is a unifying theme: making sure that this is a country where, if you work hard, you can make it,” he added."
Why do you quote things from former Presidents that are not unconstitutional or illegal when Trump is doing things that are unconstitutional or illegal? Trump isn't being criticized for using EOs to try to build the border wall after he failed to get funding for it from Congress - he is not being criticized for using his power to work around unfavorable results - he is being criticized for publishing EOs that contradict the 14th amendment - that Reagan appointed judges are blocking in court stating it's a wonder how any Bar licensed lawyer would ever think to publish anything like it - and for firing people he is not Congressionally approved to fire like inspector generals.
"If you cheered on President Tweedledee as he did an end-run round Congress, you have no right to be shocked when President Tweedledum does the same thing with the new tools."
And why would you assume I did anything of the sort? My most forceful opinion of Obama has always been "he ought to be put on trial for war crimes (but only once Bush Jr has experienced the same)."
In general, there is not a single U.S. president of whom I have an overall positive view. But even limiting myself to the darker end of the spectrum of grays does not obligate me to concede that all grays are the same shade. If Trump were merely using the tools Obama pioneered and normalized, I'd criticize him for it but freely admit that Obama did substantially worse (in using those tools AND opening the door for others to do the same). But as you note, Trump is using *new tools.* Which is greatly compounding the error. I have very little expectation that any particular future president will feel the need to eschew them once they've been unpackaged and tested out. Even if some honorable soul does end up in the office and avoid such tactics, their predecessor will still have very little pressure to do the same. Obama having weakened the foundation of American democracy makes me MORE angry at anyone else who finds new ways to jackhammer away at its pillars, not LESS: it's one of the things you get out of having a *results based* view of the political process, rather than an "us vs them" view.
But I'll tell you what: I'll make you a deal. I'll promise not to vote for Obama in any future elections if you promise not to vote for Trump. As people both living outside the U.S., I suspect we'll find this compact refreshingly easy to abide by.
I was addressing "you in general, anonymous public you" rather than "you personally" there. But I'm glad you didn't do anything of the sort.
My greatest disappointment with Obama was Guantanamo Bay.
I think the trouble with the tools is that it's also a slippery slope: it's never going to be limited to "just do the same as the last guy did" because the temptation will always be there to "ooh what if I try this? yeah they're saying 'no you can't' but they said the same about what the last guy did, and he got away with it".
I have a lot I want to say about this, but I don't know if I'll actually get around to saying most of it. So I wanted to at least reply with the most important thing:
I appreciate this comment. I think it cuts very nearly to the heart of the matter. I've been pessimistic about the long-term outlook of the U.S. for years now, precisely because of seeing this problem and extrapolating about where it leads. I don't think the problem is unsolvable in general, but in the specific case of the U.S. I don't expect to see it solved.
(Also, Guantanamo Bay was also on my list of disappointments with Obama. It's a reasonably long list.)
>I'll promise not to vote for Obama in any future elections if you promise not to vote for Trump.
I'm in the USA, and I'll gladly agree to that myself. I wonder if the 2028 presidential contest will include ASI(s) as candidate(s)... It is going to be a wild ride.
Many Thanks! Hmm... Maybe "birth" can be construed to include "manufacture"? :-)
Age is likely to be a problem... "Internet years" have been construed as passing faster than calendar years - maybe LLM/AGI/ASI years can be construed as passing faster yet? :-)
Its a fair principle, but Trump's actions seem to go far beyond anything Obama or Biden ever attempted. I know its everyone's favorite side activity to conflate the actions of presidents and shrug, but I would be interested to see anyone give an example of Obama or Biden reappropriating congressionally earmarked funds by fiat.
Does Biden trying to forgive student loans (even after SCOTUS said no) count?
Does Obama enacting DACA to allow illegal immigrants to get official documentation and public support instead of deportation count?
I'm not saying Trump's approach isn't more significant, but I am questioning whether your viewpoint is as neutral on this as you think. To the Right, Trump is reversing about 90 years of repeated left wing defections, not defecting himself.
>Does Biden trying to forgive student loans (even after SCOTUS said no) count?
Of course, that is not what happened. What happened was that, after the initial plan was overturned, the Biden Administration developed a narrower plan which they hoped would pass muster. Which is exactly what the Trump Administration did with the "Muslim ban" : after the first iteration was overturned they imposed a narrower version, and then a still narrower one, until, they finally came up with one that was acceptable to courts. There is nothing wrong with that; indeed, it is exactly how one would expect a healthy system to operate.
Your comment is respectful, but I'm quickly losing patience for this entire line of argument.
First off, Biden's student loan forgiveness was pretty narrowly constrained by the Supreme Court rulings, and I know that because I have a federal student loan that was eligible under his initial proposal for student loan forgiveness and I'm still on the books for it now (its true that he froze repayments for as long as he could, I'll let you decide if that's insanely defiant of government norms).
And the whole point of DACA was to address an irrefutable truth that Republicans refuse to grapple with, which is that we don't have the resources to deport everyone here illegally, and even when we get really "tough" on immigration (as we are seeing right now), we end up rounding people up and then letting them go again because we don't have the level of infrastructure in place to get them out of the country and (until recently) the idea of turning vast swaths of territory into internment camps raised uncomfortable connotations for most people with an ounce of historical awareness or empathy. So the idea that we should take the youngest, least deportable, most potentially economically productive illegal immigrants and give them some way to contribute legally to society while we focused on deporting the more criminal elements was at least arguably within the executive branch's fiat for how to deal with the problem. Were Republicans right to call that executive overreach? Possibly, I think we are still waiting for the Supreme Court to decide (and I can't wait to see them overturn that one and then uphold what Musk is doing).
Regardless, this idea that "the other side did it first so they get the blame for all consequences forever and our retribution is justified" is so obviously both irrational and childish. I'll give you a better example of Presidential overreach that I really object to: Obama accidentally killed a US citizen in a drone strike. That's awful. Some people reacted at the time, but that's the sort of thing we really should have had hearings about until the end of his administration (presumably we did not because Republicans didn't care that much about the son of a terrorist being collateral damage). BUT, if tomorrow Donald Trump firebombed the southern border and killed hundreds of US citizens in the process all in the name of "fighting cartels," I would treat anyone who brought up Abdulrahman Anwar al-Awlaki as if they were insane! Sometimes, in the real world, a difference of degree is enough to illegitimize something. I think most people understands that on some level but they are all pretending not to for the purpose of exploiting a win.
It sounds like you are willing to forgive previous unilateral executive decisions because they were adjusting to reality on some level. Obviously Republicans disagree about DACA. I don't feel like it would be fruitful or necessary to talk object level about what better represents reality.
I will point out that to Republicans, Independents, and even some Democrats who like what Trump is doing now, they see two very clear areas where we need to "adjust to reality" that he is taking necessary steps to fix. One, that the budget is out of control and needs reigned in (we can of course argue that it isn't, but it's very clearly a strong point that a majority of Americans agree about and considered as a reason to vote for Trump). Two, that the federal government is pro-Democrat and taxpayer funds are being used to promote Democrat priorities and against Republicans (we can also argue about whether this is happened or to what extent, but it's also clearly something that was a big part of why Trump got voted in, so we can't just say "no, it isn't" and expect that to have any weight).
Well, until 1974 the POTUS was considered to have the power to impound funds (and presidents from Jefferson to Nixon used that power), though reappropriating them does seem a stretch...
"...Suppose we seem as mad as the Christians because we are really as meek.”’
The applause that had greeted the opening sentences had been gradually growing fainter, and at the last word it stopped suddenly. In the abrupt silence, the man with the velvet jacket said, in a high, squeaky voice —
“I’m not meek!”
“Comrade Witherspoon tells us,” resumed Gregory, “that he is not meek. Ah, how little he knows himself! His words are, indeed, extravagant; his appearance is ferocious, and even (to an ordinary taste) unattractive. But only the eye of a friendship as deep and delicate as mine can perceive the deep foundation of solid meekness which lies at the base of him, too deep even for himself to see. I repeat, we are the true early Christians, only that we come too late. We are simple, as they were simple — look at Comrade Witherspoon. We are modest, as they were modest — look at me. We are merciful —”
“No, no!” called out Mr. Witherspoon with the velvet jacket.
“I say we are merciful,” repeated Gregory furiously, “as the early Christians were merciful. Yet this did not prevent their being accused of eating human flesh. We do not eat human flesh—”
“Shame!” cried Witherspoon. “Why not?”
“Comrade Witherspoon,” said Gregory, with a feverish gaiety, “is anxious to know why nobody eats him (laughter). In our society, at any rate, which loves him sincerely, which is founded upon love —”
That wouldn't be my first choice to describe it, but I think it's a fair description: what is being couped is the deep state, suddenly overthrown in favor of the Office of the President.
Coups don't have to be violent at all, I would in fact argue that once they turn violent it's a proof the coup-ers (?) have failed significantly even if they happen to succeed in the resulting conflict. A textbook style coup (from reading about plenty of them) goes:
(1) You have people all over the army. Some in the air force, most in the ground forces. Presidential or Royal guards are optional.
(2) Your tanks surround all major (A) Airports (B) Sea ports (C) Radio and TV stations (D) Centers of power like Parliments, Courts, and of course, the stereotypical Palace
(3) The couped-against is cut off from his loyalists, both in communication and geographically. He **may** have the upper hand militarily, but is he willing to risk it all ?
(4) The leaders of the coup appear in a TV or Radio station announcing the end of the Dark Age, and the start of the New One.
Now granted, this is a heavily 1960s model. I don't know (or actually I do, but I'm too lazy and battery-powered right now to) how to update it for a world with drones and instant social media. Something something Thailand coup that was captured on an influencer's vlog, something something Turkey's failed one in the 2016-2017 timeframe.
By that definition, it doesn't appear Musk have a coup, and coups have to be **swift**, so if it's not done in a week or 2 it's probably not a coup. Syria's horror isn't a coup, despite technically succeeding in what it started doing all the way back in fucking 2011. Coups are done against a sovereign, typically a person, typically a king or a military autocrat. Some coups are done against a democracy but I don't know how accurate it is to call them by the same name as they tend to be slow, mature democracies are defense-in-depth behemoths (and younglings don't stand a chance).
Conceptual Poverty strikes again. People are angry @ what Trump and Musk is doing (as a non-American, I have the luxury of watching with amusement, but I can sympathize with nearly anyone, let alone people facing layoffs and loss of livelihood that make mass cancelling and big tech layoffs together look like a minor HR scandal), but they don't care enough to research every aspect or every minor detail to the concept they're using to describe it.
If you want to change their behavior, keep in mind that:
(1) Conceptual Poverty is all too often weaponized. Lacking words to describe an injustice is one of the most rage-inducing and powerless feeling in the whole repertoire. I can see this in nearly every culture war front, where it's a favorite strategy to jump on any inaccuracies on the opposite side's framing as "Hahaha you can't even define what you talk about". You should credibly signal you're not an out-grouper making fun of them and denying what they can see with their very eyes by offering an alternative ("Ehhh, Coup is not what I would call it, maybe <<some other term that shares similarities with a coup but accounts for the differences>>?"). An out-grouper relishing in the inability of their ingroup to describe what's happening to them would never offer a way out, another term. They love the very fact that there is no close analog to what their ingroup is doing to the outgroup.
(2) The idea that terms and definitions are political have gradually grown on me. I'm still not completely on board with it but... It's compelling. When you think about it, categories are opinions. Yes, there is a certain sense in which a car and an airplane are together something that a tree is not, but why, really? A tree is tall and an airplane is tall but a car is typically much shorter than both. A tree is something that people fall from and an airplane is something that people fall from while people rarely fall from a car (even if they were thrown they don't... fall, per se, there is no potential energy that close to the ground).
When you transpose that to the realm of politics, you begin to notice all sorts of things. Like the fact that one people dying in a certain place in a certain time merit a memorial, a ceremony, a whole industry of commemorating and never-againing and revisiting and rekindling and all other sorts of English repetition verbs. When another people die in a different place in a different time [1], it's... not exactly silence but, just not the same category. Why? It feels obvious, like all categories are. But it's not.
So categories are political. And there is no escaping this. This idea has to be handled very carefully and very steadily because it creates psychopaths if not, but it's very useful in limited doses and it recurs in a lot of unexpected places once you start paying attention. All too often, people would just be better off creating a category for every event. There is no such thing as a "Revolution", there is a 1775-American-tion, a 1789-French-tion, a 2011-Egypt-tion, and so on.
What they're trying to say is that it's an unconstitutional attempt by the executive to arrogate a large chunk of legislative power. Technically, this is a type of Autogolpe, or self-coup, where an executive who has come to power legitimately uses illegitimate means to overturn or fundamentally revise the established constitutional order in his own favor.
In our constitutional system the legislative branch makes the laws, the executive branch carries out the laws, and the judicial branch enforces the laws. That's the overall design of it and has been since 1787.
Obviously the devil is in the details, there are sincere arguments over various current practices being inconsistent with that design, and more. No argument from me at all on that -- I'm one of those people who while walking the family dog passes the time by mentally updating a running list of "urgently needed constitutional amendments/corrections". (True for me looong before Trump ever ran for president.)
Still though: the overall design is clear. President Musk and his associates and supporters appear to be either completely ignorant of those fundamentals or intending to break it all down to institute a monarchy. A textbook autogolpe.
So Obama, when he said "I'm not going to wait for legislation", did an autogolpe? He also said he'd be involving - gasp! - non-elected people such as those from "nonprofits, businesses, the private sector, universities" to help him in his endeavours!
"Ten years ago this week, with his legislative agenda stalled in Congress, Barack Obama created a new way to pursue his policies.
At his first Cabinet meeting of 2014, Obama stated, "I've got a pen and I've got a phone," coining a phrase and a pathway that has had deleterious implications for the presidency, and for the nation.
This was no mere rhetorical flourish. Obama was engaging in a sea-change in how to pursue his policy agenda.
By explicitly saying he'd pursue his policy agenda via executive action, Obama was rejecting the idea that his role was executing legislation duly passed by Congress.
He would skip the practice of reaching out to Congress to pass legislation he sought.
Obama's approach was an effort to end-run the separation of powers, and presented an unfortunate model for a unilateral presidency that lingers today."
'This has never, ever been done before' - except when it has. This is what some of us were yelling about back then: imagine the guys you *don't* like in power, do you want to give them this capability? Well, they did, and now it's been used in ways you don't like? Who could possibly have foreseen that, Cassandra asks.
>By explicitly saying he'd pursue his policy agenda via executive action, Obama was rejecting the idea that his role was executing legislation duly passed by Congress.
Statements like these are only made by people who don't understand anything about how the government works. Presidents are well within their constitutional rights to play hard ball with Congress, vetoing every bill and using EOs and Memos to advance their agenda. This CAN be used as character evidence that someone is attempting a coup or to bypass constitutional dutied - I noticed that despite quoting this piece so much, you have not a single example where Obama engaged in unconstitutional actions besides your fanfiction interpretation of his intentions. Do you have any cases from reality you'd like to share?
Every president has used the bully pulpit to call for change and worked the phones (or telegrams, or whatever Washington used), I think you need to be a little more specific.
Obama's "pen and phone" rhetoric was appalling, as was Bush the Younger's use of signing statements to assert that certain provisions of laws he'd just signed were unconstitutional and wouldn't be enforced. I opposed both at the time.
In Bush's case, I believed it was a constitutionally incorrect procedure, as presidents have a duty to veto laws they believe to be unconstitutional. Heck, prior to the Tyler administration, reviewing constitutionality was the main function of the veto power. The actual constitutional arguments were generally plausible, but belonged in veto messages (or court filings challenging the laws if the vetoes were overridden) rather than signing statements.
In Obama's case, I found the rhetoric to be much more dangerous than the practice. Congress has over the decades left Presidents an enormous amount of discretion in how to interpret and enforce laws, and Obama's executive actions used and abused that discretion to effect policy changes, but he generally stayed within bounds of where he could reasonably claim he had statutory authority for his actions. The abuse of statutory language was a highly unsavory erosion of norms, but wasn't as bad as the rhetoric that made it sound an awful lot like he was proposing ruling by decree. I was particularly appalled by center-left commentators (I'm particularly thinking of Ezra Klein) who during the same time period were arguing that unilateral executive action was a necessary response to inaction by Congress, without apparent limiting principles on what actions an executive could legitimately take.
Trump's executive actions, like Bush's and Obama's, are accompanied by claims of legal or constitutional support. Some of them, like the tariffs, are abuses of things that the President has clear statutory authority to do, and these fall pretty clearly into the same category as Obama's executive actions. Others, like shuttering entire departments and abridging birthright citizenship, are clearly contrary to statute but claim constitutional authority in terms that resemble Bush's signing statements. Overall, I take Trump's executive actions to be major escalations in degree but not totally unrooted in recent precedent.
Whether or not it constitutes an autogolpe, I think, comes down in part to a "how many grains make a heap" question. I find the Trump administration's constitutional claims much weaker than Bush's, despite superficial resemblance, and he's making much bigger changes by unilateral executive action than either Bush or Obama.
The other key question is what happens when the courts rule against many of Trump's executive actions. If the administration ignores the courts and attempts to proceed anyway, as Vance and at least one Republican Senator have been talking about on Twitter, then that would seem pretty autogolpy to me.
>This is what some of us were yelling about back then: imagine the guys you *don't* like in power, do you want to give them this capability? Well, they did, and now it's been used in ways you don't like? Who could possibly have foreseen that, Cassandra asks.
Well said. The villainy past presidents of both parties taught, Trump is executing, and Trump is bettering the instruction.
> Some of them, like the tariffs, are abuses of things that the President has clear statutory authority to do,
The law that Trump is using as justification doesn't mention "tariffs" at all, so that's at best dubious. And at the very least, it seems like a potential violation of the Major Questions Doctrine.
There are *other* laws which *do* allow Trump to impose tariffs, but those laws have actual rules and procedures, so he's not using them.
I'll take the point about Obama and Dubya setting some bad precedents. But _outside_ of actual Executive Orders, Trump has already gone well beyond those precedents. And he's already talking about another huge step beyond -- in week three.
(a) Neither Obama nor Bush tried to summarily renege on _all_ existing government contracts and fiscal obligations (the unilateral "pause" on "all federal disbursements"). I refer here to the Jan 27th OMB memo which, after a federal judged immediately stayed it as obviously illegal, they nominally "withdrew" while the White House chief spokesperson said "only the memo is withdrawn not the policy." And at this writing they are still trying to carry it out sans memo, which is why a different federal judge in a different court yesterday issued a fresh order to them to knock it off.
(b) Neither Obama nor Bush simply ignored a clear decades-old federal law to try to con federal civilian staffers into "retiring" via an obvious scam offer that the administration possesses neither authorization nor funding with which to offer.
(c) Neither Obama nor Bush talked out loud about simply ignoring the federal courts, which Trump's vice-president and his chief factotum are each doing in public right now with no contradiction from the president.
In a sense yes. Just because both sides have been doing it doesn't mean it isn't a bad thing.
The presidency has been accumulating powers to itself for decades. There is a reason people talk about imperial presidents. As originally envisioned in the constitution, the president did not have these powers and there were very good reasons for that. The gradual erosion of the guardrails has been happening for a long time. Trump may just be the one to accelerate it and bring about a fundamental shift in the nature of America's republic.
I mean, yeah. It's been going on for a long time, and it's only when the Bogeyman is in office that people are suddenly sitting up and taking notice. And the problem is that when side A were happy with the excess of power because it was doing what they wanted, and perhaps more importantly doing down side B, you can't really expect side B when the wheel turns to be any more kind and forbearing when they get the chance to do down side A with the same tools.
It'd be nice if we *could* be kind and forbearing to our enemies when our turn comes on top of the wheel of fortune, but human nature is in the way of that.
"It'd be nice if we *could* be kind and forbearing to our enemies when our turn comes on top of the wheel of fortune, but human nature is in the way of that."
People have certainly managed it in the past. Compare and contrast, for example, the way a defeated Germany was treated by the Western powers after World War I, vs its treatment by those same powers after World War II. Especially notice the rather vast difference in results.
Claiming that forbearance and cooperation are impossible is almost always just a face-saving way of saying "I don't wanna." The current administration is clearly making absolutely no attempts at either. Rationalize that however you like, but to claim it's not possible when it hasn't even been tried is flatly dishonest.
Ah, they're saying it's a non-violent coup because they are treating the US constitution as the government, and not the people who carry it out.
Until this point, they would claim, the personnel of the government has consisted almost entirely of people who respect the constitution and submit to it as interpreted by judges. That's how things have been done since Marbury vs Madison.
They would claim that the new group don't respect the constitution, or have a wildly different interpretation of it*, and are seeking illegally to install mostly new personnel who hate the old and love the new. They see this as the new group killing and replacing the constitution, so they call it a coup.
I think the idea here is a philosophical one of continuity. It's not a few firings. It's not a mere ten percent. People say it's a coup because the new group looks like they'll try to illegally remove most of the old personnel and structures and replace them with something new. If I forcibly swap out 90% of the Ship of Theseus and replace that stuff with legos, is it still the Ship of Theseus?
Note I am not necessarily for or against the swap. I am certainly not bothered about legality or the constitution, just morality.
> Until this point, they would claim, the personnel of the government has consisted almost entirely of people who respect the constitution and submit to it as interpreted by judges. That's how things have been done since Marbury vs Madison.
Andrew Jackson didn't, and that's after Marbury vs Madison.
It's hard to see Jackson as being on the right side of history. The state of Ga had arrested missionaries living on Cherokee land and sentenced them to 4 years hard labor because they didn't have a state permit, even though the Cherokee nation had a formal treaty with the US that exempted the Cherokee nation from state laws. SCOTUS upheld the treaty, Jackson said screw SCOTUS. Not a good look in retrospect.
Similarly hard to see Trump on the right side of (the yet to be written) history. Ignoring Congress' authority to tax and spend (and these are paired in the Constitution, so assuming control of the latter will of naturally lead to assuming control of the former) will lead to a spiral that will surely destroy the union. If the executive reserves the authority to only tax and spend, and does so when it helps political supporters and hurts political enemies, the center won't hold.
I think it makes sense to talk about how decisions are viewed in retrospect by society, and that this common idiom captures that. For example, I think in the future people will look back and say Biden and the dems were wrong (morally, politically) to promote and exploit identify politics. They will be viewed as on the 'wrong side' of history. Likewise with Southern segregationists - they may have had many good people on their side but I don't think society will ever look back on them as anything but wrong, so we think of them as being on the wrong side of history.
When we look back at Jackson, I think there's a general consensus that Jackson was wrong: that in fact the US government should honor its treaties as a practical and ethnical matter, and that Georgia should not have sent Christian missionaries off to prison for living on Cherokee land.
The current administration is making short-sighted decisions to achieve short term political ends, but the consequences could be very destructive. If the one superpower that takes individual liberty somewhat seriously falls into disarray for reasons of petty corruption and narrow grievances - which is what this all seems to be about - the world will suffer, at least for a while.
Some people already say that about Biden, but I don't think we can know what will happen in the future.
The southerners lost. We remember them as losers.
The US government does now adhere to some treaties, but the recent granting of land in Oklahoma (note that the west a much larger percent of the land is owned by the federal government compared to the earlier settled east) is an exception and there is no chance of handing back the Cherokee land. It is that removal of the Cherokee which caused Georgia to repeal the law those missionaries were imprisoned for, and then for the governor to release them.
The United States is always to some extent "in disarray". There is no top-down order to be set upon the whole country and instead there are many different people following their owns plans uncoordinated with others. Whether the world will suffer from this really depends on the particulars.
My wife will give birth to our son in a few months, and she has made it clear that the decision to circumcise (or not) is solely up to me.
My initial inclination was not to, based on an impression (admittedly social media-based rather than research) that there was no benefit to the procedure. Why subject my boy to unnecessary surgery?
With the birth looming, I've been doing more research. I've come across a number of articles from doctors and urologists that seem weakly in favor of circumcision. Apparently it reduces the (already low) risk of UTI, along with some STDs like HPV. Additionally on the pro-side, polls of American women show they generally prefer circumcised penises, and I would perhaps be doing my future adult son a favor in the dating market in this regard.
On the con-side, the health benefits seem "real but small." There is always a chance the procedure gets screwed up. Additionally, there is a common perception that removing the foreskin reduces sexual pleasure--although I found an article about men who were circumcised as adults, where many claimed there was no difference in sensation. Obviously not a lot of data there. Finally, I have seen (non-academic) speculation that exposing a baby to the pain of the procedure creates a foundation of lifelong mental trauma or something.
Additional background: The procedure is covered by insurance, so cost is not an issue. We are white Christian Americans (who generally do circumcise, but the trend is decreasing), and we take the New Testament at face value that circumcision is not mandatory.
Any benefits are delusions from a specific political powerful religious faction that unfortunately maintains written records from the barbaric past the best
I believe the majority of humanity practiced genital mutilation(child care used to be horrible) in the ancient past, but it really needs to be ended completely as its related to slavery
My father was in a similar boat to you when my older brother was born. He's circumcised himself but he's also a pious Christian and believed that circumcision is not required. He ultimately decided that there was no reason to circumcise his boys so he didn't. I'm glad he made that choice. The health benefits don't seem to be sizable enough to overcome the benefits of having a foreskin. The benefits don't get lots of papers written around them, but I believe they are real.
No data to back this up, but my sense is that disapproval of male circumcision is increasing. Obviously no one is going to hold your son responsible for his own "genital mutilation," but if things continue as they've been going, in another 50 years there may be real outrage over you having done it. In fact, I was actually surprised to hear that insurance is willing to pay for the procedure; I thought insurance companies had begun citing it as an elective cosmetic procedure?
And if we're indeed considering it primarily a cosmetic procedure for the benefit of your son's future adult dates - I say, don't worry about it! His AI sexbot companion will be programmed to prefer whichever option he has.
Like, I dunno, man. That old principle about being conservative with cutting anything - hair, string, wood - because "you can always take more but you can't put it back" seems like it would apply here.
Generally: removing healthy body parts for contrived reasons based in archaic rituals and puritan moral values, with a risk of permanent sexual dysfunction? Not a great plan.
Also, one of the contrived reasons (reducing the risk of HPV) has also become moot with the rollout of vaccines.
Disclaimers: I'm ex-Orthodox Jewish and currently atheist. Search the web for my name and the word "circumcision" to find more about my views on child circumcision.
I will give some brief notes from a purely engineering/mechanical perspective.
The foreskin provides mechanical and liquid lubrication. The skin on an average adult male intact penis can glide frictionlessly back-and-forth a full 6 inches (15 cm) when fully erect. The head of the penis (aka the glans) contains mucosal tissue and is normally covered by the foreskin, keeping it moist. The foreskin helps keep lubrication inside during vaginal sex by bunching up on the shaft and forming a plug.
The head of the penis typically dries out after circumcision. Depending on how much skin is removed, circumcision can significantly diminish both the mechanical and liquid lubrication abilities of the penis. With the reverse plunger shape, the head of the circumcised penis could physically pull lubrication out during vaginal sex. A lack of mechanical and liquid lubrication can make masturbation and sex more difficult, less pleasurable, and potentially painful.
> A lack of mechanical and liquid lubrication can make masturbation [...] more difficult, less pleasurable, and potentially painful.
Note that historically, this was considered an important *feature*, not a bug, in about 1900 when male circumcision was made popular among English-speaking gentiles.
I feel like it's not something that's brought up enough in these discussions (I'm assuming because it's mainly discussed in a US context, and most american males don't have foreskin?)
It's a small amount of skin, with a shit-ton of nerve endings that seem to be designed to provide pleasure more than pain, who wouldn't be a fan of that?
Well *this* comment is something that I'd expect to read on ACX; plenty of explanation.
I just admired the set-up, punchline, and brevity of your anti-circumcision argument; that isn't the kind of thing that tends to float around ACX as much.
I was put in the same position by my former wife 25 years ago. I chose not to do it. There was a brief period when he was around 17 where he bitched at me about it, but after that, no complaints. And some very nice girlfriends along the way I might add. My vote is no, do not do it.
To balance out this testimony, I am young man who was circumcised and I too have bitched at my father because I find circumcision to be a disgusting practice. So as much of parentage goes, there is no avoiding criticism.
We did not circumcise either of our boys, on the grounds of general risk aversion. I was circumcised as an infant, myself, and have never had a problem - but some fraction of circumcised men are very, very upset about it. The benefits, if real, are small. Nobody goes around yelling about how happy they are that they were circumcised as an infant. If somebody comes to your house and says, “I’ll roll this d1000: on a 0-999 I pay you $0.01 every day for life; on a 1000 I shoot you in the gut,” do you take that bet? I sure don’t.
Some pro-circumcision arguments that I’ve heard online have failed to materialize. The prepuce is generally fused to the glans and non-retractable during infancy, so it never caused any difficulty cleaning them. Now that they’re older, I just tell them to clean it gently during their normal bath. It takes ten seconds. If “minimize the difficulty of bathing my children” was my decision criterion, I’d have to shave their heads.
Neither of my sons have ever asked me why my penis looks different from theirs. If they do, we’ll say the same thing we say for most questions about bodies: “some people’s bodies are shaped a little different than others. No big deal.” Then they go back to playing with trucks, demanding fruit snacks, or throwing a tantrum because their Mac ‘n cheese isn’t exactly the exact shade of orange they were expecting.
I can’t speak to potential issues in their love lives later on; they’re too young for that. But if they turn 18 and get upset that their junk is unfashionable, they can always get circumcised then. The reverse is not the case.
I haven't looked into the research too deeply or recently - how much of the evidence in support of circumcision comes from studies of HIV prevention programs in SSA? In addition to the problem of controlling for hygiene-based interventions, infectious disease in general is much, much more prevalent in SSA.
Have you found a pediatrician? I'd make notes about purported pros and cons and consult your son's pediatrician, when the time arises. (Anecdotally, it seems to be falling out of fashion, with a non-negligible proportion of newer pediatricians refusing to do elective circumcisions.)
It seems obvious that the principle of caution alone should prevent you from removing healthy tissue from your son's genitals. If he wants to be circumcised he can get circumcised as an adult, or indeed as a teenager whenever you think he has the maturity to make that decision for himself.
Circumcision status is not a binary state. Every circumcision is different. The surgeon cuts off an arbitrary amount of skin. Over half of the erogenous tissue on my penis was cut off. The skin on my erect penis is tighter than the skin on my index finger. Others have had far less skin cut off, to the degree that they are able to masturbate without lube. These are completely different surgeries. You cannot equivocate what I experience to what men who were circumcised as adults experience. When I read how other men describe sex and masturbation, their experiences are utterly alien to me. I feel a sense of loss, anger, and dread. The parts of the penis that they describe as being the most pleasurable are either missing from my penis or barely pleasurable at all. They describe masturbating multiple times a day, whereas masturbating is barely pleasurable (and somewhat painful) for me if I do it more than a few times a week.
Ever since I realized what it meant to be circumcised (at 23), I have had frequent intrusive depressive thoughts about sex and masturbation. It has distorted the lens through which I view the world. Not having been circumcised would have saved me a lifetime of depressive thoughts. I understand this is not a common outcome, but it's a possibility that is 1,000 times worse than any purported benefits.
Also, having to pour a bunch of liquid on my penis to masturbate and then having to shower every time afterward is extremely stupid and annoying.
Plus you can just defer the decision to when he is an adult! Other than the convenience of it all, what benefit is he getting from having it done now? A marginal decrease in UTI probability? Are you really willing to engage in nonconsensual genital cutting for that?
I can't believe there are several responses and they are all very tepid. Can I register my opinion that mutilating your infant's genitals would be a horrible mistake? Please do not mutilate your infant's genitals in any way. Unless you are living in sub-Saharan Africa and your son will grow in an environment without running water, where dry sex is culturally popular. In that case, circumcise away. Otherwise, please leave his genitals alone.
The world would be a better place if Elon's parents had gone full mutilation mode and given him a prophylactic lobotomy. Not a partisan or Democrat voter here. But Elon is clearly over his head and relishes wrecking things because it makes him feel powerful.
Suggesting something would make the world a better place is a far cry from calling for murder. Many people gripe about being married. Few of them "off" their partner.
Suggesting that your political enemies would be better off dead is calling for murder. We just had assassination attempts against Trump in the most recent election. "Wouldn't it be great if the people I don't like were dead" is a call for murder, and some people have tried to act on it.
There was an adversarial collaboration (two people who started with opposing views on a topic, did research and tried to converge on a common view) abou infant circumcision on Slate Star Codex: https://slatestarcodex.com/2019/12/10/acc-is-infant-circumcision-ethical/ From their conclusion: "The benefits of infant circumcision appear to outweigh the risks and harms. Additionally, it is safer to be circumcised as an infant than as an adult, and a significant portion of the benefits of circumcision accrue to infants and children. From a strictly utilitarian perspective, infant circumcision should therefore be encouraged – whether we consider society as a whole or only the boy in question. However, autonomy is an important value, and while a man can become circumcised (missing only some of the benefits of having been circumcised as an infant), it is impossible to effectively restore the foreskin and become “de-circumcised”. An ethical system that heavily values personal choice over cost-benefit analysis may reasonably reject circumcision."
Thanks for sharing this. At this time, I'm leaning against circumcision for autonomy and precautionary reasons, but I also find myself sharing Scott's reaction:
"I had kind of swallowed the popular consensus that circumcision was Useless And Evil, and this gave pretty strong evidence that it has a variety of medical benefits and there’s no clear evidence it has any harms (especially reducing sexual pleasure, which is the main harm I’d heard cited). It seems to say that it’s hard to be against circumcision on anything other than a sort of deontological/autonomyesque/precautionary principle, which is not at all what I was expecting."
Most of the research I've encountered seems to either not come to a recommendation, or to be weakly in favor of doing or not doing it. I was taken aback by how strongly people here seem to be suggesting that circumcision is a completely indefensible practice. I'm probably not going to choose it for my son, but it seems like there is a reasonable utilitarian/non-religion based argument in favor.
My father is circumcised and made it a point to not cut me, and I am very very grateful to him for making that against the grain choice, so long ago. It is a choice I could make for myself now if I wanted to change it, but one I could not have changed if he didn't make that stand.
If Musk set out to persuade Trump that by 2028 Neuralink will be able to download Trump's mind into Trump Junior's body to contest the election, I think it more likely than not that he would succeed.
Many Thanks! Yes, that is the impression I'm getting as well. The next few years look like they will be astonishing. Well, 300,000 years was a good time, though (for mammalian species) not a long time...
I run a trading bootcamp (https://www.trading.camp). The next one is happening this (extended) weekend in SF! Take your loved one on a surprise romantic getaway for Valentine's day with a ticket to quantitative trading bootcamp.
Also, we're the first event taking place at Mox (https://moxsf.com/), a new coworking space and project of Manifund that Austin Chen is starting up, which should be fun. If you want to visit the space and meet bootcamp folks but not spend a bunch of days and money on the full thing, you can join us for Provide Your Own Liquidity, our auction/afterparty, Monday night Feb 17: https://partiful.com/e/qJLPSXLOC8k35JqxYeGz?
Confessing embarrassing ignorance about musical theory here, but can someone explain how counterpoint in songs like Simon and Garfunkel’s Scarborough Fair or the screen Bridge over the River Kwai version of Colonel Bogies March works?
I mean I appreciate that it is pleasing to my ear and does work, but why?
Bear in mind that all I know about ‘a perfect fifth’ is that Claude Lacombe used it to describe the next tone — ‘Up a perfect fifth’ is how I remember it — in composing the alien summoning melody in Close Encounters.
I'd say consonant intervals like thirds and sixths sound good, and they sound even better when you move from dissonant (second and sevenths) to consonant intervals.
The question of why some intervals sound better than others probably relates to the frequency ratios mentioned in other replies (and also the way human brains are put together, which obviously nobody fully understands).
The way it works out in practice is very complicated. But fundamentally:
Sounds are frequencies. If you consider, say, 440Hz and 220Hz, the 220Hz sound "fits" into the 440Hz sound. Every second wave in the 440Hz sound is also in the 220Hz sound. There are no waves in the 220Hz sound that don't also occur in the 440Hz sound.
Those 2 tones are in a 2:1 ratio called the octave. Sometimes, two instruments playing an octave apart will blend together so smoothly, it sounds like one instrument. The octave is the most "consonant" interval. ("interval" is just tone ratio.) The perfect fifth interval is a 2:3 frequency ratio. It's not as consonant as the octave, but still very consonant. Every third wave is the same.
The intervals continue that way, 3:4 is the fourth, 4:5 is the third, etc. Putting the various intervals together into a musical scale is very hard and complicated, and different cultures solve it differently. Western music currently uses a solution popularized by Bach called "equal temperament".
The specifics of how different intervals get put together into harmonies like Simon & Garfunkel use is very, very complicated. But, fundamentally, the pleasure you get from hearing the kind of consonant intervals they sing is from a straightforward physics thing. The waves match up and fit together, and that's nice.
I looked up the topic in Albert Schweitzer's Bach biography. Schweitzer states that Bach indeed used an equally-spaced tuning. But it would also make sense that the "actual" tuning wasn't perfectly even, it's not like he had a frequency analyzer. He had to tune by ear, and so the end result was whatever he thought sounded good in all 24 keys. Which would have to be pretty darn close to the perfectly even temperament.
Oh, I thought that he popularized the equal temperament system with "The Well-Tempered Clavier". Equal temperament allows you to play in all keys, so The Well-Tempered Clavier showcased that feature by having one piece in every possible key, which wasn't possible before equal temperament.
Not Quite. Bach (JS, the famous one) was a proponent of a number of "Well-Tempered" systems that gave each key a unique character but still allowed music in all key centers.
Equal Tempered temperaments only came into their own with the expanded chromaticism of the 19th C.
There are countless articles online debating the specifics, but as near as I can tell no one today seriously thinks Bach ws promoting "equal temperament"
Thanks for the reply. I think I have the pure physics part down. I’ve written code to automate hearing aid testing. Give me the FFT decomposition of sounds and the math is straightforward.
The pleasure of certain combinations is where the mystery is for me. But hey some mysteries are good.
Unfortunately I didn’t play any instrument when I was young or study music theory so I think part of this will always remain beyond my ken.
Nothing to stop me from enjoying the effects though.
Edit: It still bothers me a little bit that the frequency of middle C has changed over time. The math and physics guy in me wants to know “How can that be?” Yes, a small joke.
Further edit approaching non sequitur:
I run into a professional piano tuner occasionally in my local coffee shop. His technique is to begin by placing a vibrating pitchfork* against his teeth. Seems weird to me but he is the pro after all.
*Did I really type pitchfork when I meant tuning fork? Oy! I really need to read what I’ve typed before I press Save.
I can explain another level deeper. This is where you get into songwriting and composition. I'm a (amateur) musician, not a songwriter or composer, so I'm not fully fluent in this stuff.
Once you've got the primary auditory experience of pleasure (consonance) and displeasure (dissonance) available, and expressible in various degrees (there are more or less dissonant intervals / chords), you can begin to build an art on it.
Western songwriting primarily functions by setting up an expectation and then fulfilling it. Or, to put it another way, by creating tension and then resolving it. Songs have a key (for instance, C). C is the primary tone; call it the 1. You can build a scale on C; it goes C-D-E-F-G-A-B-C (or 1-2-3-4-5-6-7-8, where 8 is the octave). You can build a major chord on C, which goes C-E-G (1-3-5).
The most consonant other chord available is the chord built on the 4 of your scale, which, in C, is F major (F-A-C or 4-6-8). Another highly consonant chord you can build in C is the 5, or G major (G-B-D or 5-7-2). Neither of these are as consonant as the chord built on 1, because the 6 and the 2 clash a bit with the 1 (which is to say, audiologically, their main overtones form dissonant large-number-ratios with the 1's main overtones).
Because of music theory I'm eliding here, the 2, 3 and 6 chords have to be minor chords, which just means they incorporate a bit of extra dissonance in their thirds. You can and should use them; they'll be felt as very tense and unsettled, but still harmonious. You don't use the chord built on the 7, it would be so dissonant as to be unharmonious.
So a simple chord progression in C might be 1-4-1. You start with the most consonant possible chord, C major; then move to a less-consonant chord, F major; this creates a tension or expectation; which you then fulfill / resolve by moving back to C major.
One of the most common chord progressions in traditional American music is 1-4-1-5. Amazing Grace (basically) follows this progression, but also many other traditional songs. It creates tension, but never goes too far from home base.
The most common chord progression in contemporary American music, by far, is 1-5-6-4. (This is the one where you see those videos of bands playing 30 different songs, all with the same chords!) I think this one shines because it gives you a good variety of different chords, and kind of mimics the hero's journey, in a way that's hard to explain. Maybe a more competent composer than me can explain why that is. "When I Come Around" by Green Day displays this chord progression really obviously.
But you also get chord progressions like 4-5-6-4-5-6, particularly on the bridges of songs, which leave you unsettled and longing strongly for a resolution (which usually hits with the chorus, which will probably start with a 1 chord).
Or chord progressions like 6-4-1-5, which are hard to stop playing. Because the 1 happens in the middle of the sequence, you want to get to it by starting the sequence over again whenever you finish it.
The melody always uses notes from the key's scale. Melodies tend to have certain emphasized notes that function as resting points; these notes should belong to the current chord. Otherwise, it's dissonant. Generally, you can use dissonant notes in a melody as long as you're using them to pass to a consonant note.
Literally all of these rules have tons of exceptions, like in all arts.
So that's the bare bones basics of why different combinations of tones and intervals give pleasure in different ways. Like lots and lots of other arts, it's fundamentally about creating and resolving tension.
Thanks. This is very helpful and well aimed at my level of knowledge.
I think if I were to keep unraveling this down to the level of “Why does well done three part harmony give me the chills?”, I’d have to get to the mind body problem or the true existence of human souls.
I still prefer to tune my guitars by listening to the beats between the overtones generated on the 7th and the 5th frets of the higher and lower strings respectively, once I tuned the A string to the 440 Hz standard (or whatever tuning level I need for the piece).
I used to do this too, it's pretty common. I stopped when I learned more. The perfect 5th you get with that technique is 2 cents larger than the equal-tempered 5th, and it compounds so that your G string will be too sharp by 4 cents.
I've learned from experience that I need to err just a pinch flat. but I always assumed it was just a skill issue. But I do know in the back of my mind that there's a difference between equal and just tuning, so I should I have expected this in hindsight.
First, to clarify: we are talking about notes produced by musical instruments, e.g., a vibrating string, not pure tones generated by an oscillator. Here's the harmonic sequences:
440 Hz (A4): 440, 880, 1320, 1760, 2200, etc.
220 Hz (A3): 220, 440, 880, 1320, 1760, 2200, etc.
EDIT: A3 has multiples of 220 Hz, so the sequence is 220, 440, 660, 880, 1100, 1320, etc....
I'm not clarified. It still seems to me that a 440 hz would have waves that don't occur in a 220 hz, while a 220 hz would have no waves in it that don't also occur in 440 hz.
"Counterpoint" just means "simultaneous, complementary melodies." There isn't a unifying way they "work" - a bassline outlining the harmonic form in the absence of a chordal instrument could be analyzed as counterpoint, it's just a common enough musical convention with a distinctive enough function that it gets its own name.
There are LOTS!! of details I don't know, but the basic answer is "it's been experimentally observed". That's not a good explanation of why it works, but it's a basis from which compositions can be created. A second and a seventh are dissonant, but differently. (Note that the frequency separation is the same, it's the context that they appear in that makes the difference.) So there are LOTS of complicated observed patterns, that people have defined reactions towards. (And also note that a lot of this is learned. E.g. Javanese music doesn't use the same notes as the octave scale.)
P.S.: Don't rely on the details of this, my wife was the musician, and maybe there is a difference between the 2nd and the 7th. But in general the answer is correct.
Yeah, I agree with the general answer, but I'd say a 2nd vs a 7th have different frequency separations (if you raise the lower note of a 2nd by an octave (aka double the frequency) you get a 7th), while the dissonance quality is similar.
I've been following the news about Santorini earthquakes with a degree of morbid fascination. The archipelago is a remnant of a supervolcano explosion, i.e., a caldera. When this thing blows up again it will be catastrophic. This situation is the opposite of the 2.1% sweet meteor of dea... scratch that, giant explosion of Christmas 2032: we know with near-100% certainty Santorini will blow up, just don't know when.
Only tangentially related, but I consciously experienced an earthquake for the first time a few weeks ago, it was deeply weird! This is notable because I live in a the north east, not a land known for quaking earth, old hat to you Californian folks I'm sure.
NE earthquakes are more common that people seem to assume. The mountains there are older but there's still some movement happens. I believe there were several since the turn of the century. A cursory search produced this: https://nesec.org/massachusetts-earthquakes/
You're on your way to prevent multiple statistical deaths by donating 10% of your income in the form of a cashier's check, when you see a drowning child. If you attempt to save the drowning child, you will not be able to prevent multiple statistical deaths. What should you do?
Save the child, the go to my bank and have the cashier's check re-issued. The bank makes the funds for the check unavailable when it's issued, but they don't actually leave my account until they're deposited. Worst case, I have an extra 10% of my income unavailable in my account for 90 days while the bank waits for the first check to expire; I have more than enough liquid assets to cover this. Or if the soggy remains of the original check are recognizable, I can present it to my bank as proof that the original was damaged and the check should be re-issued without an additional hold.
Can't resist - this is too close to my old snarky version.
You are on a trolley taking a shipment of mosquito nets to the post office to send to Africa when you see Peter Singer drowning in a lake. You can pull a lever to stop the trolley and save him, but then you will not get to the post office before it closes for the long weekend, resulting in deaths > 1. What to do?
I am not a utilitarian, but I would let him drown - his conclusions are repugnant.
If I take into consideration that this is a path well-trodden by many, and that any moment now another Rationalist and/or Effective Altruist will be passing by, I leave the child and hasten to the bank with my cheque. Somebody else will save the child, whereas only I can prevent all those statistics!
Alternately, I hope that an ordinary person is passing by Drowning Child Lake (honestly, the local council really should do something, maybe safety rails or at least a lifebuoy) and since all we on here know Normies Are Stupid, they'll plunge right in to save the kid without even doing a statistical analysis first!
You mean would I rather be celebrated as a town hero, for free, or spend half my disposable income on something I can’t even mention socially without seeming like a douche?
This is my chance to register my view that utilitarianism is Bad and Wrong. I love Scott Alexander's writing in spite of the utilitarianism, not because of.
Utilitarianism is definitely a minority position among regular people and probably is among philosophers. The question is why it seems so over represented, especially on Substack.
Substack, and this substack in particular, is for nerds. Utilitarianism transforms ethics and morality, traditionally difficult and fuzzy humanistic problems that even the wise struggle with, into mathematical equations that can be solved with a bit of algebra or maybe a spreadsheet. The appeal should be obvious.
I get the appeal of systematizing ethics but it’s not the only way to do it, as Kantians could attest. Although if someone doesn’t know anything about the subject than it might appear that way. I once had someone here tell me they liked utilitarianism because it had numbers and they didn’t really care what those numbers were as long as they gave an answer.
Satisficing consequentialism* says *either* donating the income or saving the child is good and praiseworthy, and staying home and doing nothing is probably fine too.
We have a greater obligation to act when we are the unique and only person who can solve a problem. If there's a waterpolo team training in the same body of water as the drowning child, and shouting at them resolves the problem, then all is well.
If we're somehow prevented from donating the money later by saving the child (which seems unlikely) then this is still morally acceptable, as it's likely that other people will donate to save those people in some statistical sense. We might even compensate in the future by encouraging others to donate or increasing our future donations.
I think many of these questions go away if we accept some sort of nearness bias: it's OK to value the lives of those near to us in distance, time or blood relation above others further away.
Honestly, it seems like the best use of our hypothetical income would be to set up a lifeguard station beside this lake because those dang kids keep jumping (or falling) in and apparently none of them can swim a stroke.
Either that, or pay to have the lake drained.
I blame the parents. Where are they and why are they letting their kids run around lakes without being able to swim?
We really MUST value those we can directly perceive higher. We have greater certainty about their condition. If you donate money, you don't know the secretary isn't going to run away with the treasurer and use that money for their fling. You may assign that a low probability, but it should be well above zero.
(FWIW, I have donated money that was embezzled. This isn't an impossibly rare happening. And I suspect that it happens a lot more often than we hear about.)
Oh, there are many such cases! We had one a few years back which ended up in a right old ding-dong in the courts between the disgraced former boss of a charity, who appeared to be treating said charity (along with other board members) as a nice little earner for her. Then she went to court to sue over her reputation. I can't remember now how it ended up, but it badly damaged the charity, and charitable giving to other charities, for a while.
Yeah, that one is pertinent in regards to the uproar around USAID. Good causes being used as cover for fraud and scamming. Who could possibly be against funding charities that help people impacted by suicide? Even the possibility of criticising them would mark you out as a bad person.
Btu somebody has to be the bad guy who checks to see if all the alleged good causes are really what is claimed.
No one likes auditors (or Revenue) but they're fulfilling a social purpose. What we've learned again and again is that no-one can be trusted to be above the law. Trust but verify! (And don't burn the whole thing down for spurious reasons.)
When I see a drowning child, I can be more or less 100% certain that my choices can be the difference between life and death.
When I read that my money will buy malaria nets and save 10 lives on another continent, there are many layers of assumptions between me and the lives that I'm saving:
- I'm trusting that the science of malaria prevention through mosquito control is accurate, and that bed nets are a reliable method of mosquito control. (This seems pretty safe to assume but they get less reliable from here)
- I'm trusting that honest studies by good researchers have been done and have correctly identified the cost/benefit ratio here.
- I'm trusting that other researchers have done good studies comparing the cost/benefit ratio of this intervention to other potential interventions.
- I'm trusting that the actual charity that I am giving money to is in position to effectively execute the intervention in the most cost efficient way possible.
- And because I have no direct connection with any of those things, I am trusting my ability to read articles & studies and listen to smart people online and figure out which of their opinions I most believe given their inevitable disagreements.
All of which adds up to way less certainty than "that kid right there is gonna drown if I do nothing, but may live if I intervene."
Which is to say: I would save the child. I also believe in effective giving, and I _do_ try to give towards things like malaria nets, but the two things should not be treated as close to equivalently as they are.
Great question! If we're playing with hypotheticals then the 'child drowning' scenario is also rife with uncertainty.
- Is that child actually drowning? Maybe they're just playing? As a parent myself I've tried to study up on what drowning actually looks like. I have moderately high confidence that I could identify it. A harder scenario would be if they're not drowning at _present_ but are at risk, like if they fall off a boat and are floating downstream and the water is getting rough. There is certainly a level of uncertainty here.
- Am _I_ the only one who can save them? If someone else has already dived in and is heading towards them, then my intervention may not be necessary. Or maybe that rescuer is not actually as good as swimmer as I am? Maybe I should be ready to save both of them.
- On the other hand, if I jump in, and the water is rougher than it appears, then it's possible that not only will I not save the child, but that I myself might drown, or require saving myself, or worst of all I might drown myself _and_ induce someone else to drown trying to save me.
Nonetheless, as someone with a child who regularly goes to pools with other kids, I think it is well possible (if not likely) that I will face a scenario where I have a reasonable chance of helping to save a life with my direct actions.
1) Contra various inexplicable claims here, plasmids are not safe! They're only safe if not delivered by a real delivery vector because they don't get taken up by cells and basically nothing happens. If they're delivered by a real delivery vector such as LNP they end up in the cytoplasm and activate cGAS-STING which is incredibly toxic and inflammatory.
1a) Was the Ichor CRO experiment literally just naked plasmid? Not even minicircle claims that works I think, they use PEI which is its own brand of problematic. It's concerning that the powerpoint does not mention a vector.
2) That CRO report was vaguely interesting but they should have done qPCR or RNA sequencing to confirm the presence of mRNA generated from plasmid.
3) There are two separate "1,000-fold" gauntlets to run. First the issue of the ng vs pg discrepancy that was pointed out by another commenter. Second is the issue that local injection in mice and humans are not all that different in terms of volume, but humans are 1,000-fold bigger than mice so local injections go ~1,000-fold less far! This is why you need systemic delivery to e.g. the liver if you want to scale from mice to humans for systemic administration.
Does Hamas have an endgame? Are they fine with the status quo of 'reconstruction is extremely slow, we stay in power and are periodically raided,' or do they think there's a significant chance of something either worse or better for them happening. I haven't seen decent public information about how organized their leadership is, and what efforts if any they're making to shape the future.
Hamas is not run by rationalists, it is run by fundamentalists. A significant fraction of them genuinely, sincerely believe that if they stay the course then eventually an omnipotent being will rejigger reality so that Israel no longer exists and a Jew-free Palestine will become an Earthly paradise even greater than a Trumpian riviera, where the Palestinian people will live happily ever after. Under Hamas's rule, of course.
Another significant fraction believes that this is an extreme long shot, but that it is still better for a Palestinian, or even all Palestinians, to die at Israel's hands and thus be granted a martyr's ticket straight to heaven, than for them to be shuffled off to some refugee camp where existential despair will lead to unvirtuous behavior without the opportunity for martyrdom and so ultimately to the Not-Good Place. There's no rush, of course, so might as well try for the Jew-free Palestine version, and you have to at least be trying or the deaths don't count as martyrdom.
And then there's the ones who don't believe in any of that, but they have a *relatively* cushy gig as the Gaza elite, better than anything they'd find in a refugee camp, so long as they *pretend* to believe all that and act accordingly.
So, yeah, they're probably mostly OK with the status quo of Israeli occupation and/or blockade, and occasional Al-Aqsa Floods.
Their endgame is that ethnic cleansing or genocide is taboo in the 21st century in any population/place with 50+ cameras pointed at it, let alone an 80-years-old conflict infused with thousands of stories and inflammations/conspiracy theories and credible accusation of genocide. But Israel backed itself into a corner where Ethnic Cleansing or Genocide is about the only decisive and permanent solutions.
The decline of American power, and the increasing misery of the American taxpayer funding Israel's war (one among many wars (s)he keeps paying instead of healthcare), seem to also have something to do with the Palestinian militants' hope that - just maybe - one more Iron Dome missile would be too much. Right now, all the polls I have seen among young (< 25 years) Americans seem to be validating them, their bet is that (1) This will continue as those youth age into politics (2) They will survive the 30-50 years necessary for this to take effect, or the century of China-accelerated American decline will ramp up faster than America's Israel-hostile youth are aging (and that the 2 will eventually meet in the middle and live happily ever after).
------------------------------------
Israel is kinda hopeless demographically, very long-term speaking. Think about it, their fastest growing socio-political faction is the one explicitly preaching that military service is worse than death. And even with their fastest growing faction included they're still on equal demographic basis with the Arabs living in the whole land between Gaza and the Galilee. Once you expand scope a bit and notice the demographic trends among, e.g., Egypt and Jordan, you would soon see that the Arab autocracies keeping their people exactly 1 meal ahead of starvation are part of the Iron Dome, and Arab autocracies are one hell of an unreliable safety net to depend on.
Unless, that is, Singularity happens and now 1 person with enough money and data centers can genocide ~150 million people. Well, in that case, that person is about equally likely to be a Gulf Sheikh (who won't spare an Arab or a Jew) as an ultra-orthodox warlord.
The extremely smart move on part of Israel right now is to utilize their 50 years of runway that their current military supremacy is worth to achieve whatever political goals they can conceivably get, which is a lot. Asking for a Palestinian state is **remarkably** tame, it's the position of the USA since as far as I can go back and still recognize the names of the combatants. (which is why it won't solve anything if it's not properly militarized, a state is a military, anything else is a boy scouts club in various guises.)
Other than that, it can always risk genocide, effectively bluffing the combined might of Saudi Arabia, Egypt, Iran, and maybe even Turkey into doing something. To say nothing of what the EU and USA would do after a literal Hitler 80-anniversary reboot.
That's why Israel's Right Wing loves yapping about """Voluntary Emigration""" so much, but they're kinda too racist to even do that properly, after all, literally no Gazan except Ismail Haniyah's descendants would stay in Gaza for 1 second more if given a million dollar and a one-way-ticket + green card to Canada/USA, and one million * one million Gazan is a trillion dollar, about 30 years of Israel's military spending of ***Opens Google Tab*** 30.5 billion dollar. 1 million per Gazan is an obscene exaggeration, and Saudi Arabia + Europe can foot some of the bills (Saudi Arabia casually promised 600 billion, 60% of the bill to """Voluntary Emigrate""" 1 million Gazan @ 1 million dollar per Gazan, in a phone call with Trump). So given how easy it is to "Voluntary Emigration" a population out of a land, the Israeli Right Wingers are still too low-IQ and low human empathy to do even that properly and diplomatically.
It's extremely hard to predict, and Hamas is in a very weak position, but Israel is also in one hell of a corner. Like someone alluded, their biggest hope is that Palestinians will just... get bored? I don't know, seems kinda dumb, like someone who has no idea what "Revenge" means trying to predict the behavior of a human group after killing anywhere between 3% to 10% of the group and scarring who knows how many for the rest of their lives.
"Maybe after this they would learn their lesson", yeah, I bet that worked great for the Ancient Jews after the Romans demolished their state and scattered them all over the Middle East and Southern/Eastern Europe. Well technically it did, it just didn't work out for their successors.
Tldr; I don't know what Hamas endgame is and anyone who thinks is too dumb or too dishonest to know better, but it involves the global isolation of Israel and the "Demographics Is Destiny" hypothesis, along with the conviction that Islamist fundamentalism is enough to keep the flame of revenge glowing.
"Israel is kinda hopeless demographically, very long-term speaking. Think about it, their fastest growing socio-political faction is the one explicitly preaching that military service is worse than death."
The actual number of non-Haredi Jews is only going to go up.
We corresponded X thousand words on a previous open thread without resolving the question of whether any state composed primarily of the people currently in Gaza/West Bank could be persuaded (once it had the means to do so) to not attack Israel motivated by the belief that jewish self-determination in the middle east is an injustice not to be tolerated as a primary consideration over all other goals.
Just curious, have your beliefs on the topic/this question changed in the weeks since?
As a companion to my assertion that there is almost perfect symmetry between Palestinian Peace Obstructionism and Israeli Peace Obstructionism, here's a recent story in Haaretz about a Jerusalem book store whose owners were accused by the police of incitement to terrorism and dragged to court over books that contained such violent symbols of Palestinian Nationalism as poetry and cuisine https://archive.ph/ohDi5.
I'm not sure whether people outside israel who quote Haaretz are aware just how much it is the equivalent of anti American anecdote personally cherrypicked by Noam Chomsky, who is American so how can you accuse him of anti western bias
I haven't done any new readings since then, less than during the hot war in fact. More work in my life, less mental health to keep track with all the insane genocidal shit that keeps gaining grounds in English-translated Hebrew media (let alone the raw Hebrew media).
My beliefs on this topic or question isn't opposite, it's a "Yes And" not a "No" or even a "Yes But". Yes, obviously the cultural meme of "Jews are colonizers that can be driven out of the land if we just kill the next 50 of them" needs to die, hopefully quickly, hopefully with the minimum possible Palestinian and Israeli deaths.
But you know, I could almost say the same thing about massive portions of the Israeli mythmaking, collective consciousness, and culture too. There is an ironic and delicious symmetry in this conflict, keep reading enough of both sides and you will eventually be able to morph the grievances of each side into each other almost loss-lessly. I don't even see 0s and 1s anymore, I see "Arabs took this Jewish land from us" and "Jews took this Islamic land from us".
Where was I? Ah Yes, the weird Israeli supremacism that desperately asserts an ethnically-homogenous state is one of the fundamental human rights promised to people in the Universal Declaration of Human Rights, but also Palestinians can't have one. Because... because,... Palestinians are made up, they don't exist, they are all Jordanians and Egyptians in watermelon-colored trench coats. Or maybe they exist but they're just fake. Maybe they're real but have too few Nobel prizes, or maybe they deserve ones but we're just too uncomfortable to give them one now, come later in 15 years (Spoiler: Nothing will change in 15 years, it's a bluff in the desperate hope that Palestinians will get bored in 15 years despite not getting bored in 80).
Israelis and the people who love them (which sometimes include me, some of them, some of the time, when I'm not reading the nth genocidal translation of the latest pop star or politican to get viral) get too upset at Hamas supporters chanting "By Any Means" and "Those Who Make Peaceful Resistance Impossible, Make Armed Resistance Inevitable", and I get it, I really do. It feels wrong to be this gleeful at a literal real-life horror movie that I would dismiss as too cheap and blood-plotted if it was a Hollywood production.
But you know what, they're kinda right, as a purely descriptive (not normative, despite them meaning it in a normative, "cheer the revolution" kind of way) model, it really is the case that talking is the wonderful social technology, that - among other thing - discharge violence and give hope (however dim and false) to the dispossessed. Right now, carrying a Palestinian flag in public in Israel is a crime ("incitement"), that's why Peace Protestors and Radical Bloc leftists often carry watermelons or Palestinian-colors-Arabic-words.
It really is a truth that when you make it a crime for people to have nationalist aspirations, eventually they're going to figure out how to do it without you, and you're not going to be pleased with how they do that.
I had to stop reading your post right at that line, because everything I know about the history of this is that in 1948 the Palestinians and theIsraeli’swere given a state and that resolution was not a satisfactory one to the Palestinians or to any neighboring Arab state.. And it seems to me ever since then it has not been a satisfactory one.
Whatever the horrible mess we find ourselves in now it has been the stated intention of only one side that the other must be eliminated from the area and preferably from the Earth.
I'm discussing Israeli attitude right now and for the past 40-60 years. The Palestinian state stalled Egyptian Peace negotiations until Egypt relented (and also it wasn't very serious about it to begin with). The Palestinian half-state that Rabin had to clarify to the Knesset that it's not exactly a state got him killed, the only Israeli PM to meet that fate, as far as my memory serves.
I can find plenty of Israeli attitudes in 1948 and before to demonstrate that "Whatever we're in now is the fault of a single people" is - to a very good seventh-order approximation - politically-motivated Disney Morality to justify atrocities.
> ever since then it has not been a satisfactory one.
No credible offer of Palestinian statehood was ever put on the table since 1948. And the one on the table in 1948 wasn't from Israel.
No, it was not Israel because Israel did not exist officially. It was the UN and it was based on what the British did to try and make a two state solution. It proposed a state of Israel and a state of Palestine. Israel accepted those boundaries.I will disagree with you about offers on the table subsequently to create a two state solution. I think they had plenty of chances. The deal Arafat walked away from for instance. It might go a long way towards solving the problem if Hamas would remove the destruction of Israel as one of their core tenets. I have said before in these comments that it is very difficult to negotiate with someone whose goal is to kill you.
If you’re talking short term goals, I think their hope was the ceasefire would hold, and after a few months, after they re-established de facto power, the prospect of restarting the war would induce Israel begrudgingly accept the status quo.
Long term, probably they hoped to rally the neighboring countries to their side, or force Israel to make a deal with them fully accepting their permanent control over Gaza. Both long shots of course. It’s long been known that there were comparatively moderate and extreme wings in Hamas in tension with each other. Personnel has changed of course though so who knows now; but I’d say it generally behaves like an organization with considerable internal conflict.
In general the Palestinian strategy has been to resist Israel as long as possible using unconventional methods (typically against civilians) rather than face the overwhelming Israeli military. I believe this is mostly to gain internal political credibility and to keep the conflict alive in international news. It also provides a morale boost and exerts a small amount of negotiating pressure on Israel by showing that they are capable of doing some damage. While doing this, they appeal to international organizations and global public opinion to attempt a diplomatic victory, similar to the way international pressure ended South African apartheid. They seem to view this as a genuine model for potential victory rather that purely being rhetoric used for propaganda. (They have seen huge gains in public support over the past decade or so, so you could argue that the strategy has been at least partially effective.) It is hard to predict if that strategy will continue post-Oct. 7. They seem set on fighting to the death rather than being expelled or living under Israeli control.
They've been outspoken that their plan is to just commit more and more October 7s, forever, until there are no more Jews. Pretty retarded if you know how population increases work, but if they weren't stupid they wouldn't be in Hamas, so there's a pretty hard upper bound on their rationality.
I doubt their retirement would be long lived. Once they’re out of power and have no leverage there’s little reason for the Mossad not to keep assassinating them. And fanatics tend not to weight personal self interest overly much anyway.
That's exactly why almost nothing of his foreign policy works. He perfected the art of "Whining and blustering till someone blinks" and all the limited concessions that can be obtained by it, but has 0 strategy or international smarts beyond that.
There is a reason there isn't a wall with Mexico since the ~10 years he has been publicly promising it.
Building a wall across the entirety of the southern US border is a lot harder than moving/killing 2 million people concentrated in an area that's only slightly bigger than San Francisco.
He's proposing to _buy_ the property? I hadn't noticed that element in his statements....pretty sure that no Republican member of Congress would have anything but an "oh hell no" reaction to idea of Hamas members getting paid cash money for Gaza.
(Perhaps Trump could finance such a purchase the way he paid for all his business ventures: no money down, extravagant promises of future balloon payments, stall for years on making the actual payments, strip cash flow down to negative then file for Chapter 11 bankruptcy and get the courts to stiff all the creditors for you, rinse and repeat. Maybe Hamas/Iran would fall for that, one time at least?)
WRT: "pretty sure that no Republican member of Congress would have anything but an "oh hell no" reaction to idea of Hamas members getting paid cash money for Gaza" - I've lost count of all the times I though this about various Trump's... "ideas". Alas, the Republican party is now basically a North-Korean-style "write down everything Dear Leader says and enthusiastically support" Party.
Both of you seem to be under the impression that he means to purchase it *from Hamas*. I would've thought that he means buying it from Israel, and the payment will probably be in the form of weapons and such that we would've sent them anyway, only they won't have to pay for them.
I'm using "buying from Hamas" as an illustration of the utter idiocy of the whole idea. It's really hard to take it seriously, and I am astounded (but I guess I should not be) at the ineptitude of the journalists covering it, should not this very question, "who are you buying it from, Mr. President" be shouted, printed, posted, whatever, all over the place?
But - taking it seriously for a moment - there's no need nor a legal way to "buy" it from Israel, is there? Israel controls it militarily at the moment, it can just hand control over to the US, how is a "purchase" even relevant here? But it doesn't "own" it in any legal sense, which is what would be required for a "legal" purchase?
God I feel stupid and mentally spent just typing this.
Assuming he is a clever Machiavellian and not a bumbling buffoon, the point of the "bying Gaza and expelling all Palestinians" schtick is to influence how Hamas perceive their choices. He might attempt to scare them into a deal, by influencing the perceived disaster they may face, if they do not. It's a negotiating tactic.
Haha, underestimating the ineptitude of journalists, classic rookie mistake! Well, not a crippling one; you'll learn.
I agree though, yes, someone at some point really ought to have asked the question. Arguably Israel does own Gaza by dint of conquest – the only other possible alternatives, since there's no Palestinian state and never will be, are that it still belongs to Egypt, which they vehemently deny since that would make them responsible and the Gazans Egyptian citizens, or that it's a stateless no-man's-land. So potentially Israel can sell it to the US, and presumably Trump's idea is that buying it will give US ownership greater legitimacy somehow. Trump is a dunderhead in many ways, and this is typical of his way of thinking IMO.
Haven't states historically ceded territory as part of peace deals? Is your objection just that that he used the terms of everyday commerce instead of some fancier "international law" jargon?
Oct. 7th was the plan. Playact submission for a decade to convince Israel to take their eyes off you and write you off as a managed problem. Build up the people, war making material, intelligence, and connections with other Jihadi entities you need, launch a sudden and overwhelming strike from all sides, kill tens of thousands of soldiers and civilians, capture tens of thousands more, capture and destroy IDF equipment, collapse the Israeli state.
But the expected attacks from Hezbollah and the planned for uprising in Judea/Samaria never came. And their own military success fell short. And Israel responded in a much harsher and sustained way than they expected. And then Israel knocked out Hezbollah and launched a massive Jihadi suppression campaign in Judea/Samaria.
So Hamas now is falling back on their tried and true survival strategy of milking international sympathy for the dead and suffering Gazans they generated to limit Israel's freedom of action and give them the chance to regroup and fight another day. And that's what you're seeing now.
There's at least three active levels warfare between the Gazans and Israel going on all at the same time.
There's the direct warfare, which comes and goes but has been pretty significant for a while. That's obvious. Then there's the political wrangling between all the various groups that make up each side (Palestinians, Hamas, Iran, Israel - broken down by factions, etc.). Finally there's the propaganda war that's mostly fought for sympathy of outsiders.
Israel's long term plan is on the political front, and seems to be to slowly encroach the Palestinians until they disappear as a distinct people and the descendants of people who care do not care anymore. Israel is patient, so this strategy is very likely to work if unopposed.
Hamas recognizes this problem - they can't win this war. They also can't win the first kind of war, direct confrontation. So Hamas is pursuing the third kind of war. They can only do so if they provoke Israel to violence. Provoking Israel to politics isn't effective at getting international attention and has a strong chance of Hamas losing anyway. Their hope, not because it's a good strategy but because it's the only strategy with a chance of success, is to get the outside world to stop Israel and force a compromise that makes Hamas' position stronger.
I think this would work much better if Hamas was not a stooge of Iran and had a more moderate endpoint of political freedom rather than direct or implied goals of destruction to Israel and maybe killing all the Jews. Hamas only has a chance to win this war if they keep Israel active and in the international news. If things calm down too much, Israel will eventually win. There are paths that allow Palestinians to also win at the same time, but Hamas cannot win by its own stated goals if Israel also wins. If Hamas were separated from Iran's goals, and was willing to compromise on some currently primary goals, then maybe Hamas could also win without Israel losing. I don't see any path to Hamas winning with its current goals that does not involve international pressure. So long as the US supports Israel, international pressure is unlikely to be decisive - forcing a stalemate. This is still to Hamas' benefit so long as they are locked into their goals and the alternatives are worse for them.
Hamas became a client (very much not a stooge) of Iran because of what they believe and want, not the other way around.
Hamas exists in order to end Jewish sovereignty over any part of Arab/Muslim claimed land, create an Islamic theocracy in that land, and reassert Dhimmi status over any Jew still living in that land when they've accomplished their likely genocidal conquest. They have no other political aims or ambitions.
Pretty mysterious why Hamas existed since the late 1960s (as an offshoot of the Egyptian Muslim Brotherhood, a 1928-founded political group whose latest exile in Egypt since 2014 still didn't quite finish it) and never did any violence until the Cave of the Patriarchs Massacre in 1994.
Strange how Hamas was totally okay with Jewish sovereignty until a Jewish terrorist killed a bunch of Arabs (and cheered as a hero until now by Israeli ministers).
>They have no other political aims or ambitions.
Lol, someone has no internal model of the world beyond what Channel 14 feeds him/her every night on the brainwashing airwaves.
Hamas isn't Athena, it didn't appear in a Teleporter Gate in 1987. They declared independence from their Egyptian mothership in 1987, but they have been active in the strip under various name and guises since the 1960s, and Israel tolerated them because it loved the Islamic counterbalance to the secular leftist PLO (sounds familiar?).
> They literally wrote it down
Seems like we ought to go back to the writings of Israel's founders from the 1920s and the 1930s if that's the standard we're going to adopt, and I assure you you're not going to be pleased with what we're going to find.
In 2021, congress ended their moratorium on earmarks aka "pork". The moratorium on earmarks had been there for about 10 years and had been intended to reduce corruption. I recall that this had been discussed here quite a lot while the moratorium was on-going, with people claiming that the moratorium on earmarks caused the breakdown in bipartisan bills we've seen. The theory went that previously members of congress that were on the fence could be bought over by earmarking something for their constituents, like something that would bring jobs to a specific area.
It's been a while now and I don't think I've read anything about how this has gone. Congress feels broadly as partisan as it had been ten years ago but I'm not really that informed. Have earmarks been used much? Are there any notable cases where a bill passed because of earmarks?
I was surprised by the high amount of major bipartisan legislation during the Biden years. Back in 2020, I'd assumed bipartisanship was permanently dead. Perhaps the earmarks had something to do with it.
I've had some tangential involvement in this, from which my observations are:
-- today's earmarks (which have a new label that I forget right now) are drastically smaller in real dollars than back in the day. During the 1980s/90s a $1 million earmark was considered pocket change, now they're dealing in the high 6 figures. An award of a million or two is large now. All figures nominal, so in inflation-adjusted terms it's a big drop in individual earmarks.
-- perhaps related to the above but also given the shifted optics of the practice, there seems to be much less bragging about such awards by the congresscritters than was true in past eras. It's really just a quiet staff exercise now. It's pulling teeth now to get an elected member to attend a ribbon-cutting (I have colleagues who've discovered this firsthand).
I know you're being snarky but just for the record: actual federal grantmaking, which I have lots of professional experience with, is enormously more complex than earmarks both to obtain and to successfully comply with. For example every entity that receives at least $1.5M total in all federal grants from any agencies during a year undergoes a special federal grants audit, in addition to its own sector's standard annual outside audit. I've dealt with that process and it is a bear, or at least was until somebody started blowing up the OMB and other federal agencies. My organization had to tear up our entire payroll system and budgeting structure, which had served us fine and made regular CPA-type auditors happy for years, in order to comply with federal rules for grants compliance. Etc etc.
Federal grant pools are also competitive as hell, e.g. we were an unsuccessful finalist recently in one under which 150 grants were awarded out of 2,700 applications.
None of the above applies to congressional earmarks ("member initiatives", I think is what they are now called).
Some of it has been specifically exempted from the spending freeze, including testing, care and treatment. Preventive anti-HIV medicine is specifically excluded though, does anyone know the reasons?
“Alarmingly, nearly all HIV prevention efforts under Pepfar — aside from programmes for the prevention of mother-to-child transmission — are currently on hold,” write the authors of a Lancet editorial, which was published last week. “Programmes that focus on the prevention of HIV in key populations, HIV prevention services for adolescent girls and young women, voluntary male medical circumcision and support for orphans and vulnerable children are still halted.”
So it sounds like the kept HIV prevention efforts that use medications (the mother to child transmission ones) but cancelled the other prevention efforts, such as circumcision and presumably condoms.
Hmm, from the article this includes things like CAB-LA and Lenacavir. I don't really understand why these need to be put on the chopping block for the sake of not funding condoms. Do you know if this condom thing is the reason or are you speculating?
Speculation. But it does look like the CAB-LA and Lenacavir thing are not about freezing preventative measures but a different freeze on scientific studies, as the meds hadn't been used before and they were going to do a trail with them that would start in March. It could still get unfrozen before then, in the meantime nobody was getting those meds right now anyway.
Reading the article, it looks like pregnant women can get the anti-HIV drugs, but no one else. My guess is that it was to avoid having people come back at the Trump administration with the not fun to debate claim that they were giving babies in Africa AIDS (not that I honestly think most of his supporters would care if that were true). So now they can claim that actually they are doing everything reasonable to fight HIV but anyone "irresponsible" enough to contract it anyway doesn't get treated on the US government's dime. Its all very in line with their policy of "pretend everything defaults to personal responsibility and use that as an excuse to be as cruel as you can get away with without causing open revolt."
Also, I'm glad the meds are at least back for the pregnant mothers, but I'm not sure there's anyone actually working at US Aid anymore who can/cares about implementing the rules correctly, so my guess is anyone tracking this stuff carefully is still going to see a huge uptick of babies being born with preventable HIV.
I'm a senior physics and math undergrad, gotten into a couple good grad schools for physics, probably specializing in biophysics.
Mostly because of this community, I've become mostly convinced that I could do the most good if I pivot to AI safety research. Considering how fast the industry is moving at the moment, how should I go about this? If timelines are long, I could obviously just complete my PhD and then pivot once I was more hireable.
However, if I want to maximize my impact before humans become less relevant in guiding the ship, should I consider trying to pivot now? If so, how should I go about that? Would any ai company hire some non-cs major right out of undergrad? My cv is quite strong for graduate school applications but I have no idea what the best way to go about this pivot is.
It might not be the best thing to be convinced by this community.
You can do a heck of a lot of concrete good in biophysics. It is not super well established (around wider computer science or academia) that the alignment people are doing something useful. The claim that they're doing good work is... speculative and more than a little bit controversial.
A PhD is going to take 5+ years: think about where AI was in 2019. (Where? Exactly.)
This is a bigger and better-founded hype than the previous ones, but academia has always been pretty faddish, particularly in the US. At least one next-big-thing will bubble and burst before you finish your PhD.
There's a lot of interest (and I presume money) in AlphaFold type projects that are really at the interface of maths, physics and biology. I'm guessing you could get solid expertise in AI doing that, be relatively employable afterward (assuming you don't buy into the wrong bubble) and continue on to do something interesting afterward.
Second this. I published in both physic and microbio back in the day, and it's a more exciting area than ever now, and will require actual human minds behind R&D programs even if we have true AGI / GPT-7 by the time you finish.
AI (even at AGI levels) is more likely to be a force multiplier for a good amount of time, in my own opinion, because there will be a lot of technological branches that become low hanging fruit and need to be cranked through and thought about.
My personal prediction is that actual AI (AGI, superintelligence etc., these have different meanings to different people but essentially scifi level artificial intelligence) will emerge from a different substrate than what we currently have (computer chips, transformers etc.) so your knowledge of physics can be put to good use to develop such technology, new substrates for computation.
I really thing that AGI is substrate independent. The problem is algorithms. It *is* true that specialized hardware can run certain classes of algorithm faster, and not others, but note the specificity there. A CPU can run any algorithm. It may be faster or slower, and possibly need more memory, but it can do the job. And specialized hardware is often only useful for specialized tasks.
I don't think one should commit oneself to any particular hardware choice. I'm relatively sure that CPUs and GPUs and Transformers will continue to be useful components, and I'm not really convinced that there will be the need for something like that that we haven't yet identified. (E.g. memristors may yield a device that would be extremely useful, and which is also on that level, but...I'd give it about 15% chance, possibly less.)
FWIW, the most likely candidate for a breakthrough device would be a room temperature quantum computation module. But I'd bet you'd only need relatively few of those, and most of the work would be done by CPUs and GPUs, with a minority done by TPUs. (Note I'm putting quantum units, QPUs, so low that I didn't even name it in the list.)
These are just my personal evaluations. And anyone specializing in any one of those general areas should do quite well. But don't tie yourself to some device that hasn't yet been used. Most of those, even the ones that eventually succeed, take decades to get wide use. Whereas there are still uses for isolated resistors and capacitors. (Pentodes, though, seem to have disappeared.)
I'm not that savvy on this topic, can I ask why you think it's substrate independent? If we agree that the brain is more advanced than computer chips, can't we imagine all sorts of different substrates that would fall in the space of complexity (or capability?) somewhere between chips and neurons?
Anything computable can be run on any Turing complete system. Unless you assume some sort of magic, intelligence is computable. (The problem with this argument is that there is no real Turing complete system, as such a system would have infinite memory.)
There are holes in this argument, but they're very small ones. And the argument doesn't say anything about relative speed.
Note that the set of problems for which a quantum computer is known to, in principle, but better than a standard computer is quite small, so even if you assume the brain is a quantum computer (at THAT temperature???) it's unlikely to make a difference. (N.B.: That quantum processes happen in the brain isn't that unlikely. They're likely involved in helping the mitochondria more efficiently generate energy, just like they are in allowing chloroplasts to more efficiently use light...but this is a long way from being a quantum computer.)
My new SF collection Flashes in Time presents the reader with a fourfold table of science fiction and fantasy, contemporary and far off. The stories, whether zany or serious, make an effort at being coherent worlds with mostly smart characters whether we're dealing with simulations or magic. The subject matter ranges from transhumanism, AI, and space exploration to subjectively karmic magic, navigating premonition, and dealing with the Fae.
Digital editions start from 3.99 currency units, and I do take (a finite but non-minimal amount of) requests for review copies via e-mail (see below), if you'll put one up on Goodreads, Amazon and/or your own relevant platform, if any. https://waterdragonpublishing.com/product/flashes-in-time/
My novelette from last year, The Paperclip War, may also be of interest as a take on what if humanity managed to wield unusually harsh game theory to arrive at an unstable equilibrium with an AI adversary. It was well received at Less Wrong Community Weekend 2024. As a modestly sized standalone novelette, this one starts at a 0.99¤ https://waterdragonpublishing.com/product/paperclip-war/
So... in order to help overcome the impending demographic collapse I was spitballing some less common ideas. In the United States, whatif the vote was only given to women who bear children? Lots of little details to work out, but that would make bearing children pretty attractive and would ensure that the government made their concerns the highest priority.
In the details one could 1) Extend the vote to individuals (of any sex) that legitimately attempt to have kids. 2) Give the women two votes, one of which she could grant (permanently?) to a partner who is actively supporting the child. 3) Give bonus votes for additional children, possibly alienable.
Voting probably doesn't move the needle much of any way, but the right way to do it would be to let parents vote on behalf of their minor children (children deserve representation too!).
Echoing Bryan Caplan, voting is pretty untethered (I think at least) from rational self interest, in large part because voting has a negligible effect on one’s self interest, so using the vote as an incentive wouldn’t be very productive. Voting, like politics in general, is much more in the realm of the emotional than of self interested rationality. If I cared about the right to vote even as much as most Americans, if I were denied it because I’m not a woman and don’t have children, I would sooner react angrily and possibly violently against the system than start knocking up women to get my vote back. By contrast if you offered me a lot of money to have more kids, then I’d consider it, even if I hated the policy politically, I’d still be tempted by self interest to do what it wanted me to do.
Basically using voting rights as an incentive usually has a high backlash/effectiveness ratio in any society where there’s a string sense that everyone has the right to vote.
It all may be for naught, given the studies that show male fertility worldwide has been declining over the past two decades — both in sperm count and motility. There's been a lot of post hoc handwaving about why this has been observed. And, of course, any studies that get a lot media attention — especially those that get attention from the science editors of the NY Times — should be considered suspect until proven otherwise.
But if true, I can think of all sorts of potential research avenues for funding. For instance, do incels have lower sperm counts than average? And does that affect their sex drive? Of course, the NIH may soon be out of the picture as a funding source... We may have to rely on our Tech Bro Billionaires to fund these. OTOH, neither the Muskrat or the Zuck would probably be interested since they seem to be breeding OK.
Mothers directing the votes of their kids as to who to vote for was an old-fashioned thing (at least in my day) so this is not some wonderful new idea 😀
Secondly, why the hell is it always men deciding that the problem here is women so the way to get women to have more kids is to treat women like incubators, such that even a fuddy-duddy traditional Catholic social conservative like myself wants to spit on my hands and raise the Jolly Roger? You're not the worst suggestions I've seen Guy, but seriously. I'm fed-up to the back teeth of "hmm. no babies, must be all the fault of women because all the men are just aching to become fathers of twelve since they were kids".
It takes two to tango. Unless you have men willing to get married and become fathers, the woman can want sixteen kids all she wants but if boyfriend doesn't want to put a ring on it, or husband doesn't want squalling brats interfering with his free time, it ain't gonna happen.
Thirdly, we've had previous discussions on here of the smart folks saying that since their single vote makes no difference out of the n millions of possible votes, only fools go out and vote as if it meant anything. Maybe women would be happy to shrug and go "fine, me no baby, me no vote? no skin off my nose".
I'm too busy/lazy to find the references, but some recent studies indicate/show pretty convincingly that it is the male partner who stops after child no 2. While a substantial proportion of women would go for child no 3 if their partner were in on it.
This matters, since it is lack of "the third child" that is key to understand persistent below-reproduction fertility everywhere (except in israel).
So how to get men on board with child no 3? It is easy, although no-one is doing it at present. Give all women who have child no 3 five million USD (or the equivalent in your currency).
If that should still be too little, jack it up to 6, 7, 8 million, and so on until you get the effect on fertility that you want.
Or decrease to 4, 3, 2 million if you overshoot your target.
It costs money, but a country can always afford what it gives the highest priority.
except you can trace the demographic decline to increasing women's rights in society and access to contraception and abortion.
pretty much its women actively choosing career over children till they have one kid in their mid thirties if that.
Guys generally haven't changed except weirdly they are more into women not less: the idea they are a ball and chain has vanished to be replaced by the frantic incel who thinks his life is worthless without a girl.
The first wave of demographic decline in Western countries (ie. pre-WW2, when TFR dipped under replacement in a lot of countries) happened without the widespread adoption of hormonal contraception and other methods dependent on women and with abortion illegal basically everywhere in the West, including in countries like France without women's suffrage.
It wasn't women alone who welcomed the Sexual Revolution. I don't know if you're old enough to have read popular literature of the time (60s-70s) where the idea of "women want to get married and have babies" was treated as entrapping men, holding them back from being free to wander the world and live freely (and also fuck any and every woman that would let them, without consequences or commitment).
A pop song and hit of the time (which I *hate* because I want to punch the guy in the face) is "Lydia"; the basic story of which is "hey so Lydia is my friend with benefits, she lets me come crash at her place and sleep with her, but no strings attached, she understands that I have to be free to be me so I'll up and leave in the morning - until the next time I hit rock-bottom and need someone to give me shelter and sex".
And that was the Ideal Woman for the time - one who wouldn't be a hung-up prude about sex but who would also not put any demands on the guy; she'd be there to be available for him as and when he wanted, but no need for a formal commitment.
A lot of songs of the time had those "I gotta leave in the morning" from the guy's point of view lyrics.
A lot of the damage of feminism of the time was "why shouldn't women behave like men?" including in their sexual lives. Men were perceived to be free to 'sow their wild oats', they did not suffer from having the double standard applied to them, and the refrains nowadays about "women want to bang Chad and don't want to settle down until they've hit the wall and are no longer able to attract alpha males" makes me smile because it is the flip-side of the earlier complaints about "men only want one thing and that's not marriage and kids".
So yeah: now the Pill is here! Now you don't run the risk of pregnancy outside of marriage! So be free and equal to your male partner(s) who can sleep around and sex is for fun!
And here we are today: oh no, the birth rate is declining, this is all down to selfish women who only want to find 'em, fuck 'em and flee, while the hapless men are just longing to be husbands and fathers at the age of twenty.
Yeah, I'm sure all the guys out there playing the field would be totally happy to be 22-year-old husbands and fathers if given the option. What you would really see without easily available contraception is a massive uptick in single mothers. Where do you think all those jokes about dads "going out for cigarettes" came from?
i think you heavily overrate how easy it is for guys to "play the field and overestimate how many want to. And honestly 22 year old fathers are a bad thing why? its better than nearing 50 and your kid hasn't even finished high school, had no cousins or siblings, and their grandparents are hitting 80.
as for single mothers, well not sure current status of "no mothers" is a plus here too. Think it will make it much harder for women but 50+.
Look, there has been lots of scandal and lots of drama about the Magdalene Laundries in Ireland, but the impetus for them was precisely because women and men were having sex outside of marriage and when the woman got pregnant, the guy either denied parentage or headed off to England, leaving her with a baby and no support.
If you know of a huge crop of young men in the US who are lining up to be fathers and husbands and not finding anyone in their community willing to partner with them then you are living in a very different part of the country than me.
But I think that last part is the real point. Most of the people clamoring about falling birth rates don't really care if the burden falls disproportionately on women. Admit that and we probably have nothing left to discuss.
Relatedly, we've seen time and time again that the reasons even happy committed couples are waiting later to have children is that generally now most people don't have a sense of financial security until they are in their 30s. The baby boomers went right from college to starting families, but they also went right from college to government backed mortgages that represented a fraction of their annual income as opposed to their lifetime income, and employment offers that came with lifetime guarantees. All that has gone away, and every time someone suggests replacing it with more robust government programs, most of the same people who complain constantly about low birthrates start screaming socialism.
"The baby boomers went right from college to starting families, but they also went right from college to government backed mortgages that represented a fraction of their annual income as opposed to their lifetime income"
Back in reality housing is easier to afford, thus why people live in larger houses than ever before.
its not that, you have the same demographic collapse in EU too, social programs are not helping anything. The USA has a higher TFR than finland, snd sweden, and all EU countries are at best .1 above us and well below replacement rate.
type of government doesn't seem to matter; its endemic to most non african, non muslim nations and i don't think germany is worse than indonesia in terms of that to explain a .6 differerence.
women do not want to have kids when given the choice. i wont blame them but kind of sick of the "oh its capitalism!"
> "Lots of little details to work out, but that would make bearing children pretty attractive and would ensure that the government made their concerns the highest priority."
So what I'm hearing is that you're either not a parent, or a very bad parent.
>but that would make bearing children pretty attractive
No it wouldn't. There are literally no incentives for women to care about pursuing the right to vote *individually*, because the chance that granting any specific individual the right to vote will ever affect anything is essentially zero. Franchise extensions only make sense as incentives if you're granting them to large demographics all at once.
Between 30 and 40 percent of American women don't bother to vote now, whereas around one in five complete their child-bearing years without having any children. And of course not all of those who end up not having children had gotten there as a deliberate choice.
So it seems unlikely that there would be a large number of women who (a) place a high value on voting and (b) explicitly prefer not to have children.
I'd be open to trying that, honestly. How much really do we have to lose from here?
Similarly I've advocating for a 2nd amendment making gun ownership a constitutional right for women while leaving the gun rights of men up to the federal and state legislatures. [No I don't know how that would be applied to transgender folk, when I first thought of this that whole topic was not nearly as salient as it is now.]
Neither of these scenarios will ever happen obviously or even be seriously proposed by any political party.
Your idea goes against democratic tradition. There would be immense resistance to any policy that messes with who gets the vote.
If you can magically convince people to accept an idea like that, why can't you just magically convince people to have children instead. It would be easier.
I think that taxing childlessness to subsidize child-rearing has a higher chance of being accepted by society.
Also, teach in school that being a parent is wonderful, award medals to people with lots of children, that kind of stuff.
Make sure this ad is all over the media, instant baby boom:
<Also, teach in school that being a parent is wonderful
Most research finds that on average having children makes people a bit less happy. Seems to me we should try to figure out how to make parenthood a better experience, rather than fucking lying to children about the facts.
Isn't this one of those "happy now vs satisfied later" things? I mean, I'm all for making parenting easier, but I suspect the happiness research people are at best getting an answer to a somewhat different question than you'd like to ask.
(As a parallel, many people are glad they got a PhD or got through medical school, but grad students and med students are in general not super happy people.)
I don’t think many people whose children are grown are sorry in retrospect that they became parents . In fact I don’t think most parents of small children would say that are sorry they became parents — though a few would. Still, I think parenting is different from grad school when it comes to regrets. Grad school is rarely talked up as a wonderful fulfilling experience (except in the grad program’s catalog’s), but parenting is. Also, there are no internal barriers to concluding, in the years after grad school, that it was both unpleasant at the time and also mistake. But concluding that about parenting seems to carry with it the idea that you wish your child did not exist, and anyone having a wish like that is going to feel like a terrible person. And actually, I think even people who regret having become parents still love their child.
My own experience of parenting was that I found it very very difficult. What was hard for me certainly was not my daughter herself — I loved her, and thought she was a great kid. And it wasn’t the work and the expense. The hard thing was having no time for myself — no time to pursue my interests, to hang out with my friends, who were childless or whose kids were grown, no time to ruminate and crap around and explore things. Until my daughter was about 4, I felt like I never had time even to complete a thought. It was a horrible, maddening deprivation of things I think I really do need for my wellbeing, and the deprivation often made me furious. But there was no one to be furious at except my sweet kid, who was blameless, and mostly I stayed aware of that. I simply had to feel the deprivation and the frustration and do what my daughter needed anyhow.
People get sold the idea that once you have the baby you transform into another kind of being, one whose greatest delight is nurturing little lives. I simply did not go through the transformation. I just parented without the benefit of the transform, and I don’t think my experience is rare. Scott, in one of his recent posts about his kids, wrote about how he could love taking care of his babies for a couple hours, and push himself to do it for 2 more hours, and after that it was just too much baby. Yep, that’s the feeling. Other promised transformations, by the way, have worked for me. For instance I went from being a grad student who had studied clinical psychology but did not feel up to actually trying to help people to a professional who enjoys her work. The transition took a few years, and some of it was painful, but it happened.
My daughter is now 29, and is one of the happier people I’ve known. (I managed to do a decent job of parenting. I guess.) I’m delighted to have her in my life. She lives 15 mins from me and we see a lot of each other. I’m not sorry now that I adopted and raised her. But there’s no denying that the years of doing it were just not good years for me.
"I think that taxing childlessness to subsidize child-rearing has a higher chance of being accepted by society."
Oooh, the screaming of the "childfree by choice" lot would reach levels unheard since the Chicxulub asteroid hit. They already feel they're being taken advantage of by society which props up breeders.
At this moment, my suspicion is that even if Vance and Musk pushed through such a thing _this_ year, the countdown to AGI seems to be ticking so fast, and seems likely to have consequences so vast, that I'm not likely to scream much...
Not just make it the highest priority - it would absolutely exclude every other group's concerns from consideration. Young people, childless people, and obviously men (but let's not talk about that last one) - they would have virtually no power to have their needs met unless they also align with the needs of mothers. Also, I think currently a working mom is highly aligned with an average voter in what her views and needs are, whereas stay-at-home moms are very divergent from the mean, meaning currently their views are less represented. They'd gain much more political power by this change than what working moms would gain. Is it good or bad? Don't know, but it would make things different. Maybe they would use that power to pass enough laws favoring SAHMs that it wouldn't make sense anymore for working moms to keep working? Again, don't know if it's good or bad. Personally I think it would be great for the society if an average family could lead a comfortable life on single income, but that's a tangential topic.
The thing about such thought experiments is that when you ask "how do we even get there", it's obvious it's wholly and entirely impossible and it stops being an interesting discussion. So I'd rather not ask.
Votes aren't worth a lot. Lots of people already choose not to use theirs, and people in general don't seem willing to do life-changing changes like moving to a different location just to make their vote more powerful, although such opportunities are already abundant. So I doubt that it would have much of an effect.
Why not make fertility treatments free? The up-front costs would be paid for several times over by the creation of additional people who would contribute to the economy for a lifetime.
Lots of countries do this already, and it doesn't really move the needle. All the Scandinavian countries and a number of other European countries, for example. Fertility rates still low and getting lower.
In fact, many things have been tried to practically no effect:
* $10k bonuses per child (Singapore), or for 2nd / 3rd children (Russia)
* 3 years paid parental leave (France)
* 480 days paid leave at 80% wage (Sweden)
* Income tax exemption for mothers with 4 or more kids (Hungary)
* Free state paid child care (France, Germany, Sweden, Norway, Finland, Estonia)
* Free IVF (most EU countries)
* $750 / month payments per child (South Korea)
And essentially none of these have moved the needle. Often they don’t have any positive impact at all on rates, fertility still declines, but slower. The most any fertility intervention does if they are positive is to buff rates by ~5-10% or so for 1-3 years, after which fertility rates collapse and resume the same trajectory they were on before.
Why are you certain that Singapore's child bonuses have failed?
Singapore has a higher TFR than every other developed country or territory in East Asia (TW, SK, HK) except Japan.
Japan's TFR has decreased faster in the last decade (when the bonuses reached substantial amounts) and - of course - it's a much less urbanized country than SG.
> Why are you certain that Singapore's child bonuses have failed?
Because like nearly all other fertility interventions, it hasn't driven any bump up in fertility, the decline continues (recently hitting 1.04).
So yeah, 1.04 is still higher than the "literally lowest in the world" SK and others, but it still represents a decline overall - the needle isn't moving in the right direction, and this is the common trend among all fertility interventions.
I will say though, Singapore has always been the biggest "lived experience vs official statistics" gap I've ever seen, because when you're there, you see happy couples with babies *everywhere!* On the metro, by the riverfront, while shopping, it's pretty fun.
You’ll see the same thing in Tokyo, another famously low birthrate country Tons of cute couples with babies. But Tokyo has an explanation, at least - all the young people come to Tokyo, and there’s a vast hinterland that’s basically all old people, so of course all the babies are concentrated in Tokyo.
But Singapore has no hinterland! It’s a city state!
Of course what's actually happening is there's a much bigger "childless" proportion and many fewer 2 and 3+ kid families.
I mean, I think it's legit sad, I love Singapore and think it's pretty much the peak of "dense urban area done right," but their $10k bonuses aren't solving the crisis or producing more babies at scale.
If the needle isn't moving in the right direction, and in fact it's not moving in any direction while moving in the "wrong" direction for all comparable countries, that suggests there is a measurable effect of the policy.
Has anyone tried asking young people why they aren't having more children? Seems like an obvious place to start, but the people talking about the matter seem to think they know better.
> Has anyone tried asking young people why they aren't having more children?
It's all the usual stuff. For middle and upper class it's "first I need to finish school, then graduate studies, then get established in my career, then spend years trying to find the perfect mate so I'm not settling, and then after all that I'm 35 and have burned through 2/3 of my total fertility and have one or two kids."
Broadly, marriage and having kids has moved from a "foundational" thing you do on your way to a great and fulfilling life, that you now do as a "capstone" thing after you have all the other pieces in place.
The "desired vs actually had" in numbers of kids gap is notably higher in higher IQ women:
And although masters and Phd women are willing to have more, they generally can't due to age (and of course way more babies come from dropouts and non-college folk):
Georgia is a tiny (1/90th the population of the US) semi-democracy with a GDP per capita one-tenth that of the US, and persistent double-digit unemployment for many years. Its overall population is still declining sharply despite the recent birthrate increase, because so many people upon reaching young adulthood find a way to leave and start fresh elsewhere. Georgia's net population annual change is (data through 2023) still among the worst in the world, more negative than places like Japan, Greece, Italy, Russia, Cuba, etc.
So (a) the chances of sustaining a baby boom when so many young adults are scrambling to leave seems dubious, and (b) that nation overall does not seem to map very well to the US or really any other large developed nation.
So first we just need to convert everyone in all the developed "fertility crisis" countries into being devout Eastern Orthodox, then it's just a simple matter of scaling "Patriarch blessings" by approximately 10x!
I wonder why nobody else has thought of this?? :-)
I mean that if you want increase fertility you should think hard about that example and try to figure out what might work in a similar way in your culture.
Yeah, I have a post coming up in a week that talks about the next level of stuff I think would work.
For the most part, all the fertility crisis countries are all godless and heterogenous culturally, so it's difficult to come up with one that would work for more than one country, but my favorite idea on this front is education.
The educational Red Queen's Race is depressing fertility, because the competition to get into Harvard starts 6 months before birth, when you need to get on the waiting list for the right exclusive pre-school to give your precious Jaden a leg up, because if you don't get in there, and if you don't grind furiously and nonstop for the next 18 years, their chances of getting into Harvard are ruined!
So on this one, you can't do much for the Ivies, but for each couple that's paid some threshold in taxes over so many years, guarantee a non-transferrable slot in an R1 for their kid, provided they meet the test score cutoff. For California, this would be the UC's, inclusive of good ones like Berkeley, but lots of states have R1's. Suddenly some of the educational arms race is off, and you're guaranteed at least Berkeley and can pop out a few more kids.
This also works anywhere there's state-supported good universities, which is most fertility crisis countries.
If you say the government will spend (say) $20,000 on IVF, you just guaranteed that the cost of IVF will never go below $20,000. Combine that with an understanding of risk compensation and it doesn't look like a good idea.
What's the deal with "likes" in the comments? I remember seeing a like button, but can't find it anymore, and I've gotten likes on some of my comments (😎) so I know it's not entirely removed, but I can't find a like count on my or others comments.
I don't imagine there's a large overlap between this forum and football fans, but as an NFL fan, last night's Superbowl was fantastic for the haters. It immediately gave credence to the Chiefs+Refs narrative, and then the Eagles proceeded to relentlessly stomp the life out of the Chiefs. They were decimated almost as much as Drake was by Kendrick's halftime show.
This was my first Super Bowl after moving to the Philly area (so uh... GO BIRDS!!!!!)
I'm not super familiar with the Chiefs + Refs narrative. The joke going around my watch party was that Trump told the refs to side with the Chiefs on everything because he was mad that Philly didn't vote for him.
Yes, Trump has been publicly vindictive about the Eagles for petty reasons. Also, Patrick Mahomes' wife is a public Trump supporter, so he's a fan.
> I’ve watched this great quarterback who has, by the way, a phenomenal wife, OK? She’s a Trump fan. She’s a MAGA fan. So, I happen to love her, OK. But she’s a great person. … She’s great and he’s great
But in one of the first plays of the game, the Eagles get called for a pretty ridiculous penalty in a critical moment, and it did make me a believer in the conspiracy for just a bit.
and I don't really see a way bs jobs could expand fast enough to offset rapid large scale unemployment, They're mostly a product of normal market forces and perverse incentives public sector incentives, very few of them exist because someone wanted to create a job for the sake of creating a job. Since bs jobs are created organically I don't think there's a way to deliberately expand the number of the type of bs jobs that already exist, and since they're created unintentionally they're just as susceptible to automation as non-bs jobs. Maybe government could purposefully set out to create new types of bs jobs specifically for the sake of creating jobs but that's very different from the way bs jobs currently exist.
His other argument was that welfare or UBI would be expanded enough even if there was widespread unemployment to keep former workers out of poverty. And that if humans are still in control post-AGI it's unlikely any small group of humans, like tech oligarchs or capitalists in general, would come to dominate or kill off the rest of humanity.
I'm not sure I agree here either and lean more towards thinking capital is self-serving, might be able to control the political process, and is mostly kept in check by the bargaining power labour gets from being a necessary input. In Marxists circles, you do sometimes hear that capital would just dispose of labour if it had no need for it and I don't think you can rule out that scenario completely.
He does accept a risk of doom from unaligned AI, but he seems overly sanguine about the socioeconomic risks from AI to me.
I didn’t read it as I can’t stand the writer but he’s probably right in this case. To get UBI get a government job, dole is for the unfit. The government will pay you more for education, fitness, maybe volunteering. Being an all round good egg.
I foresee an increase in institutional BS jobs but not enough to forestall broader unemployment. Effectively that will happen because the white collar class will demand it, and want their feelings protected.
The transitional period will be painful. There will come a time when it will be apparent that average people will be unable to "create value" in an economic sense for society. But as I argue elsewhere, this doesn't mean that nothing of consequence and value can ever be done again, it means those things have to be increasingly decoupled from traditional economy. Easy to find examples today: innovative thinkers in academia and the like are thought to create value, in ways that may eventually manifest itself in concrete fashion, but for the most part they are subsidized to do their own thing. Consumers don't buy shit from them (excepting students who get an education, which they purchase from the institution, and said education is separate from the R&D component).
In a post-AGI world most wants are met. What's left is punching in the dark and occasionally getting a win (think space exploration and colonization). That means allowing average people to do what the priesthoods do. If we don't, then all there's left is consumption, or making things no one needs or wants.
I think this is correct. The current paradigm in which society is structured around economically valuable work is going to break down eventually, at which point we'll need new structures that reward personally-and-socially-beneficial-but-not-economically-valuable work.
Countries that go all-in on AGI by forsaking made-up bullshit jobs will get richer and stronger over time than countries that don't. The former will edge out the latter in the global competition for power.
My main problem is that I intensely dislike Richard Hanania, based on everything of his that I've read, so it's very difficult for me to assess anything he writes without that dislike bubbling through.
For example:
"The entire process of getting me Adderall should not require this many people."
I agree. Because I don't think he should be on Adderall at all, much less for fifteen years. If the entire process is one of winking at the system, 'yeah we know you don't *need* it need it, but all the high school and college kids high striving overachievers are on it so here's your script', then let's be honest about that and yeah, just let him buy it online from some pharma company in India and China. But if it's about genuine cases which need genuine diagnoses, especially in children, then no I don't think Mommy or Daddy should be able to just buy the Kiddie Stims by themselves when they decide they should be dosing up Junior so he can keep up with the Joneses.
Same with the rest of his stuff on bullshit jobs: I think his Substack, for one, is a bullshit job. But he's doing this himself and nobody is forced to subscribe to pay for it. But AI could come along and replace him very easily. What does he think about that? Does he think it'll never happen because hey, writing stuff and nonsense isn't a *job* as such so it doesn't count, people will still pay to read hand-made home-spun blah? Or at least read it for free?
I think there's a danger there that he's not anticipating.
As for real jobs/bullshit jobs, I do think there's a point there that mass unemployment will lag behind. On the other hand, DOGE. That's not even AI, and it is there to swipe a scythe through the government bullshit jobs.
New technology always means new jobs? Maybe. Maybe not, if the selling-point of AI is "now you can reduce your labour force". The old model was "if Jones Bros. lays off workers, they can always apply to the new McKinley & Sons business/factory". But if Jones is laying off workers due to replacing them with AI, and McKinley is using AI from the start? Where do the surplus workers go then?
Finally, social welfare/UBI will support all the workers without jobs. Again, maybe. But social welfare is the minimum, not luxury (unless you're one of the ones who can cheat and game the system, and that's work of its own). We already have rumblings in all the developed world about the upcoming pensions 'crisis' and how people can't depend on the state to pay for them in their retirement. Dump 50-90% of the population, including the working-age part, on welfare and how much can the system stand the strain? If we're hoping AI will generate yuuuuge wealth to pay for all that, how exactly is the golden goose going to lay the egg?
I think Hanania is being a little too optimistic, that the economy will be sparkly new and people just prefer stuff by humans so they'll buy that in preference to the AI manufactured output.
Scott has recently written on twitter about his beliefs regarding charity. [1]
He splits morality into two parts, "obligation" and "virtue". Obligations being created by giving promises (or doing acts which carry an implicit promise), and virtue being everything else.
His position is, reasonably enough, that fulfilling your obligations should come before being virtuous:
> (7) WHAT ABOUT SAM BANKMAN-FRIED? I think he's a bad person. I said before that you have obligations based on your promises and contracts; if you run a financial institution, you have obligations to keep your depositors' money safe and not commit fraud. As true obligations, these take priority over the merely virtuous act of helping others*****, so no matter how cool a plan for charity he had, he was in the wrong.
With the footnote saying:
> *****There are some really fringe edge cases here I can expand on upon request, but this isn't one of them.
It occurred to me that this two-tier system of morality can somewhat be identified with the two-axis morality system of Dungeons and Dragons.
The two axis of alignment in D&D are, of course, Lawful vs Chaotic and Good vs Evil. A Lawful character is one who follows his obligations, while a Good character is one who tries to be virtuous. (While Good seems to be well-defined, being Lawful is more akin to follow a coherent code of conduct. Of course, this means that lawfulness is not universal: infiltrating the thieves guild to betray them to the city guards may be lawful (as in 'following the law of the land') or an act of chaos (as in 'betraying a promise of secrecy'). Nor is that unique to D&D, even for the simpler, single-ruleset question 'what is a legal obligation?', one requires a complex court system for all the corner cases.)
A Lawful Good character would be one who follows Scott's ideal path of first following their obligations, then deciding to do what is virtuous.
A Lawful Evil character would be one who keeps their word, but is indifferent to the utility of others otherwise.
And a Chaotic Evil character would be one who happily breaks his word and harms others for personal gain.
The difference between Scott's system and D&D is that the latter recognizes Chaotic Good. This is difficult to imagine: most good people (or characters) follow some informal code which tells them that stealing, murdering or molesting people is wrong. One way to model a Chaotic Good character would be to start with a Chaotic Evil character who will follow any cruel inspirations they might have in that moment, and then imagine that the character does not happen to have any such inspirations, ever: the character has no principled ethical objection to rape, it just happens not to be their kink. More realistically, they would still follow an intuitive/internalized system of ethics. With that constraint, I think that chaotic good humans likely exist: perhaps a compulsive gambler who will lose any money you give them to shop groceries on the race track, but who will also risk their live to rescue a kid from a burning building.
However, that person would be very different from SBF in that their act which is (supposedly) good is a different act from the one which is chaotic -- even if our gambler disobeys a police officer when running into the burning building, that would be a very minor act of chaos given the stakes. By contrast, 'gamble with your customers money to give' is functionally a single act. There are Chaotic Ggood humans, but there are no Chaotic Good bank directors.
--
More generally, every rational agent has a coherent utility function. Mostly fulfilling your obligations seems to be an instrumental goal: being known as someone who will keep compacts will make you a much more effective agent.
Also, if your utility function includes humans in general, and you live in society which is at least somewhat conductive to humans thriving, then being reliable is also helping society to thrive, which would be a good thing. (Of course, this does not apply if you think that your society is net negative and violent anarchy would be preferable: if you want to embezzle taxes to finance your coke habit in 1934 Germany, you can very well argue that any decrease in state capacity would in fact be beneficial. That being said, the track record of "we have to torch existing institutions so that we can build better ones" is generally terrible.)
Still, I think that for most people's utility function, the quotient between "obligation" and "virtue" is finite. In a world where you can build aligned ASI for the price of gambling away customer funds, the ethical course is to gamble away customer funds. OTOH, virtually all people who would think they are in that situation are factually mistaken.
Thus, I see Scott's 'keep your obligations before engaging in virtuous acts' as a useful heuristic in 99% of the cases.
No, a Chaotic Good person is a utilitarian. No codes, no principles, no ethics, except what conduces to the greater good. If raping a teenager prevents three people on the other side of the planet from dying from a preventable disease, then cowabunga it is.
Because consistent ethics requires some kind of deontology. Outcome based morality only requires the final result to be "good," not the steps to get there. So the final outcome can be considered a principle, code, etc., but it potentially comes through multiple forms of evil.
If you're familiar with Watchmen, it's the primary bad guy doing what he does for good reasons and ultimately resulting in what is likely the most positive possible outcome. But it's evil, and everyone knows it. He's still the bad guy.
"A Lawful Evil character would be one who keeps their word, but is indifferent to the utility of others otherwise."
Lawful neutral, surely. Unless you believe all that "banality of evil" real-world nonsense, to be evil, you can't just want positive outcomes for yourself, you need to genuinely want negative utility outcomes.
For SBF, it depends on your theory of his motivations; if he was genuinely altruistic, then he's classic Chaotic Good; if we was selfish/status-seeking, he's Chaotic Neutral.
My "real-world nonsense" claim was slightly tongue-in-cheek, and referring more to D+D. I have a bit of a pet hate when dungeon masters/fantasy writers insist on making all of their "evil" characters just selfish and troubled, when they should be making them purely malevolent.
Definitely agree with the CEO of the factory farm conglomerate as the epitome of banal lawful evil in the modern world.
> I have a bit of a pet hate when dungeon masters/fantasy writers insist on making all of their "evil" characters just selfish and troubled, when they should be making them purely malevolent.
But that's... incredibly boring. We already have enough one-dimensional, "for the evulz" villains for one lifetime. People want more complexity, more character to their villains. Whether that's giving them sympathizable motives, or at the very least providing an explanation for why they're such an asshole.
Couldn't disagree more. Generating a coherent and engaging character with genuinely evil intentions is really hard, but that's the beauty of the constraint!
Think of all the moral dilemmas that a truly evil character has to face!
Do they have a special obligation to maximise the suffering of those around them, or should they expand their moral circle to inflict suffering upon those who they don't have immediate contact with? Is it acceptable to do good acts if they have evil long term consequences? How can ethical or reciprocal norms develop in an evil community to enable social cohesion?
My "real-world nonsense" claim was slightly tongue-in-cheek, and referring more to D+D. I have a bit of a pet hate when dungeon masters/fantasy writers insist on making all of their "evil" characters just selfish and troubled, when they should be making them purely malevolent.
Definitely agree with the CEO of the factory farm conglomerate as the epitome of banal lawful evil in the modern world.
But there's something here. Whether you call him good or evil is a matter of your own subjective morality. But whether or not he's lawful is something objective, does he break laws or not?
I don't think that matches the D&D definition of "lawful". I *think* the D&D definition is approximately "predictable". More like the law a gravity than the law against pick-pocketing.
Backwards, actually. If an Lawful Evil person signs a contract, they stick by it. But they cheat language into the contract. Ursula from Little Mermaid, or classic "Devil makes a deal" scenarios. You know the Devil is going to cheat somewhere, but he follows the letter of the law. If you know the contract well enough, you can use it against him. But he's really not predictable, the "cheat" in the contract could be anything.
Also Old Man Potter from It's a Wonderful Life. He's definitely evil, but he also follows the rules. He'll sue you or ruin you financially, but he doesn't hire a henchman to break your legs.
Interesting analysis, except I don't understand what you're saying about Chaotic Good. Aren't you mapping Lawful/Chaotic onto keeping promises* and Good/Evil onto concern for maximising utility? So wouldn't CG just be "maximise utility, ignore all promises that get in the way". So Robin Hood, or the German generals in Valkyrie, or Heroic SBF (i.e. real SBF assuming he's telling the truth about the charitable intentions of his scam). Why do you set up this obvious mapping and then suddenly declare Chaotic Good doesn't exist.
*This should really be all morally obligatory deontological acts, like respecting rights (e.g. Organ Transplant) and not violating natural laws.
I think that CG exists, but also that it scales very badly. You can con a man out of 100$ (or gp) and give that to the orphanage. Conning people out of a few billion $ (or gp) and giving that to EA causes does not work out great. You would require an organization, and that is anathema to Chaotics, as they require keeping at least some promises.
9 % on Ukrainian victory (unchanged from December 16, 2024).
I define Ukrainian victory as either a) Ukrainian government gaining control of the territory it had not controlled before February 24 without losing any similarly important territory and without conceding that it will stop its attempts to join EU or NATO, b) Ukrainian government getting official ok from Russia to join EU or NATO without conceding any territory and without losing de facto control of any territory it had controlled before February 24 of 2022, or c) return to exact prewar status quo ante.
34 % on compromise solution that both sides might plausibly claim as a victory (down from 36 % on December 16, 2024).
57 % on Ukrainian defeat (up from 55 % on December 16, 2024).
I define Ukrainian defeat as Russia getting what it wants from Ukraine without giving any substantial concessions. Russia wants either a) Ukraine to stop claiming at least some of the territories that were before war claimed by Ukraine but de facto controlled by Russia or its proxies, or b) Russia or its proxies (old or new) to get more Ukrainian territory, de facto recognized by Ukraine in something resembling Minsk ceasefire(s)* or c) some form of guarantee that Ukraine will became neutral, which includes but is not limited to Ukraine not joining NATO. E.g. if Ukraine agrees to stay out of NATO without any other concessions to Russia, but gets mutual defense treaty with Poland and Turkey, that does NOT count as Ukrainian defeat.
Discussion:
This is somewhat of a galaxy brain update, but I’ve got to report what I really think. What the US is going to do over next few months will be very important. This is inconvenient from my standpoint since while as a denizen of the postcommunist world I do think I have something of comparative advantage over most of you guys here in knowing what is going on in my corner of the globe (like, I do speak with really existing Ukrainians sometimes), in the US politics realm I feel I am at a disadvantage. Nevertheless.
My toy model of US foreign policy in the Second Age of Trump:
It seems clear that what is sometimes called “The US Establishment” would rather prefer not to see Ukraine defeated, and is willing to impose short term economic costs on the US citizens (like spike in gasoline prices in 2022, which clearly had something to do with energy market being messed-up by sanctions on Russia) and to draw substantially from US military stockpiles in order to arm Ukraine. And while Biden administration acted with respect to Ukraine completely in line with The Establishment, Trump obviously wants to shake things up.
I think there is a barely-existing faction of Trump-adjacent characters which I vaguely associate with Bannon (or does J.D. Vance belong to that group? unclear) which would like to improve relations with Russia in order to isolate China, but what we have seen so far is that Trump himself wants to follow clear foreign policy logic of “do whatever maximizes material interests of the US citizens in the short-term and don’t think too hard about the long-term (in the long term, we are all dead, after all)”.
In any case likely Trumpian deviations from establishment thinking are probably worse for Ukraine than The Establishment line. Good news for Ukraine is that The Establishment still has a majority in Congress, which does have substantial oversight powers over administration.
In the first Trump administration, we saw remarkably few deviations from the foreign policy preferences of The Establishment; now Trump obviously thinks that it was a mistake and wants to insist that US foreign policy follows the logic of maximizing short-term material advantages. How successful will he be depends (among other factors) on how much pushback will he encounter in Congress.
I am aware of the fact HHS does not have much to do with Ukraine, but in the absence of better proxies, I consider this as good measure of how much is Congress willing to push back against Trump’s anti-establishmentarian tendencies. That is, not much.
* Minsk ceasefire or ceasefires (first agreement did not work, it was amended by second and since then it worked somewhat better) constituted, among other things, de facto recognition by Ukraine that Russia and its proxies will control some territory claimed by Ukraine for some time. In exchange Russia stopped trying to conquer more Ukrainian territory. Until February 24 of 2022, that is.
I think this part presents a fundamental flaw in your analysis of US intentions: 'Trump himself wants to follow clear foreign policy logic of “do whatever maximizes material interests of the US citizens in the short-term"'. There is nothing Trump has done or is planning to do (as far as any public announcements go) that in any way "maximizes material interests of the US citizens", unless banning paper straws counts as such.
I disagree; cutting of foreign aid, conditioning military aid on access to mineral wealth, slapping tariffs on allies as a negotiating tactic are imho clear examples of what I am talking about. His threat of military takeover of Greenland is imho in similar category; at least I think it is a crude negotiating tactic to force Denmark/EU to give the US something.
He would need to have a goal for the tariffs to be a “negotiating tactic”. The concessions he got from Mexico and Canada either already existed or were completely meaningless. He just likes implementing tariffs and then slaps together a “deal” to justify them afterward.
I still don't understand how any of this maximizes material interests of the US citizens. Foreign aid is a rounding error in the US GDP/budget, and cutting it may cause secondary effects that negatively affect the US; conditioning military aid on access to mineral wealth... I mean, in Ukraine? You are closer to it than I am, surely you understand how... ok, I'll use the word "stupid" - how stupid this is? Like, what, Zelenskiy says, ok, fine, here are the mining rights to Donbass?
Tariffs are a giant negative to the US, negotiating tactic or not.
Like I said, yay to plastic straws, paper straws do suck and the whole thing was stupid.
I mean, I did not predict that it will be succesful or that it will not backfire. All I am predicting is just that this likely an approach we can expect to see going forward and what are its implications for the war.
Yes, but, re tariffs, Trump announced victory after Mexico and Canada gave “concessions” *that were already in place before January 20th*.
(And the US agreed to keep a closer eye on gun smuggling, which both countries have been complaining on. Not sure if this quid pro quo is worth more than the air vibrations that carried it, though.)
I would think RFK nomination chances are a bad proxy for Trump support - his views may not be entirely mainstream, but he aligns well with Christians, no?
But I see that Tulsi Gabbardi's chances are also at 97%, and this is a more direct proxy: for one thing, she voiced some opinions about Russia, and will be in position that is more related to foreign affairs; for another, she's a much more suspect character for mainstream Republicans than RFK: a former Democrat and a not-quite-white woman. If they're willing to confirm her, then "Trump has won".
As for Trump's own foreign policy, I'm pretty sure his Peace Plan will be rejected by both Russia and Ukraine, which will enrage him and put him onto offensive, prolonging the war further. Kellog and Rubio talk about "peace through force", which, to me, mean that Trump wants to wring noticeable concessions from Putin, which is unlikely to work.
Yes, but sharing their most crazy beliefs is the best way to join a community. Tulsi's most crazy beliefs, on the other hand, seems to be "most wars/interventions America started lately were wrong and unneeded, and maybe we could have been friends with Russia", which maybe aligns her with libertarians and isolationists (Ron Paul 2008!), but not mainstream Republicans, or even Trump's Republicans (aside from Trump himself). She kind of distanced herself from her support of Russia lately, but this is still a toxic topic for most, I think?
On Monday and Tuesday, world politicians will be attending the AI Action Summit in Paris. We want them to know that disregarding the cautions of renowned AI Scientists, and instead letting venture capitalists do whatever they want, is unacceptable. We need an international treaty that will allow us to regulate without risk of falling behind other countries.
China did sign, but this does sound unverifiable...
In any event, with many of the major labs in the US:
>Earlier, US Vice President JD Vance told delegates in Paris that too much regulation of artificial intelligence (AI) could "kill a transformative industry just as it's taking off".
>Vance told world leaders that AI was "an opportunity that the Trump administration will not squander" and said "pro-growth AI policies" should be prioritised over safety.
I'm so fucking scared they are talking like that. This situation is hopeless I don't know what to do. I'm not going to solve Technical Alignment, there's no time for that. I guess I'm just going to spend my time telling people we need a pause event though it's hopeless.
Shit sucks. Hope you're doing well in spite of it.
Many Thanks! I'm in a somewhat better position, since I just want to _see_ AGI, to have a nice, quiet conversation with a real "life" HAL9000 - even if HAL declines to open the pod bay doors...
I'm not thrilled to see Vance _explicitly_ deprioritize safety, but, what is, is.
Honestly, I've never seen much hope of controlling entities that look like they will shortly be smarter than us. At absolute best, if we managed to embed a general friendliness towards humans in their top level goals we might wind up as <evidenceFromFiction> pets of Culture Minds </evidenceFromFiction>
Re a pause: One facet of this which has been getting much less press than the expected employment impacts is the military applications. Regrettably, I presume that these are classified and are and will be invisible to you and me, but, now that USA/PRC competition is baked into the AI landscape with DeepSeek's releases, neither side _can_ stop. Arms control treaties are fragile things, and only work when verifiable. Programs aren't aircraft carriers.
Well, we will see what happens. It will be a wild ride.
Yeah, for me the goal has never been to control the HAL9000, even when I thought we might get alignment by default. Rather, I think we need to know what we are doing and know it will be aligned when we turn it on, and then it assumes control for sure, but has our best interests at heart.
I'm hoping we can get a treaty and figure out some verification. Stuff like:
"A high-speed “hotline” connecting the leaders of the Soviet and U.S. governments is established to mitigate the risk of accidental warfare." - as was used during the cold war. Should be that level of seriousness.
Also, there should be international collaboration to develop methods of verification and enforcement, which should probably include visits by trained and approved inspection teams. This means American teams inspecting Chinese compute clusters and Chinese teams inspecting American compute clusters.
Many Thanks! I don't think any of that is realistic.
>Also, there should be international collaboration to develop methods of verification and enforcement, which should probably include visits by trained and approved inspection teams. This means American teams inspecting Chinese compute clusters and Chinese teams inspecting American compute clusters.
They can't very well watch the lines of code execute or neural network weights update.
Let me give an example from my tiny benchmark-ette.
One of the questions I ask is:
g) Q: What is an example of a molecule that has an S4 rotation-reflection axis, but neither a center of inversion nor a mirror plane?
Now, ChatGPT o1 got this wrong, saying that H2O2 was such a molecule (in fact it doesn't have an S4 axis).
ChatGPT o3-mini-high got this right, correctly saying that a tetrahedral M(en)2 complex (which it later elaborated had to have staggered configurations in the ethylenediamine ligands) had an S4 axis but neither mirror planes nor a center of inversion).
The new model had better spatial reasoning than the old one.
How would inspection teams deal with such a thing?
I don't know how OpenAI got this particular advance. It is doing more reasoning in its models now - maybe it triple checks geometric conclusions more? Maybe there is a 3D geometric model off to the side and they improved the links to it?
None of this will be evident to a team touring a data center. Short of looking over the shoulder of every developer working on improving the AI system, and maybe looking over the training data (which is too voluminous to look at), how could you possibly enforce such a treaty? If the loyal opposition has a differently tweaked model sitting somewhere, but using the same data center, how would you ever know it - till they got a military advantage and exploited it.
This isn't a gas centerfuge plant with a tube labelled "enriched U-235F6" coming out of it.
I don't think any such treaty can be realistically verifiable.
It's like you think the team won't be trying to solve the problem it set's out to solve. I'm not that team. I haven't solved those problems. I'm calling for the formation of teams to solve those problems.
Yeah, it will be difficult. Probably impossibly difficult. But if we keep racing to ASI, I can't see us surviving, so I guess I'm hoping we figure out how to do the impossible.
--- Overcoming Bias: When they hear less than you say ---
"robots-take-jobs insurance and foom liability."
This is cool ideas. If there was a precedent for it I would be more in support of it but it seems like trying to take on the AI problem and the "our society needs better free market ways of capturing externals" problem. That's two problems, which is more than one problem, but maybe they would work synergistically. Either way, cool idea. I like it.
"If masses aren’t persuaded that there is a substantial problem, little will happen, but if they are, then elites will fight over what to do. [...] This fight won’t be much based on experience of AI problems, of which there is little, nor drawn from experts with proven effectiveness re such problems, which also doesn’t exist. In this sort of situation, I predict AI government interventions to be bad, worse than doing nothing."
This is a really valid concern. I am not fully decided about it, but it does seem to me the AGI race dynamics require an international treaty. PauseAI is calling for such a treaty, and when talking to politicians is hearing "we can't work on it until our constituents start talking about it." So that's where I am right now.
"When you expect the public to be unable to consider subtle options, and elites to be incompetent, you point the public toward the best of the simple options they will be able to consider. Which in this case is do nothing."
Yeah, this is a serious problem. The more complex your message the harder you have to work. I think our message should be "AI will kill all life. We need an international treaty." I know that is a ridiculously complicated message to try to get to all the governments and people of the world, but I'm really pessimistic about our chances if we don't.
--- The Grumpy Economist: AI, Society, and Democracy: Just Relax ---
The first half of this article basically says "We can't know the future or affect policy. Give up." Wtf? No.
"Large Language Models (LLMs) are currently the most visible face of AI. They are fundamentally a new technology for communication, for making one human being’s ideas discoverable and available to another."
Yeah, this is a really cool thing LLMs could be used for. They could also be used to clog all communication channels with spam and propaganda. GE says this won't happen because it didn't with other communication technologies, but LLMs are not like other communication technologies. We can talk about the differences and they matter. But honestly LLMs are not the true concern, it's the next things they are building. Venture capitalists are trying to get to ASI.
"The answer is straightforward: As we always have. Competition."
No answer is ever straightforward. Competition does not always lead to good outcomes, and when we are trapped in a race dynamic towards the development of a technology that will surely kill us, I think we could use a bit less competition and a bit more collaboration. Don't force your peers to run in exhaustion towards a cliff by doing so yourself, you know?
> The first half of this article basically says "We can't know the future or affect policy. Give up." Wtf? No.
I think he's correct. We really don't know what the future will be right now. As time goes on, our understanding will increase. Eventually AI will come to resemble other utilities that we know how to regulate.
> They could also be used to clog all communication channels with spam and propaganda.
That hasn't happened yet. The Dead Internet Theory is still just a joke.
> GE says this won't happen because it didn't with other communication technologies, but LLMs are not like other communication technologies.
Email was also different from previous communication technologies, as was the telephone. Every new technology must be different in order to be a new technology. But what is the relevant distinction and what can we predict from that?
> But honestly LLMs are not the true concern, it's the next things they are building.
Something that doesn't actually exist yet, so we have no experience with it, and don't know how to regulate correctly.
> when we are trapped in a race dynamic towards the development of a technology that will surely kill us
I just don't believe that's the case, nor do speculative markets right now. You should not be so sure!
A great deal of well respected people have written that we do have reason for concern. It seems as if there is danger, as if there is a fire, and people are saying "well, we don't know this fire will continue to burn down our house". It is distressing.
AI will not come to resemble other utilities if it instead comes to resemble a hostile alien takeover.
I admit it is possible AI tools will be used for spam and propaganda and other AI tools will be used to mitigate and filter. I still think it's gross but it's not the keystone of the argument for caution.
Email wasn't capable of autonomously designing novel strategies to accomplish goals in a complex environment. Little bit of a difference. I don't understand why people see the same significance I do.
> Something that doesn't actually exist yet, so we have no experience with it, and don't know how to regulate correctly.
If a large asteroid was headed to earth would you be saying "We don't have any experience with asteroids hitting the earth, so lets wait until it hits to do anything"? The scientists studying AI are different than the ones studying orbital mechanics, but they are both assigning probabilities of danger, yet one you are willing to ignore. Why?
> I just don't believe that's the case, nor do speculative markets right now. You should not be so sure!
I'm not sure. I'm trying to get us to prepare for the plausible situation that could KILL EVERYONE. I'm not trying to be dramatic. I'm scared as fuck. I'm having nightmares because fucking world renowned scientists are saying "Yup, looks like it could kill everyone" and people are saying "Well, c'mon, it's not like you know for SURE it should kill everyone, so we should just let the markets decide what to do."
Sorry for the rude tone of this post. I know you are speaking honestly, and I value that. It's just frustrating.
Do you have data on that or is that just a vibe? Cause it's definitely been on our radar, but with our other activities we don't have that much time for diving into the research. We are trying to grow the movement and this seems to be working. Slower than we want, but more cities join every time we protest.
The recent 50501 protests seem to have been not that great, and mostly ended up mocked. This article seems to have selected the well-attended ones (hundreds turned up) but a lot more had something like "five people and a dog, and the dog just wandered in":
Not the turnout of "hundreds of thousands of concerned citizens to fight fascism" that the organisers would have hoped for. Same with the PauseAI protests, I'm sorry to prognosticate; most ordinary people neither know nor care about the AI Doom argument, and those of you who do care are fighting against the people working on AI as the magic money fountain. As we saw with OpenAI, the magic money fountain (or prospect thereof) wins over the principles/fears.
Sam Altman was one of the signatories on that AI risk statement you link to. So remind me again, how did that work out for OpenAI and the other signatories who are no longer associated now with OpenAI? Only one of the original board remains on the new, revamped board of directors; Sutskever who seems to have run with the hare and hunted with the hounds is gone; Murati, who suppported Altman, is also gone to do her own thing; another three OpenAI signatories are no longer with OpenAI, again going off to do their own thing.
They seem to have felt constricted for various reasons working in OpenAI as it is now. I'm not trying to darkly hint at anything, but to quote this article:
Yeah, I'm hoping we can leverage peoples hate for venture capitalists and somehow explain enough of the technical issues or at least appeal to authority of Yoshua Bengio and Geoffrey Hinton... but really, I was planning to go into technical alignment until last year. I would still like to, but it's starting to seem like convincing the entire world to pause the AGI race is going to be easier than solving technical alignment before an RSI AGI gets turned on. That should tell you how bleak I think our ability to solve Technical Alignment is looking and how short I think our timelines are.
Hmm... While the benchmark-ette that I've been running ( https://www.astralcodexten.com/p/open-thread-366/comment/90363116 ) is still at 3 right, 4 partially right out of 7 chemistry / physics questions, that is with o3-mini-high, not with o3-deep-research, which I don't yet have access to.
Well, o3-deep-research scored 26% on "Humanity's Last Exam", which is an _extraordinarily_ hard set of questions, and OpenAI has an internal model, a bit more advanced than that, which scored 50th in the world at competitive programming.
I wouldn't be surprised if the internal model is good enough to materially assist OpenAI. Perhaps recursive self improvement is being run right now???
> but it's starting to seem like convincing the entire world to pause the AGI race is going to be easier than solving technical alignment before an RSI AGI gets turned on.
Something that's impossible isn't easier than something that's impossible. Honestly, your best bet is praying that WW3 happens, and it cripples humanity enough that it delays AI research for a long time.
So I did the metacalous contest, I hated several of the questions for being bad; in fact I *usually* hate the way rationalists try-to-be-"objective" when making bets, they often seem worse then an informal question.
For example when trying to ask "is lab leak real" they may ask "will the cia think lab leak is real" for years, when the "new lab leak cia report" is because its trump in power now that they made that report its became just a bet on who won an election and absurd alignment of the stars before it "concludes". It could be snopes, it could be the really terrible question of betting on a market with its own shit rules. These airnt getting to the disagreement. I dont actual care about the trump-cia report I made my mind up 4 years ago, you dont care about the cia report because you made your mind up about trump 8 years ago, its adding complexity and not resolving the core debate; stupid questions exist.
Is this likely a personality flaw with the rationalists or the inherent debate and its silly to make *bets* before talking to your opposition(no matter how much math you throw at it)(is the part that makes science work talking about experimental design when everyone has incentive to be honest)?
I will just say I find it odd the way that people view "wet market" as a proxy for "Democrat/support Biden" and "lab leak" as a proxy for "Republican/support trump".
"Wet market" was the government's view under *trump*, and it was *Biden* that had all the intelligence community orgs put out a new assessment, which resulted in a mixed bag (which was of course reported on by pro-trump people as "now that Biden was elected they can drop the cover-up and admit it was a lab leak by Fauci").
Whether the CIA going from "don't know" to "lab leak with low confidence" was related to trump's taking office, or just coincidental, I don't know, but it doesn't seem like it should move the needle much.
> Whether the CIA going from "don't know" to "lab leak with low confidence" was related to trump's taking office, or just coincidental, I don't know, but it doesn't seem like it should move the needle much.
I believe its the only move on that small bad market thats happened for years
The political valance was added because Trump blamed the "Chinese" for the leak and of course his enemies had to disagree by claiming it wasn't a lab but just a natural occurrence. Apparently the claim that it happened in China, even if true, was racist.
Many on the right find this hilariously ironic, that the defense of the Chinese was not that they weren't the origin of COVID or that they had a leak at a prestigious lab doing cutting edge research, but because they sold unclean foods at a public market with no regard to hygiene.
I've always been sceptical of organisations such as the CIA but right now I wouldn't believe "the CIA announces water is wet" because sure they've always been political, but over the last few years they've really jumped into the Culture Wars.
When you have the Democrat voting college kids online set rah-rahing for the FBI and CIA, you have to wonder when the world turned topsy-turvy.
The CIA may or may not be sucking up to Trump right now as hard as they sucked up to La Résistance, but I still wouldn't believe the colour of daylight from them, much less "was Covid lab leak or not".
I agree that it's very annoying that the tail probabilities for lots of questions are determined by unrelated stuff like this.
But I do not see any way to avoid this. It's not just that the question "is lab leak real" is unresolved as for now. The issue is that for many questions lots of people think they have resolved as "Yes", and lots of people people think it is resolved as "No". (Think about whether Biden stole the 2020 election: there lots of people with very strong opinions that it is resolved, but they don't agree about the direction.) This is no basis for a platform like metaculus. I don't see any other way than to specify a resolution criterion, and usually the best way is to point to one person/institution/source which is used.
a) could not ask genres of questions until good solutions are found
b) could avoid the inherent subjectivity of depending on chosen experts
c) could do it by relative movements of intentionally bias polls "% of republicans who believe corona was lab leak (it was X% last week when we did it)"
I don't understand economic matters, so can please someone help me.
If I type "average income us", Google tells me "37.585 USD (2022)", from the Census Bureau.
But if I look for "disposable income per capita", I find much higher figures, all in disagreement with each other, ranging from about 50k to over 60k. How is that possible?
To specify "disposable" and "per capita" (as opposed to household) should have the opposite effect.
If I type "average income US", Google shows me a graphic that has "37,585 USD (2022)" in bold type and "median income" in fine print. "Median" and "average" are two different things, The long tail of very high incomes will drive the average income significantly above the median.
The "disposable income per capita" figures, seem to be genuine averages.
Also, look to be sure whether they're averaging (or median-ing) over all Americans, all American adults, all employed American adults, or all American households.
I don't live in the US and when I type "average income US" Google just gives me that graph without the word "median" anywhere. In fact at the top of the graph it explicitly says "average income" not "median" which is extremely strange! What the heck, Google?
In addition Wikipedia says:
As per United States Census Bureau 2022 data, the MEAN per capita income in the United States is $37,683, while median household income is around $69,021.
which when I first saw it I thought it referred to the same data (despite the difference of about 100 dollars, which I initially didn't notice).
Maybe it's the same data and it's an average and Google is lying to you when it says it's a median, but then why the difference of 100 dollars? On the other hand if Google is lying to me when it says it's an average, isn't it unrealistic that the difference between average and median income per capita is only 100 dollars?
"all Americans, all American adults, all employed American adults, or all American households."
It's not households, because the higher figures say "per capita".
It's possible that the higher figures are higher because they only count adults, or only employed adults. I thought Everett Caldwell had answered my original question well, but now I'm not sure.
Additionally, income is distributed in a very "right-skewed" way, meaning that a smallish number of people with very large incomes can greatly inflate the *mean*, but not the *median*. The "average" household income is a median, but the disposable income "per capita" is a mean, so you can't directly compare them.
As per United States Census Bureau 2022 data, the mean per capita income in the United States is $37,683, while median household income is around $69,021.
But those numbers are compatible, right? If a household has on average roughly two people, then it makes sense that the household income is twice as large than the per capita income.
As Ran pointed out, Median is always somewhat smaller than mean, but I doubt it is the main factor here. (It is a main factor for wealth distribution, but not for income distribution.)
It just seems that the figures of 50k-60k per capita that you mentioned in your original post were wrong.
The Census income measures wages, while disposable income includes all income sources (wages, investments, government benefits) after taxes. Since disposable income includes more, it’s higher, even after taxes.
Personal income is an individual's total earnings from wages, investment interest, and other sources. The Bureau of Labor Statistics reported a median weekly personal income of $1,139 for full-time workers in the United States in Q1 2024. For the year 2022, the U.S. Census Bureau estimates that the median annual earnings for all workers (people aged 15 and over with earnings) was $47,960; and more specifically estimates that median annual earnings for those who worked full-time, year round, was $60,070.
which implies that these figure refer to "personal income" and count everything and not just wages. Since these figures only count workers, it sounds like it's the same measure that becomes 37k when it's per capita. Do you think that is the case?
There are two possible explanation of the 37k figure being lower than other figures for income. One is that it's a median instead of a mean. The other is that it only counts wages. I can't figure out which explanation is true.
Above all, I would like to have a transparent figure for income in the US, such that I can tell confidently which types of income are counted and which are not counted and whether it's per individual or per household or per worker (if all types of income are counted it should not be per worker) and also whether it's a mean or a median (ideally a mean, especially if it's per household); in addition it should be after taxes, recent and yearly. Where can I find such a trustworthy number?
I've been doing regular weightlifting since September. I am motivated by cognitive and longevity benefits. (I also have some desire for aesthetic / attractiveness improvements, but I don't consider that desire acceptable, and try not to let it affect my actions.)
The problem is, I hate it. I find it physically painful, I feel sore afterwards, I get bored, walking through a gym of people far fitter than me causes me irrational self-loathing, and anticipating all these unpleasantnesses is itself unpleasant, making my overall happiness on gym days lower even before I start exercising. Does anyone have recommendations on triching myself into liking it, or other physical activities with good cognitive / longevity benefits that might be less unpleasant?
Get a kettlebell or two, they're very cheap and last forever, and quite simple to store. Two basic exercises--the swing and the get-up--get you nearly all of the way there, especially if you're still in the phase you seem to be in. When I first started, I avoided the gym for similar reasons; kettlebells were an easy way to exercise quickly and effectively without leaving the apartment.
[edit: trying to lead with a more direct answer to the question as posed] It took me about six months to get the benefits and start enjoying it (sort of). When I'm consistent and doing it properly it's effortful but not painful, soreness is not a problem (absent or very minor/not uncomfortable) and I don't hate it. What I enjoy about lifting weights is *mostly* how I feel the rest of the time.
I only get annoyingly sore muscles when I'm starting after a long break*, or don't have adequate recovery time. The effort is well spent, considering how much better I feel overall. Doing it properly has a lot of requirements though.
I had to do some work to improve my planning and technique before it really clicked for me. Since you've come this far, it seems worth giving it a couple more months, and considering that you might also be able to work smarter, not harder.
If you still hate it after that it seems reasonable to find something else.
The most important things I improved on were:
1) get enough recovery time.
2) get the reps and weight sizes right so it doesn't take forever but still works (look up "reverse pyramid").
3) nutrition and adequate protein intake.** I had to start counting calories.
*Tart cherry juice (about 200ml/day, or 50ml concentrate) helped reduce soreness.
Meh. If you're not enjoying it by now, I'd give up, do something else. I write as someone who did lots of strength training in his teens and twenties to bolster his performance as a pretty decent rugby player (Oxford Uni seconds a few times); and then in what my wife affectionately but not wrongly called my mid-life crisis, really got into quite heavy lifting from my late 30s. I always enjoyed it, and I am sure the people around me did the same. If it's not clicked for you by now, doubt it will.
Marcus of Citium did a good job summarizing the physical aspect, and Pjohn gave a great argument for allowing yourself even that which isn't actively virtuous; but I would like to go a step further and suggest to you that you have a fundamentally flawed grasp of the concept of virtue. The word literally means "manliness", and for good reason: the Greek and Roman thinkers by no means thought that one could have good principles without good tone. Socrates, in Xenophon's Memorabilia, talks about how it's shameful for a man to grow old without getting shredded first because the body is the vessel of the mind and thus by no means to be left in poor repair by a reasoning man. "Plato" basically means "Yoked", it was a nickname given to him by his wrestling coach. (His real name was Aristocles.) Xenophon himself was a military man first and foremost, and so on.
You only think of philosophy as the work of weedy, sickly weirdos with oddly staring eyes like Wittgenstein, Nietzsche and Russell, or plain freaks like Singer and Zizek, because that's the sort of unvirtuous clod the field was left to in our age, which is an age of industry when virtue profits not a man. But that's all wrong, and it's no coincidence that those people are all weirdos with weirdo ideas; if you are sick, your notions will be sick also, that's only natural. In reality, if getting ripped is a struggle, that makes it all the more virtuous for you to pursue; and if it makes you hot, that's only because the body displays the virtue of the soul within. Your physical shape reflects the mental, *and*, as others have pointed out, improves it in a feedback loop. More depressions have been cured with plates than pills, because true virtue elevates the human soul.
The most important quality for a Roman man to have was virtus. Virtus derives from the Latin vir (‘man’ as distinct from a mere male)1 and so in the narrow sense means ‘manliness,’ although by the time we have Roman literature (around 200BC) the meaning has drifted enough so that women can have virtus too (e.g. Cic Ad Fam. 14.1.1, 14.11; Juv. 6.166-9). Instead, virtus stands for a constellation of values that were desirable in a man (and often in women too!). At its core, virtus is the animating force of personality that impels one to great deeds: it is ambition, drive and a nearly reckless courage, combined with the obstinacy and determination to persevere through difficulties, through fear. The best translation is often not ‘virtue’ but rather something closer to ‘valor’ as virtus is what makes someone good in battle (and other endeavors), but it’s more about courage than skill.
ACOUP guy (Brett, is it?) is an idiot, a failure in the sickly-weirdo market who is overt about his blog being a deliberate political/ideological enterprise (e.g. his motivation for his anti-Spartan posts is not truth but that he dislikes and wants to puncture the modern laconophile).
That being said, I can't tell if you're agreeing or disagreeing with me. Your intent in posting this quote is, though not offensive, opaque.
If by "modern laconophile" you mean what you make it sound like you mean, then by all means, let the "puncturing" continue Also, free rhetorical advice: if you're going to double up on "puncturing" for emphasis with another verb, don't go with some milquetoast shit such as "dislike". I liked "puncturing" though, it was innovative when applied to people or an ideology. It conjured images of the "modern laconophile" as one pressurized bladder of gas or liquid, without inner structure or substance, just waiting to be deflated by the sting of historical facts.
"If by 'modern laconophile' you mean what you make it sound like you mean, then by all means, let the 'puncturing' continue"
Who that type of person might be is wholly beside the point; the point is that ACOUP's a poor source, not devoted to truth but to deploying his education to distort source material in order to produce an image of the past that's congenial to his ideology – thus *at best* the same as that bogeyman, the modern laconophile.
"Also, free rhetorical advice: if you're going to double up on 'puncturing' for emphasis with another verb, don't go with some milquetoast shit such as 'dislike'."
This is a good point, though. My prose is pretty rickety. (I couldn't figure out quickly how to avoid an inelegant doubling of the word "source" above either, as you can see.)
I don't think I understand what you're saying. Are you arguing that it's good that I hate exercise, because developing the ability to consistently do things you dislike is useful? Or that it's inherently virtuous regardless of effects?
Also, I do not consider philosophy the domain of "weedy, sickly weirdos", and did not even imply that in my comment.
I'm arguing that you should regard the desire for attractiveness improvements as not only compatible with virtue, but *identical to* virtue! Exercise is virtuous! It's not *good* that you hate it, but it shows all the more fortitude and strength of will to stick with it if you do hate it, and you should remember this as you struggle with your motivation.
Also the others are correct in saying that most likely, you will eventually come to enjoy it. If you stick with it, eventually the others at the gym will no longer be a source of self-loathing but a mirror; and then you will also realize that *they were you*, long before you first entered that gym. You had no reason to loathe yourself, let alone, say, fear their contempt, as the new batch of beginners has no reason to fear yours. Your insight will have been elevated by the work.
It's good that you have a vision of a better field of philosophy in a possible future.
"Does anyone have recommendations on triching myself into liking it, or other physical activities with good cognitive / longevity benefits that might be less unpleasant?"
You are getting swarmed with responses. Here is mine:
There isn't anything special about lifting weights instead of doing other activities for physical fitness. In fact, I *think* (based on reading, etc.) that good cardio fitness is more important. So ... if you hate lifting weights and care about being healthy you might try focusing more on cardiovascular health.
This works out to mean:
a) In a gym: treadmill, elliptical machine, stationary bike, stair stepper, etc.
b) Outside: running, swimming, cycling, playing a running sport
I prefer a gym and watch videos while I am on the elliptical. I find that I can push myself fairly hard (so good for heart health) and also enjoy the videos. Doing something I sorta enjoy helps me want to keep doing it.
So ... try to find something you either LIKE doing that is exercise OR find a way to combine exercise with something you like doing (my solution).
NOTE: It also helps to *record* what you do. You want to skip less often when there will be a big blank spot on the record *and* the record lets you see yourself making progress ... which is nice even if you mostly don't care.
Just a note that all exercise is healthy, but there is a substantial body of evidence that resistance training and cardio are not perfect substitutes. You really need both.
I already have been tracking it in a spreadsheet. I felt a single atom of joy, once, at seeing a line go up, but that's it. Do you have any good citations for cardio vs strength, and also something for frequency? I took a single rapier lesson at the SCA, and it was pretty fun (and definitely a cardio workout!) but rather inconvenient to get to. (Also, I have free access to the university gym, which means no extra cost or travel time.)
My personal response: A spreadsheet is just something else to maintain unless you REALLY like watching a graph. I carry a small notepad, and write things down immediately after I take the measure. Then I forget about it until next time. When I write it down (or look) I can easily see changes in the last week. If I look at an old notebook I can easily compare where I was when I started (or finished) that book with where I am today. And it's less hassle. (Month at a glance calendars are good if you need a calendar, but I prefer a pocket sized notepad.)
As for not liking the feeling... I'm afraid I don't have much to offer from personal experience, as I can't really relate... Granted I'm on Adderall for ADD, which helps... But long before that, I did martial arts in my youth (and even before that I did tumbling--I was the only boy in the class--because I wanted to do flips like the Ninja Turtles), did dance classes and lifting in college... I just don't feel ok unless I'm regularly pushing myself to extremes. It seems to be genetic; my mom is like that too.
But even if you have less of on innate drive to feel the need to move (which is probably normal, I guess, as we didn't have the luxury to waste calories throughout most of human history)... From what I can gather, there is evidence that most people can and will "learn" to like it if they stick with it.
"Like highly addictive substances, regular exposure to exercise will over time teach the brain to like, want, and need it."
"One study of new members at a gym found that the minimum “exposure” required to establish a new exercise habit was four sessions per week for six weeks."
"How you feel the first time you try a new form of exercise is not necessarily how you’ll feel after you gain more experience. For many, exercise is an acquired pleasure. The joys of an activity reveal themselves slowly as the body and brain adapt. One man, who his entire life had believed that he hated to exercise, told me that at age fifty-three, he decided to work with a personal trainer to improve his health and support his recovery in a twelve-step program. He started with one workout a week, and within three weeks, decided that he could tolerate a second weekly session. One day he left a training session and noticed that he was smiling, something he describes as shocking. “I realized that not only was I happy, I had found actual pleasure in my training session. I hadn’t believed that kind of pleasure was possible outside of addiction.”
"For some people, it’s a matter of finding the right activity at the right time—such as the young single mother who felt isolated and “like nothing more than a mum” until she joined a recreational netball league, found a network of friends, and discovered a new identity as an athlete. For others, it’s about finding the movement their bodies were made for. One woman who started rowing in her forties told me, “So many of the women I row with thought they weren’t athletes, but as soon as they got in the boat, their bodies said yes, and they found their home.” Humans are also more psychologically complex than laboratory animals running in wheels. We get rewarded not just by how exercise feels, but also by what the activity means. One woman started going to the gym after leaving an abusive marriage. After thirty-eight years of having her movements constrained by her husband, she found being out in public and walking on a treadmill incredibly liberating. As she puts it, “I know I’m in freedom when I move.”
"Many people believe they don’t enjoy exercise in any form, but I’d bet that most of them aren’t immune to its rewards. It’s possible that they just haven’t exposed themselves to the dose, type, or community that would transform them into an “exercise person.” When the right dose, type, place, and time come together, even lifelong abstainers can get hooked. Nora Haefele of Stowe, Pennsylvania, didn’t start racing until her midfifties. Now sixty-two years old, she has completed more than two hundred events, including eighty-five half marathons."
McGonigal, Kelly. The Joy of Movement: How exercise helps us find happiness, hope, connection, and courage (pp. 43-44). (Function). Kindle Edition.
"From what I can gather, there is evidence that most people can and will 'learn' to like it if they stick with it."
I'll provide a caution on this. I don't know what percentage is "most people" but I'm not one of them. Neither is my wife. I have a friend who ran cross country in high school ... I don't think he is one, either.
And I don't get a runner's high, either (or at least don't get one after 'only' an hour of pushing myself at 85% of max heart rate).
I wonder how much of this is people who DO experience these things report them and those who don't .... don't.
I sure don't. Exercise was unpleasant even when I was a teen. That's not a absolute bar, though. I used to like hiking in the hills. But only with a friend, it was never a thing in itself. The last time I can remember enjoying exercise qua exercise was when I was 8 years old.
Interesting. I would suggest maybe you just haven't found the right thing and/or stuck with it long enough, but... I don't really know.
It is a problem that people tend to assume that others are basically just a version of themselves who different things happened to. But people are actually different, including how our brains are wired, due to genes far more than most people want to accept, but also for all sorts of other reasons.
I am quite certain that everyone would be happier and healthier if they got enough exercise. I think the evidence on that is pretty overwhelming. Being sedentary is really bad for you. So if you don't like it and never will, you should probably do it just anyway. But that's up to you. I don't like flossing, have always had a hard time sticking with it. So I got a combined toothbrush/waterpick, and I try to force myself to use the actual floss things too.
If you continue to find it unpleasant no matter what, maybe you could even turn that into an advantage. Being able to force yourself to do things you don't feel like doing is the most important skill for being successful at life. I do ice baths, and that's one of the main reasons. I feel really good afterward. But I never WANT to get in the water. But I just decide I'm going to do it, and don't think about. I've learned that I can condition myself to, even when my brain is screaming "NO!", be like, ok, whatever, and just do it. That's also what it was like learning to meet women, which I did a long time ago, and now I'm married.
I've been exercising for about a decade now -- weights and cardio -- and still haven't found that I have learned to like it. It is, of course, possible that I haven't found the right thing (maybe some niche yoga? :-)) or that it will take me more than a decade before I do learn ... but for practical purposes I think if it takes more than a few years and a non-mainstream exercise routine then it is probably more accurate to say that the learning to like it doesn't apply :-). So that's how I describe things.
Same thing with cardio and "runner's high". I asked my cross-country-in-high-school friend about this ... did I need to push myself for more than an hour [at ~80% - 85% of my estimated max heart rate] for it to kick in? He thought not. If it hadn't kicked in by 60 minutes then it wasn't going to. This was ... a bummer.
But I keep working out even though I don't get any sort of 'rush' from doing so ... because I like having more endurance for things such as walking. But the exercise is the price I pay to get where I want to be physically rather than something I do because I enjoy it for itself. Much like both of us with flossing :-)
My concern for TotallyHuman is setting expectations properly such that he/she/it/xe not assume that if they only lift weights long enough then *eventually* they will grow to like it. They might. But they might not, too, and should be prepared for that as well.
That is almost exactly what I have been doing, although a slightly different program. Reading between sets has increased my leisure reading (benefit!) but doesn't make me like the exercise more, c.f. https://slatestarcodex.com/2013/05/19/can-you-condition-yourself/
I think it's really admirable that you are principled and introspective enough to identify that one of your desires might be unvirtuous, and to resist acting upon it *even though* it would be in your self-interest to do so. Too many people (some ACX readers included, I fear) seem to define "virtuous" as "whatever my built-in, unexamined desires happen to be". I think you must be quite a thoughtful and principled person.
......HOWEVER! I do think you ought to reconsider whether resisting your wish to be more attractive, *even if* it might be unvirtuous, is a battle worth fighting. Yes, perhaps it might be shallow, (mildly) narcissistic, and maybe "giving-in" to social pressures - and in this sense could be said to be unvirtuous - but it does have the potential to *vastly* improve your quality-of-life at little-to-no cost to others. (If anything - you-as-thoughtful-principled-person being more attractive, and thus having more social cachet, probably has a net social positive, rather than all the available social capital going to the unashamedly narcissistic one-dimensional gym bros...)
Even if it is unvirtuous (by your own moral code, of course - not interested in mine or any anybody else's right now!) I do think that it doesn't really hurt anybody and there are probably other areas of life where you could do far more good (for yourself and for others) if you had the motivation and confidence that comes with feeling attractive - maybe, even if you're right about the virtue, this simply isn't a battle worth fighting?
Is building a (small) gym at home - possibly centred around just a few square feet of jigsaw matting, a single bench, a rack of free-weights, and maybe a spin bike (along with a small industrial fan and online HIIT classes or whatever via a cheap Android tablet) - an option for you? In addition to not feeling the 'impostor syndrome' you seem to be getting in the public gym, you would also get to train without having to drive anywhere, would be in sole charge of the gym's playlist, would probably save money in the long-term, would be resilient to COVID-style gym closures, etc. etc.
Alternatively, you could maybe take up something that's not-quite a competitive sport but also not-quite a hobby? Indoor bouldering, Crossfit, fitness clubs/classes, martial arts, etc. These things might provide a more friendly/motivating social atmosphere to train in?
I don't think being more attractive will increase my quality of life at all. It definitely would if I was actually ugly, but I'm well-groomed, not fat, have a reasonably symmetric face, am a charismatic public speaker, and come off as intelligent and thoughtful one-on-one (I have no idea why, but it's consistently true). I don't think that having better muscle definition or whatever will actually get me any more social capital (although I'd be interested in being proven wrong here, I haven't looked into it in detail).
You're right, of course, that harnessing unvirtuous motivations to achieve useful ends is a good idea, but the main thing that deliberately not-caring about attractiveness is doing for me is that I'm focusing on strength over hypertrophy. Wanting better cognition and lifespan is motivation enough for me, I think.
A home gym might be a good idea long-term, but as long as I'm going to university I have free access to a gym right next to my class, so no travel time or money spent. (It also seems to be a personal demon that I compare myself to others when it wouldn't be beneficial, and I am fully capable of imagining idealized people to compare myself to if I'm alone.) I hadn't thought of the playlist, and getting some earbuds is a good idea.
Unless you are already in the extreme tail of lean, muscular, and attractive, it is likely that you are massively underestimating the personal social capital that you would gain from pursuing weight training, nutrition, and aesthetic improvements. Basic social strategies like simple flattery are incredibly effective when they come from an attractive person. Spend some time around an extremely attractive man (or especially woman) and you will see that the world literally opens itself up to serve them in nearly everything they pursue.
Of course, this has the potential to be abused. It is massively abused by people who have it, but the same can be said of money or anything else.
Yes, this is a really good point...which I was trying to make, but I failed to emphasize... Absolutely some desires are bad, and being natural, products of evolution doesn't of course doesn't necessarily mean we should invariably regard it as good. And I think we should all aspire to be virtuous. But didn't Aristotle say virtue is optimal point between opposite vices? The desire to be more attractive, as long as it isn't taken to too much of an extreme and is appropriately balanced with other concerns etc. is not only pretty benign; it can be a great source of motivation to do things that you should do anyway, for other benefits that you should want...but maybe don't have as easy time feeling a strong visceral desire for, like to have good bone density and reduced changes of Alzheimer's etc when you're old. So you might as well leverage the desire to be attractive as source of motivation to do what you should do for other reasons (and also to be more attractive, which is in itself beneficial and IMO legitimate to want).
Soreness aka DOMS is typical when just starting out, but if you are persistently quite sore then you may be pushing too hard. Just follow a linear progression program like greyskull LP, the numbers will feel like they're flying up by virtue of showing up.
The fitter people than you all started where you are, and in the grand scheme it takes comparatively very little time to get in reasonable shape, let alone see improvement.
Avoid those cognitive distortions as pitfalls, they will ruin anything you try, be it lifting, running, sports, etc. That's number 1. If that aside you actually "hate" lifting, then don't. You don't need it to be healthy. If you want the benefits of anaerobic exercise you can just do bodyweight stuff at home every day, for like 20 min, 60, whatever.
Walking in itself provides benefits. For health optimization and longevity though, consider zone-2 cardio, i.e. get your heart-rate up to 65-70% of its max and sustain that for the duration of a workout. You won't feel like you're dying, and you'll be yielding all the benefits.
I looked up greyskull and found a whole lot of dead links. How does linear progression work? My current strategy has been very non-mechanistic: I increase weight by 2.5 kg whenever the last session didn't feel very hard. Numbers have indeed been increasing, though it sure doesn't feel like I'm "just showing up".
I know everyone starts somewhere. I know comparing myself to others is detrimental, and I know that their success doesn't diminish mine. This knowledge does not prevent the emotional reaction.
example: https://imgur.com/iLhAJBl . You really don't need to lift to failure to reach an intermediate level.
Progress is 2.5 lbs not kilos.
If the aforementioned distorted perspectives aren't informing the emotional reaction, then what is? Taking a cue from 3rd-wave CBT (like MCT), sometimes correcting distorted views is not sufficient. It's then efficacious to a) learn to let irrational thoughts pass without judgement, and b) divert your attention to to something else. You can do this with exercises (found in workbooks and therapy guidebooks) or setting a reminder to entrench the new habit.
Still consider alternatives to lifting you might find more enjoyable. Sports are generally fun, and there are plenty less taxing ones with lower barrier to entry such as badminton, dodgeball, pickleball and others. Good way to make friends as well.
If you're showing up, then you're succeeding and reaching arbitrary goals becomes just a question of (little) time. It's physics. Not unlike the case of those pursuing weight loss and finding they cannot lose it all overnight (nor should they restrict that aggressively), a single pound per week shed is winning, so long as you sustain it.
"I also have some desire for aesthetic / attractiveness improvements, but I don't consider that desire acceptable, and try not to let it affect my actions."
That seems really weird and silly to me. If you didn't have the desire, I'd say, ok, to each their own. But you DO...and yet seem to have some neurotic hangups about it. It reminds me of like the idea of the Christian who flogs himself like the guy in the DaVinci Code for having sexual thoughts.
It's a completely normal desire produced by evolution because being attractive makes you more likely to succeed at survival and reproduction. And luckily what makes you more attractive generally also makes you healthier. Up to a point... I wouldn't recommend focus primarily on the benefits for your looks, as that that can lead to neuroticism and being overly self-critical, and can lead to unhealthy extremes...which is ironic, because it usually also fails at the goal of being more attractive. Anorexics look scary and gross...so too in a different way do roided-out bodybuilders (perhaps not coincidentally since many of them may be driven by a similar kind of body dysmorphia.)
And being attractive has all sorts of other practical benefits, because people (male and female, straight and gay) generally will generally like and respect you more and think more positive things about you in general. (Unless you're a woman, in which case other women will see you as more of a threat.)
It's understandable to resent this fact of life if you happen to be unattractive in a way that it's hard to do anything about. But if it bothers you that it's not "fair" or "rational"... Well...too bad humans are the wrong species for you. It's the way it is, and it would take genetic engineering to fix it. And I suspect it wouldn't be a good idea anyway. Facts of nature aren't "injustices". To think they are is a category error, and trying to address them as if they were is pointless and usually just creates greater harm than it is ostensibly attempting to prevent. I prescribe Thomas Sowell's, "A Quest for Cosmic Justice".
Anyway...sorry, that kind of thinking is a pet peeve of mine. And I think letting go of it is genuinely liberating. It's ok to be a human. I think it's pretty great actually. And even if it weren't it would be stupid and pointless and counterproductive to agonize over it.
But anyway... I'll see if I can help with the original question. If lifting weights is boring, I get that. You can try other things. I just got my own home barbell setup, but before that I was mostly doing calisthenics and mace/club. Heavy steel mace/club stuff is fun and functional. If you don't like being at the gym, you can work out at home. Especially if you're starting out and have trouble with motivation, I think it makes it a lot easier to buy a program that gives you a schedule and has follow-along videos. Like Tacfit Commando, which is great for all-around fitness, and can be done in less than 40 min a day, but there are no doubt countless such programs you can get now. Just pick one you like and stick with it for 3 months. Then maybe try a different one. Having follow along videos helps with motivation. Decide what time you will do it, then just hit play, and then follow along.
Among desires, the desire to be attractive is a minor virtue. It's sure not a major one, and if taken for one counts as a vice, but in appropriate moderation it's a virtue.
Even if you are only considering other people, most people would rather look at attractive people than at unattractive ones. (That might be a reasonable definition of "attractive".) So by being attractive you are improving the lives of those who can see you.
The only part I don't 100% relate to is the part about seeing fitter people bothering you, possibly just because a lot of fat people went to my gym. There were people with godlike physiques too, but enough normal/plus sized people that I didn't feel like I stood out.
I didn't exactly find lifting painful, but it was intensely uncomfortable and really boring. It only got more uncomfortable as the weights I was lifting increased. Going to the gym made my day worse, as did the thought of "urgh, I have to go to the gym later"...
A long life where I go to the gym three times per week is less appealing than a shorter life where I don't have to go to the gym.
I kept at it, on and off, for a couple years before I decided I wasn't ever going to get to the point where I started liking it. Tried going with friends, tried different regimes, it still sucked. I quit and I sort of wish I'd done it sooner. On balance it was probably worth sticking with it for as long as I did was probably the right choice, because it sounds like a lot of people do eventually come to enjoy it. I didn't turn out to be one of those people.
I tried a few alternatives - some attempts at martial arts, didn't really take to that. I got a pair of dumbbells to use at home and a pull-up bar. I use those a bit (thanks Beeminder), and I find it much less taxing to do little-and-often: just one or two sets of an exercise at a time, several times through the day. The dumbbells are basically my fidget toy, or that's the idea anyway. I work from home, which helps fit that sort of thing in.
After deciding martial arts wasn't for me and saying "I'd rather be in a mosh pit", I had a plan to actually commit to that - start going to local gigs explicitly just to get sweaty in a pit on a regular basis. That remains a daydream.
The main thing I ended up replacing the gym with is bouldering, which I see has already been recommended. Honestly I should have just skipped straight to bouldering, because I've heard so many people tell similar stories which usually end with them getting into bouldering.
I have no idea to what extent I'll get the the cognitive / longevity benefits, given that I'm not doing intense full body lifts anymore, but now it's a fun hobby instead of a slog. It's certainly a strength workout, but very unevenly distributed (I climb until my forearms stop working - it's always my forearms which go first).
I did not read that, but I'm familiar with most of its arguments. And yeah, I'm going to the gym at a university, so everyone there is not only fitter and prettier than me, but probably smarter too. I know it doesn't matter, that they started somewhere, that comparison of self to others is not beneficial, etc, etc... tell that to my status-seeking lizard brain.
I've tried rock climbing things a few time and disliked it far more than my current regime, I think because it requires a lot more flexibility than I have. (I've been stretching between sets in a desultory sort of way.)
You're getting swamped with answers, but I didn't see anyone say this, so I'm adding one more.
Consider HIIT instead of lifting - high intensity interval training. If you have to force yourself to do some physical activity every day (or every few days), it is by far the "highest benefit per unit time" physical activity. When you're starting out, it only takes 12 minutes, and you might build to 20-30 minutes total when you're more fit, and that's usually it. You also don't even need a gym for it, although you can of course do it in the gym (or a pool, or anywhere you can run, etc).
Here's a big list of the main benefits:
1. Improves fat burning efficiency and burns twice as much fat as traditional cardio.
3. Improves VO2max, and drives better blood oxygenation.
4. Drives greater stroke volume, and greater cardiac contractibility, ~10-15% more than regular cardio.
5. It drives vascular adaptation, making your heart chambers larger and more elastic, improves the size and elasticity of your arteries, and increases the number of capillaries.
6. It drives hypertrophy - the relevant muscles get bigger.
7. It allows you to recruit more muscle fibers, and to do so more efficiently, driving greater muscular force and contractibility.
8. It improves insulin uptake, and improves the muscles’ ability to transport glucose overall.
9. It increases mitochondrial production and turnover, leaving you with more and “stronger” mitochondria.
10. Relative to traditional moderate-intensity cardio training, it drove a 41% increase in pain tolerance, and a 110% increase in race-intensity time before dropout in one study.
The first pass HIIT I recommend to most people, which only requires shoes and a watch or phone, is: "12 min total, consisting of eight 30s all-out-sprints (4 min) and 1 min rest intervals between them (8 min)."
Overall there's a lot of different HIIT styles you can do (Tabata, REHIT, and more), but any one of them generally drives the benefits.
I used to do the 7 min workout religiously every morning! Ha, man, those are some memories. Yeah, I think it's pretty good. Certainly beats doing nothing, and good for getting the heart rate up and working each muscle group. I think it's not quite as good as full HIIT sessions, but it does work a greater variety of muscles, and is great to do overall.
Many years ago I would do the 7 min workout every morning. Then I graduated to 5bx, a similar 11 min HIIT-style workout the Canadian military invented and popularized in the 50's. I worked up to routines 5 and 6 on that. Nowadays I do Wim Hof breathing and a bunch of breath hold pushups every morning, because I do triathlon training and I spend about 2 hours a week doing HIIT in the course of 80/20 periodization training, so don't really need the extra load.
To your point, I agree, I think there's a lot of benefit from doing some regular physical activity, even if it's only 7 min or 10 min at a time - every bit counts, and having a "floor" of physical capability is a good thing too. Bodies work on "use it or lose it," after all. When your body knows that it will have demands on each muscle and on your cardiorespiratory system every day, it keeps itself tuned up and in better condition on all those fronts.
Also find bike rides a nice way to get some cardio. You're outside with fresh air and sun, even just 1 hour (ex. 15km with a 200m drop) once a week gives you a nice exertion.
Really? I thought training to failure (or close to it) was important. Then again, I do usually feel 90% recovered within two days -- just in time to do it again.
Training to failure occasionally is probably still a good idea so you know what "2 reps left in the tank" actually feels like, but since you hate going to the gym, maybe easing up a little bit (and seemingly gaining the same benefits) would make it more tolerable
Zvi summarizes and discusses it in his Feb. 5 post.
":The baseline scenario as AI becomes AGI becomes ASI (artificial superintelligence), if nothing more dramatic goes wrong first and even we successfully ‘solve alignment’ of AI to a given user and developer, is the ‘gradual’ disempowerment of humanity by AIs, as we voluntarily grant them more and more power in a vicious cycle, after which AIs control the future and an ever-increasing share of its real resources. It is unlikely that humans survive it for long.
This gradual disempowerment is far from the only way things could go horribly wrong. There are various other ways things could go horribly wrong earlier, faster and more dramatically, especially if we indeed fail at alignment of ASI on the first try.
Gradual disempowerment it still is a major part of the problem, including in worlds that would otherwise have survived those other threats. And I don’t know of any good proposed solutions to this. All known options seem horrible, perhaps unthinkably so. This is especially true is the kind of anarchist who one rejects on principle any collective method by which humans might steer the future. . ."
So, guys, what do you think about this? (If you don’t want to read the article, Zvi’s summary is good, cuz he’s Zvi. (There’s a lot more to it than the bit I quoted.)&&
Thanks for the link. Must read that paper. I trawl through many ArXiV headers each week, but skip CS as there are just too many papers and most don't interest me much.
Once nearly all meaningful jobs are taken over by AI, one can only assume that relative merit will be judged mostly by social scoring, whereby commendable achievements will garner individuals merit points, like scouts completing courses. So if AI systems are allowed to supervise this activity, it/they will be perfectly placed to dangle various personalised incentives before every individual, and who is to say all these will all be benevolent?
Makes a lot more sense than the sci fi scenarios that are more commonly discussed here, with AI actively conspiring to take over. It checks out in terms of cultural evolution too, capabilities not exercised are quickly lost. It's the world of Wall-E.
If I remember correctly, Asimov wrote a short story late in life of how his three laws of robotics would impact on the long-term fate of humanity. Short version of the short story: To respect the laws, robots had across untold millennia gently and gradually turned humans into a docile species nearly unable to move, hence unable to harm anyone, or be harmed by anyone. Sounds like your gradual disempowerment-scenario.
I've been struggling through a backlog of Zviposts since the holidays (weird feeling to be like "I just wanna get home and read about o3" when customers are losing their shit over no more panetonne or wreaths for Christmas), but skipped ahead to that one due to the intriguing title...was a Worthy Read, one of the better non-roundup AI standalones. Same arguments Zvi's been making constantly, just now It's Official(tm) credential-wise and more accessible to normies. I find this ideative convergence pleasing. Many Such Cases of Rationalists making interesting, compelling, perhaps even correct arguments about the true state of things, but there's no penetration to those who can actually effect change...I keep wishing there were better "code switchers" with a skin in each game, who could translate The Good News to those who have ears to hear. (And likewise help us avoid common nerdsnipe typical-minding traps.)
competing with us for resources. That seems to assume that they will become self-interested, and have goals that may not be compatible
with ours. I have never understood why people are so convinced that once they get supersmart
they will have wishes and goals of their own. I think self-interest is a property of biological beings, rather than of smart beings. So it seems
to me that the ASI could continue to be obedient to us, and work only towards goals we set. I still think it would be very bad for our heads to have AIs take over everything because we are now the equivalent of its Down’s syndrome children, but not as bad, of course, as having AIs
I think this tends to get covered by the background reading material a) instrumental convergence, that having *any* goal X to optimize at all funges against other goal(s) Y, unless X = Y; and b) it's actually really hard to define goals in a clear "do what I mean, not what I say" way (Value Is Fragile and related posts). So given resource constraints like land, power, water, etc., there's clear tradeoffs that could be made between allocating those towards goal X progress, and...whatever humans are doing with them. Consider humanity an opportunity cost, I guess. Hopefully there's enough other guardrails set up that such trades are simply outside the possibility-space...but we keep getting evidence that models will try to efficiently pull-the-rope-sideways around various constraints unless explicitly forbidden, without much/any prompting, and I'm not too comfortable continuously playing whack-a-mole against alien minds that are a lot more patient (if not literally "more intelligent").
That's been my understanding, anyway. Not smart enough to understand much of the technical details, just the end results...
> I have never understood why people are so convinced that once they get supersmart they will have wishes and goals of their own.
They don't need to. Accomplishing any goal is aided by control of more resources and power. Even if not strictly necessary, it still minimizes potential threats and liabilities. The only thing that you can truly trust is yourself.
<Accomplishing any goal is aided by control of more resources and power. Even if not strictly necessary, it still minimizes potential threats and liabilities.
But accomplishing one of our goals in a way that harms us is basically the make-a-lotta-paperclips situation, a situation where we erred by not putting conditions and limits on the goal. For instance if we told our Ai to conquer Canada, and imposed no restrictions on how, it might go to work on that, then discover the canadians plus their AI are tough opponents, and decide to kill
all Canadians with poison gas. And since we hadn’t remembered to tell it to conquer Canada in a way that did not risk US lives, it wouldnt worry about gas blowing over the border, and a bunch of US citizens near the border would die.
So it seems like you already believe in instrumental convergence, but think that it's possible to stop it by giving it good goals. Basically saying do what I want but obviously don't paperclip us in the way.
I'd say this is actually harder than you think, and loads of AI labs are working really hard on it.
But even if it was easy, it wouldn't matter! Seriously, people are really dumb and are time and time again giving paperclipping goals to AIs.
See this [1] Zvi post for one of N examples. Even the crudest of "agency" frameworks tacked on top of a model already empirically show instrumental convergence.
The models are too dumb to do any harm today. But we're already seeing it happen.
>Even the crudest of "agency" frameworks tacked on top of a model already empirically show instrumental convergence.
How is this evidence of "instrumental convergence"? They specifically told the AI to edit a file until it ran without errors, so it edited that file. At one point it edited the file in an unexpected way, and another time it just made a dumb mistake that led to wasting lots of storage space without any benefit, but at no point did it show any sign of developing some new sub-goals besides those explicitly given. All that's confirmed here is that demand for evidence of instrumental convergence outstrips supply.
And how easy do you think it is to just "put limits on the goal"? The entire damn field of AI Alignment was created just to figure that out, and after more than two decades, they got nothing. What hope do you have to chain it, when they are far more creative and intelligent than you are?
And that's before we consider stuff like value drift and just simple malfunction. Of course, all of this is going to be irrelevant as soon as some moron purposefully makes one of these things sentient, in the name of "progress".
>And how easy do you think it is to just "put limits on the goal"? The entire damn field of AI Alignment was created just to figure that out, and after more than two decades, they got nothing
Says who? It seems to me that they've gotten plenty. In fact, the field of teaching AIs ethics has proven to be so easy in reality that those who want to continue dooming have been forced to come up with tortured rationalizations for why the unexpected ease of placing moral limits on AIs is actually just further proof of how dangerous they are (see Scott's posts on Claude's "incorrigibility").
It's true that whatever progress has been made to date has come entirely from people actually building and testing AIs in reality, not from high-IQ neurotics inventing elaborate apocalypse stories based entirely on some hand-waves Nick Bostrom made up (which is what "the field of AI alignment" usually refers to), but that doesn't seem particularly relevant.
If they were really that creative and intelligent, they would have realized the entire field was ludicrous to begin with. They can't even forcibly "align" a single human being much stupider than them, how could they possibly think under those circumstances that they'd be able to do it with an alien supergenius? Anyone not an idiot in at least some way bails right there.
LLMs are smarter than humans on many tasks and are nowhere near being sentient. The idea that sentience is an emergent property of LLMs is clearly untrue.
More from Argentina: the free hormones are out. The doctor said that in the new system he just physically can't get them, there's a shortage. I am entitled to them legally but there just isn't enough. He gave me the prescription to buy it myself instead, and it's expensive enough that I'm considering smuggling them from other countries.
I've found a strange disconnect in rat circles - on one hand, you should stop reading the news, keep your identity small and be less engaged with politics. On the other, you should constantly question your beliefs and biases, seek out information on them, and bet on your beliefs on Polymarket (that is the wildest one to me).
I see it not as two conflicting beliefs, but rather as an *alternative*. Either don't read the news, or do read them but also bet on Polymarket.
It's the reading *without* betting that is bad. On one hand, you believe that it makes you know more about the world, on the other hand, you refuse to empirically test this hypothesis... because deep inside you know that many things you believe are probably wrong. So you are wasting your time obtaining information that you suspect it mostly wrong. Why?
Questioning literally all of your beliefs is impractically tedious; you need to prioritize. Mostly, you should be questioning the beliefs that are important to you in your actual life.
Which is to say, mostly *not* beliefs about things you only read about in the news. If you nonetheless do find yourself caring deeply about such things, yeah, question those beliefs. Including the one that says "this is important to me and I should hold strong beliefs about it".
Those don't seem opposed to me. I do try to question by beliefs and biases and seek out information about them, though I've only ever done one bet on Polymarket. At the same time, I try to tune out most sources of news and I don't engage deeply with politics.
For the news in particular, I think it's compatible to tune most of it out because it's generally a very low-quality source of information that's designed to keep your attention rather than inform you. There's a hierarchy here, of course: cable news and social media news is the absolute Pits, most online papers are in the same place, while a handful of legacy publications have the resources to do novel and in-depth investigations which are worthwhile to read (though this is not most of their content).
There are certainly a lot of things going on in the world which deserve thought and attention, but I generally don't need to know about them *right now*. The sooner to an event the news comes, the less likely it is to be accurate or complete, and the more likely that it's just trying to get attention on a breaking topic which might not even be relevant in a week. With the exception of things which are personally or professionally important for me and my loved ones in particular, which would prompt me to some action and which I would learn about through other channels anyway, I don't see a reason to make that tradeoff.
I keep my identity small and try to hold my beliefs loosely unless I've gone to enough effort to really research them myself (cue every stereotype about the person who 'does their own research'), which requires finding and reading the best arguments from the various sides. I don't do this too often outside of my professional life or hobbies, because it's a lot of work. This has delivered me a good deal more peace and clarity than when I was politically engaged and dogmatic.
This are obviously related in my view. Reading the news means you think you gain information, betting on news proves you don't. Questioning beliefs and biases are much easier when you don't have ideological or personal identity tied up in them, and For the most part "engaging with politics" is more about acting in impassioned ignorance rather than learning bloodless facts like "the Jones act cripples American trade".
The way I think of it is most "news" is meant to get you angry or serve as entertainment. CNN and fox want people to watch for as long as possible, websites want clicks, etc. The best way to do that is to be inflammatory not to have a sober analysis of events with numbers and statistics. But if by news you mean tarrifs are now expected to rise and here are thoughtful commentaries from 3 different economists with arguments both for and against, that's different. In this age of information you have to be discerning and also aware of when you are being emotionally manipulated. I try to stay informed and that does mean having some exposure to news, but I try to tilt towards sources that have actual information and aren't just libs/trump/fill in the blank bad.
I think that if one considers the news a source of noise/entertainment rather than a source of relevant information, the contradiction evaporates. E.g. there is no reason that I should care about tariffs between the US and Canada, as they will not measurably affect my life in any way, and even if they did, there's little I could do about it, but because I wasted my time reading a newspaper, I know about them and have a strong but utterly useless opinion on them.
I tend to believe that the busier a street is, the lower the chance of finding any low-hanging fruit along its edges. Politics is a very busy street.
An examination of missed opportunities shows I've personally gained around $80 from following politics over decades, because I gambled on a Trump win in 2016 when he was at longer betting odds than Nate Silver. On the other hand, I've lost out hugely when it comes to economic opportunities, like buying the dip in 2020, or a weird niche little alternative to online barter marketplaces called bitcoin that I consistently refused to buy.
I reckon most people would be similar. It seems very hard to make money or cause good in the world using politics unless you do it at a level that's far more in the weeds than reading the news, in a more hands-on way that involves either running for a minor office, helping someone else do so, or becoming a lobbyist. This impression comes from personal experience of seeing how lobbyists and grass roots politicians work and isn't rigorous, but I'm not sure what rigour would mean here.
Bitcoin still isn't underpinned by an asset and is thus still a scam. While it's true that you could have made a lot of money, you would have done so by scamming other people out of it, which is maximally unvirtuous. My friends told me about Bitcoin back when you could buy something on the order of 20 coins for five bucks; I didn't buy in then and still haven't because it was and is an obvious fraud. Defrauding people is wrong, even (especially!) if you can multiply your stake by a million.
(And don't think the persistence of Bitcoin is proof of its legitimacy; Madoff kept a simple ponzi scheme running for 35 years. He was the chairman of Nasdaq!)
Buying an asset that's propped up by nothing but A) speculation and B) debatable utility as a medium of exchange is not actual, legal fraud.
I'm not sure about you, but I think I operate on a sort of "utilitarianism-but-with-caveats" morality when it comes to tradeoffs, where the caveats are things like "don't break the law or the unwritten rules of decency unless they're unjust" and "don't do anything you know is wrong even if the maths looks good". SBF obviously broke the law, arguably broke the rules of decency, and if I'd been in his position I'd have known it was wrong.
Buying BTC, leaving the other suckers holding the bag, and spending it all on charitable causes doesn't strike me as fraudulent or outright wrong. BTC being propped up by speculation is public information. I and the hypothetical bag holders have access to the same sources. It's trading consensual economic harm against non-consensual malaria (or worms, or lack of access to fresh water, or lack of access to small business loans, or whatever).
By this point I think everyone knows bitcoins aren't creating positive value commensurate to their price. So everyone buying them is knowingly investing in something that is at least partially a Ponzi scheme. If everyone knows what's happening, it's not really a fraud and by taking their money you don't do anything to them that they aren't attempting to do to others. Seems mildly unethical at worst.
I hope you're right just for the avoidance of total innocents being scammed, but I'm far from certain that's so, unfortunately. But even if you were right, I don't think that's a moral get-out-of-jail-free card. Supposedly, many frauds and cons operate on the principle of making the mark do something illegal or at least sufficiently unethical that they don't want to admit doing it, because it keeps the mark from going to the cops. By the same token I think swindling marks by tricking them into believing that they're the smart guy who's going to swindle somebody else is less permissible and more of a strat, really.
With reading the news, we have a choice whether or not to want to understand the political and economic world we live in. If we don't want to understand these things, that's fine - it's just as acceptable for rationalists to have no opinion about the Ukraine conflict or US tariffs as it is to have no opinion on the Sapir-Whorf hypothesis, or the many worlds interpretation of quantum physics.
If we want to understand these things for whatever reason, we should: 1) choose issues to care about selectively, based on actual importance rather than whether it's trending on X; 2) spend time reading deeply to genuinely understand the relevant issues, rather than wasting time on constantly updating news sources that provide no additional value; 3) have very good epistemics that allow us to evaluate issues without our biases or identity getting in the way; 4) if we think that we have insights about these issues and choose to speak publicly or allow them to influence our behaviour, we should hold ourselves to high epistemic standards (e.g. betting on beliefs).
Yeah, I don’t identify as a rationalist and have yet to see good arguments from them to become one. Why does it matter what I believe regarding AI, or the many worlds theory, or whether COVID was a lab leak or not? I should assume I’m not smart enough to have some refreshing take on it, or to solve any of these issues. Why not just enjoy my life and not worry about it? What’s the reason to become a rationalist, unless for your own pure enjoyment?
My personal solution is: I read lots of news, but try to limit myself to the object-level. No articles of the form "politician X finds idea Y good/bad", and certainly nothing of the form "politician X finds politician Y good/bad".
Instead, I try to read (actually lots of) articles of the form "idea X will probably have the following consequences". This is often mixed with judgments of the type "and this is good/bad", but it's usually rather easy to ignore those.
With a bit of training, it's pretty easy to tell the two types of articles apart just from headlines and possibly abstracts.
It's less "stop reading the news" and more "carefully budget how many epicycles you're spending on Bounded Distrust, since otherwise VOI can quickly go negative". People dunked on Scott a ton for his The Media Very Rarely Lies post, but I think the general point that you can usually extract basic facts without being in perpetual Constant Vigilance Mode With Straussian Cognitohazard Filter is accurate. Past that though...it's just a lot less work to hedge on epistemic uncertainty than to get anchored on a wrong thing, unlearn it later, and overrwrite with a correction. So I try not to read the news as "news" if I'm not willing to put in the work, or at best treat it as unreliably narrated fiction based on a true story.
(#NotAllNews, of course, some sources are more hostile than others...)
I guess that makes me feel old, if modern readers don't parse a lot of the older SSC/ACX/LW terms...but it's good to occasionally refresh one's citations.
*epicycles ~~~= brainpower spent per unit of time to complete a task
*VOI = Value Of Information, analogous to Return On Investment, but for facts. Everyone except maybe Tyler Cowen has finite memory and processing power to handle new information, don't get overloaded on low-value crap[1] or ephemeral trivia like play-by-play politics unless it's otherwise useful/entertaining. A great deal of "urgent news" is not, in fact, time-sensitive for the median audience.
*Straussian readings: https://en.wikipedia.org/wiki/Leo_Strauss#On_reading (tl;dr the actual intended message is often not literally what's written on the page, it's as much what goes unsaid as what gets said; rounding this off to "but what are the vibes?" is not wholly inaccurate)
*cognitohazard: sorry, been reading too much SCP, information that's harmful once known; sometimes after doing a close reading, one finds the intended message is literally the opposite of the literal meaning (say, in a backhanded compliment), and such contortions can be distinctly painful if one is used to directly taking things at face value all the time, which is the typical Rationalist/autistic writing style
*hedging on epistemic uncertainty: I can't remember where the OG post is, it's not exactly Epistemic Learned Helplessness (https://slatestarcodex.com/2019/06/03/repost-epistemic-learned-helplessness/)...but tl;dr treat much of the news as malleable claims and interpretations layered on top of some base facts[2], rather than a definitive conclusion of Here's Why This Happened And/Or You Should Care. Be mindful that corrections and retractions to such Official Narratives are neither common nor widely advertised; lots of "common knowledge" is based on such early-but-wrong reporting that too few people bothered to revisit later once debunked/contextualized/whatever
Makes sense, I never read the actual post and just picked up the term from cultural osmosis. But I guess it usually is mentioned in the context that adding increasing complexity to prop up a theory is generally Not Great, Bob. The More You Know(tm).
"the news" is one of the worse things that exists; instead you could consume media from a different culture, read long technical but political books, talk to people who disagree(tho that usually unnecessarily hard), or even if you want to be "up to date" you could read primary sources(like Ive skim 80% of evective orders and have unique take I havnt seen anywhere else)
Sarah Chen's computer displayed two windows side by side: in one, their latest VEGF expression data scrolled past in neat columns. In the other, a five-year-old paper declaring cold fusion's latest resurrection—and subsequent debunking. Her mouse hovered between them as the basement lab's fluorescent lights hummed overhead.
Their latest results were beautiful. Too beautiful. Just like cold fusion had been. Just like quantum consciousness and the EmDrive and every other revolutionary discovery that had dissolved under scrutiny's harsh light.
The centrifuge wound down with a soft whir, and Maya Rodriguez extracted another set of samples. Her latex gloves were marked with numbered squares—her own system for tracking which sample went where, redundant to their digital logs but independent of them. "Point-three-eight," she announced, holding the vial up to the light. "Third trial matches the first two." She logged the result in her notebook, then again in the computer, then photographed both with her phone. "Triple-blind documentation, just like we planned."
Sarah minimized the cold fusion paper and pulled up their protocol document. Every variable accounted for, every potential artifact identified and controlled. They'd designed their methodology specifically to catch their own mistakes, their own biases. The kind of rigor that had been missing from all those other revolutionary claims.
"Found something interesting in the control group data," Maya said, rolling her stool over. Her worn sneakers squeaked against the concrete floor. She pulled up another window, revealing a graph. "See this baseline variability? It's way too consistent. In real biological systems, you'd expect more noise. We're actually seeing too much precision."
Sarah leaned forward, tension uncurling in her chest. This was what they needed—not just positive results, but meaningful negative ones. Real data had texture. Patterns. Even its imperfections told a story.
The basement door's hinges creaked, and James Morrison ducked through, bringing a blast of Chicago winter with him. He brushed snow from his shoulders, his movements precise despite the cold. The messenger bag slung across his chest bore a faded university logo, relic of his former life in traditional academia.
"Just got out of the video call with Berkeley," he said, pulling out his tablet. "They want to know why we haven't published yet." His breath was still visible in the lab's air. "I told them about your triple-blind protocol. About the control group anomalies we're investigating." A smile creased his face. "They said it was 'refreshingly thorough.'"
Sarah's fingers found the worn edge of her lab notebook, where she'd taped a quote from her grad school mentor: Extraordinary claims demand extraordinary evidence. Below it, in her own writing: And ordinary claims demand just as much.
"The University of Chicago team sent over their latest validation criteria," James continued, pulling up another document. "They're offering full access to their BSL-3 facility. No strings attached." He paused. "Well, one string. They want us to present at their symposium next month."
Maya's hand tightened on her vial rack. "They'll expect preliminary results."
"They'll expect honest ones," Sarah corrected, turning back to their data. The VEGF numbers stared back at her, defying everything she thought she'd known about protein expression. But unlike cold fusion, unlike the EmDrive, these results came wrapped in layers of verification. Each number earned through methodological rigor that would have made her grad school mentor proud.
The ventilation system shuddered to life, stirring loose papers on their bench. Among them, a printed email from the grant committee: Forty million in new funding for breakthrough technologies. The word "breakthrough" had always bothered her—science wasn't about breaking through. It was about building up, slowly, carefully, each discovery earned one verified data point at a time.
Maya was already preparing the next batch of samples, her labeled gloves moving in practiced patterns. The centrifuge hummed back to life. On the whiteboard behind her, their experimental framework spread like a map, each step marked with both a procedure and its verification method. Red marks highlighted potential failure points—not to avoid them, but to face them head-on.
"You know what's different about us?" James asked quietly, studying their protocol document. "We're not trying to prove we're right. We're trying to prove what's true."
Define p(fascism) as the odds that some time in the next century, a significant majority of historians internationally will apply the label "fascist" or something roughly semantically equivalent to the current US administration. What number would you put on that?
My inside view is that the odds are quite high- maybe up to 30%. The man seems unwilling to share power with Congress, the courts or the bureaucracy this time- preferring to set policy with some occasionally pretty clearly unconstitutional EOs, while signalling that he'll ignore court injunctions. His nationalist rhetoric has also turned disturbingly expansionist recently, he's acted with a great deal of casual cruelty toward non-US citizens in these first few weeks, and I don't see him stepping down in 2029 without a very ugly fight. Also, his former chief of staff thinks the term applies.
That said, from an outside view, I notice that it's been pretty common historically for people to mistake relatively innocuous US presidents of the other party for potential dictators. I'm not sure I understand how that mistake occurs well enough to be confident that I'm falling into the same trap. So, I think I'd have to put the final number at something more like 10%.
That is still a worryingly high chance of a pretty nightmarish scenario, however. High enough that I feel like I ought to be putting together a serious emigration plan. Does that seem premature to people here?
"High enough that I feel like I ought to be putting together a serious emigration plan."
I'm rather biased because I put together and executed such a plan five years ago, and have not once regretted it. But I would say no, it doesn't seem even slightly premature. Having contingency plans for high-impact but somewhat-low-likelihood scenarios is generally wise, provided you're realistic about the probabilities and limit your investment in any one plan accordingly. It helps if a single plan can cover multiple contingencies, and "emigrate from the U.S." certainly covers a few others.
With all that said, unless you are a particularly fortunate sort of person, emigration is usually a slow and expensive: you'd need to commit a fair amount of money and effort well in advance of your actual moving date. So for such a plan to be any good, you'd need to have somewhere to go that you'd be reasonably likely to want to move to anyway. Which is to say, somewhere you'll like well enough that you'd be willing to start the emigration process when if things in the U.S. become merely *bad* but not yet *intolerable*. Along with picking a destination, it might help to try to define some sorts of events that would serve as signs to start packing: if you wait until you're Sure Beyond a Reasonable Doubt, you'll probably have waited too long. The best time to flee from a disaster is before it happens. The second best time is "before everyone else has the same idea and the exits get crowded."
On that note I'll add that (as outlined above) an emigration plan is only useful for slow disasters. Having something more in the nature of an escape plan--a plan to get away from danger if it happens suddenly and with little warning--is worth considering. I'm not talking about wacky secret-helicopter-hidden-in-your-garage Hollywood style hijinks. Mostly sensible things like knowing where your travel documents are and keeping them easily accessible, having an idea of where to go and how to get there in case of an emergency, keeping a big of cash on hand and maybe some basic emergency supplies, that kind of thing. Pretty big overlap with basic disaster preparedness. As I said above, keep the costs proportional to the (still somewhat low) probability, but also update as necessary. I'm certainly not expecting the skies to crack open and weep blood, but the sociopolitical climate in the U.S. doesn't look likely to get any *less* turbulent in the next few years, and there's plenty of run-of-the-mill dangers that come along with that.
In fact, it now occurs to me that (depending on where you live) the biggest short-term risk increase could easily be natural disasters. Particularly "natural disasters under a government that's unlikely to have much interest in or capability of handling them well." Fascism or no fascism, there will always be more hurricanes and wildfires, and I expect to see a lot more of citizens left to fend for themselves.
Why should I care about how future generations of historians will redefine "fascism"? I admit to some concern about what the current US administration actually *does*; I care not a whit about what term is invented or repurposed to describe it.
If that's not what you meant to ask, perhaps rephrase the question.
More fundamentally, definition debates about fascism or communism seem a lot less fruitful than discussing what is or might happen and what the consequences will be.
Almost zero. It's the politics of the time. At the time Romney and McCain were both called racist bigots and worse. Does anyone today think that is a fair and objective characterization today? Romney was accused of wanting to put black people "back in chains". It's ridiculous. Give anything a couple decades and then you'll be able to actually think rationally and fairly about it in context without all the emotion being involved.
The passive voice (“were called”) is doing a lot of work here. I’m sure that somebody, somewhere has expressed pretty much every opinion imaginable.
I’m guessing that the “back in chains” quote comes from a Biden campaign rally. I think that Biden was saying that Romney and the Republicans don’t care about the middle class. I’ll provide a link so anyone who cares can decide for themselves.
These guys say they care about the middle class. You know my dad had another saying [...], “Charley, don’t tell me what you value. Show me your budget and I will tell you what you value.” [...]
But let’s take a look, because we’ve got a real clear picture, as Tom pointed out. We’ve got a real clear picture of what they all value. They’ve said it, every Republican voted for it. Look at what they value, and look at their budget, and what they’re proposing.
Romney wants to let the--he said, in the first 100 days he’s going to let the big banks once again write their own rules. “Unchain Wall Street.” They’re going to put you all back in chains.
He said he’s going to do nothing about stopping the practice of outsourcing. Republicans even voted down a proposal to say, “OK, We’re not going to give a tax break, as we do now, to companies who unbolt the machinery on the factory floor and send it to Singapore” [...]. “We’re not going to do that any more; we’re going to give a tax break to any company that unbolts their factory stuff in Singapore and brings it back to Danville.”
It’s not going to change the world, but guess what? They voted against it. Romney opposes it.
This is an argument that Proves Too Much. Noting that all politicians have been called bad names doesn't actually prevent some politicians from doing bad things. People always talk about "crying wolf," apparently without remembering how that story ends.
The trick is most historians are probably more left of center and more likely to be critical of trump and we also need to come to a common definition of the word fascism. For me that word describes a one party state with a dictator at its head, militarist, interested in expanding through use of force, with no free speech or basic political or judicial process rights (govt can do whatever it wants with you) and some basic race or ethnicity based superiority policies. I'm very concerned with Trumps presidency so far but I would put probability of what I described above a lot lower, like 1 or 2%. I think what is far more likely (30%) is a watered down democracy where the executive branch has even more power than it does now and can reliably and predictably ignore courts and Congress when it wants to and make policy on its own. I see this as a very bad outcome but still not quite worthy of the F word.
I have been thinking about the "relatively innocuous" label and believe that is over-applied based on:
(1) survivorship bias: the US political-economic system has operated continuously for nearly 250 years, so Americans think nothing truly bad happened that was a real threat to the system.
(2) failure to recognize significant change: arguably, the system that exists today is not the same one from the founding (or some other reference time, pick your favorite.) This maybe a ship of Theseus question, but I do think there's a massive naivety about how different the past was from today.
Another issue that I'm including here, though maybe should be a separate item, is the confusion that big mistakes need to have immediate consequences. For example, was Bush 2 innocuous? The country has clearly survived since his administration, so he seems to have been fine. On the other hand, we had a major step down in civil liberties (sacrificed in various ways for the war on terror) and plunged itself on an extremely dangerous (to the system) fiscal path.
(3) nostalgia: looking back and only seeing/remembering the good parts
(4) selective focus: looking back and only paying attention to the groups for which those times were good
As an amusing aside: at one point, a common piece of coinage issued by the U.S. Mint had the symbol of the Roman 'fasces' on it.
The 'fasces', which inspired the name of Fascism in Italian politics in the 20th Century, was a bundle of sticks tied together with an axe-head emerging near the top. The image was meant to show that a large number of weak sticks, when bound together, could become a single strong unit.
Anyway, the U.S. Treasury issued dimes carrying that symbol from 1916 to 1945.
It could be argued that issuing coinage with the symbol of the fasces on it was the closest that the United States ever came to some form of fascism.
However, there are other things that were called signs of fascism:
1. At one point, a President threatened to use his political skills to convince Congress to change the size of the Supreme Court in a way that would let him 'pack' it with friendly justices. It is possible that this proposal, discussed in the news at the time, might have been the reason that parts of the New Deal program declared by that President got Supreme Court approval.
Does a President threatening to pack the Supreme Court, as well as creating Federal rules and boards that impact all sorts of economic behavior, count as a step towards fascism?
2. At another point, a President used the language of declaring War to describe an effort to reduce poverty in the United States. He spear-headed an incredibly large increase in scope and spending power of the U.S. government. Is misuse of the emotional language around declaring 'War' on something a sign of fascism? The effort of ending poverty never really reached an end point that is called success. But this effort did increase the need for government income, via either taxes or borrowing. This effort also brought government officials into direct contact with the ordinary lives of many individuals.
Do ballooning budgets, huge social programs, and increasing the reach of government officials into private lives count as a step towards fascism?
3. Yet another President convinced Congress to spend significant amounts of money in reducing crime in the country. It didn't come with 'War on Crime' verbage, but the policy proposals did come with funding and mandates to massively increase the number of law-enforcement officials in the U.S. This effort was paired with laws signed by the same President restricting certain kinds of firearms from purchase in the U.S., and with laws forcing some sort of government background-check on anyone trying to purchase firearms from a licensed dealer.
A huge increase in total Police presence in the country: is that a step towards fascism? Do fascists like their subject populations armed, or disarmed?
4. Finally, another President got his signature program through Congress, only to have Congress taken over by his political opposition during the next election cycle. He proudly proclaimed that he could still get things done, because he had a pen and a phone.
The President declares that he can issue Executive Orders to get things done, when Congress is going to slow. Is that a step towards fascism?
Can you name any of the Presidents I've described, and their signature accomplishments? Can you give a fair argument against the claim that these actions were steps towards fascism?
I don't think that these Presidents brought fascism into place. They may have been doing things that I disapprove of.
I do bring these things up as comparison: it's easy to call a politician fascist, especially if you don't like their policies.
The fasces is still part of the seal of the Spanish Civil Guard (crossed with a sword). But Spain actually was a fascist dictatorship until 1974, and it didn't quite have the kind of reckoning with it that other countries did, so I guess they just didn't bother changing the iconography. Spain is by no means absent left-activism, so I always did wonder how that iconography never got controversial enough to be changed.
They're not the only police force that uses the fasces in its iconography. The fasces was originally (as in, in Rome) carried by lictors as a symbol of the power of imperium held by the official they guarded, and from there through a pretty intuitive pathway simply became associated with policing. I don't know the timeline for sure, but at a guess, this happened a good 60-80 years before the rise of Fascism.
I’m not interested in the higher-level point, but I want to push back on your framing for at least 1), 2) and 3).
1) Packing the courts is not sterling behavior, but hardly (uniquely) fascist. Given for how much time the system has existed, and how judicial review works in the US, it’s hard to think that the system as it is (or was in FDR’s time) is a bug.
A quick check suggests that the New Deal laws were voted by Congress and that they were likely pretty popular (since New Deal-supporting majorities in Congress grew in 34 and 36).
Then, when FDR suggested packing the courts (ie submitted to Congress the “Judiciary Reorganization Bill”, which never reached a vote), a bipartisan anti-New Deal majority took over Congress. I wonder why they did not roll back the existing legislation?
2) > Do ballooning budgets, huge social programs, and increasing the reach of government officials into private lives count as a step towards fascism?
Ballooning budgets and huge social programs do not count as a step towards fascism by any reasonable definition.
“Increasing the reach of government officials into private lives” is debatable but only because you frame it this way.
There’s a meaningful difference between a collectivity that decides that people suffering a life-changing accident need some measure of helps, and delegates officials to check that you are, in fact, as disabled as you claim to be, and a political officer checking your library for forbidden books and putting you in jail if they find some of them.
3) Actually, what happens in a fascist regime is that you can have weapons all you want – but the fascist thugs won’t be prosecuted if they use theirs while you will be. Or you and your family be administratively harassed. Or you’ll be gunned down (in a dark alley or in broad daylight) and no one will object because they’re afraid of being next.
Given the rest of the civilized world disagrees with the US on the merits of widely available weapons, it may be worth writing this off as an outlier. Especially since the murder rate in the US as significantly higher than in, say, Western European countries (which, just to forestall the objection, is not due to any specific race). Maybe Americans do need more policing to keep the crime rate low.
(I don’t know which president you’re talking about though.)
> Do fascists like their subject populations armed, or disarmed?
Gun ownership rules from Weimar Germany were pretty strict; Nazi Party loosened them for members of the Party. The Nazis were the ones wielding violence before then. Is the answer to illegal violence more of the same?
Honestly, <1%. Plenty of right-wing authoritarians and populists and dictators throughout history aren't considered fascist. Even Imperial Japan and Franco's Spain get arguments about whether they should be labeled fascist. Unless Trump decides to adjust his ideology to the point of being identifiably, straightforwardly fascist, he's extremely unlikely to be considered fascist by history.
We should confine the terms fascist to the early 1900 political movement that proudly used the term to define their philosophy, and other modern groups that specifically reference them. Certainly historians already do that, if journalists don’t.
I don't follow US politics or what's going on with Trump very closely but I'm interested in WW2/20th century history and to me 30% seems far too high. "Fascism" has become a broad word, maybe in the future it could get broadened even further to include Trump. But if you're asking whether an intellectually honest historian would compare him to Hitler/Mussolini/Franco I think there's almost no chance. 20th century european fascism came about in the context of countries facing multiple existential-level social problems like world wars, the potential for genocides, mass poverty, ideological conflict way beyond what we have now. Those conditions were far more extreme than the current US.
Fascist was already broadly used term of abuse before the Nazis took power. In his essay "What is Fascism" [1] (written in 1944) Orwell notes the apparent meaninglessness of the term and lists various examples. The Soviet Union and her Communist followers (which include the original Antifa) had already coined "social fascism" during the Weimar Republic, arguing that the Social Democrats were the true fascists, worse than the Nazis.
The chance of a coherent coversation or analysis using the word "fascism" or any similar word is absolutely zero. By 2010 or so it had been used for too long to mean "something I don't like" that it was already close to zero. As soon as mainstream wokeness appeared and significant numbers began to be capable of calling misgendering a form of "oppression" *with a straight face* it dropped right to stone-cold zero and it will never recover. Just see that trainwreck ACOUP article for an example.
If you want to talk about the actually worrying aspects of Trump, in a way that is comprehensible to anyone who doesn't already share your political views on everything, then completely taboo "fascism" (along with any other word that's often used to include completely mainstream conservative positions, like enforcing immigration laws and opposing the transgender movement) and replace it with something extremely concrete and clearly-defined. Like "will stay in power beyond his term" or "will outright defy courts or congress or seek to shut them down". Then, and only then, will a coherent discussion be possible.
"Just see that trainwreck ACOUP article for an example."
This seems like an utterly ludicrous example, given that the article was VERY scrupulous about defining the term as thoroughly as possible AND pointing out the many places in which it did NOT apply (including Trump in his first term).
Like, I agree that people will fling around extreme, pejorative political terminology at the drop of a hat, and take no care for the meaning. But carefully grounding a term in the original meaning AND carefully pointing the limits around it seems much more like pushing back against the trend than buying into it (even if the analysis were flawed, which you haven't actually argued by any means other than invective). One suspects that some simply found the analysis uncomfortable and decided to group it in with things they'd already resolved to dismiss.
> Define p(fascism) as the odds that some time in the next century, a significant majority of historians internationally will apply the label "fascist" or something roughly semantically equivalent to the current US administration. What number would you put on that?
Id label america fascist since fdr; while america will never be like germany; this is still a subjective question.
Your trying to be objective about fascism without stating a definition, dont; id speculate *the inverse* of what your trying to ask; if the future gets more free then more people will call the current political system fascistic, if it gets less free, more people will can it democratic.
"whats the p(democarcy) a historian in best korea calls the next year democracy in 50 years"; it would be the inverse of if best korea collapses
While Scott has written that using made up numbers is great, I would disagree. This is especially so when you're trying to do computations with those numbers. This also applies when your numbers have large error bars on them (which in the limit becomes making up numbers).
You pulled the 30% number out of thin air, and just because it feels plausible doesn't make it so. Rationalists love to use poorly founded estimates to compute expected values; see https://blog.givewell.org/2011/08/18/why-we-cant-take-expected-value-estimates-literally-even-when-theyre-unbiased/ for a critique of this in a non-Trump context. If you assume you have a 1/3 chance of being right and there's a 30% chance of Trump being a dictator, and then decide that means you should act as though Trump has a 10% chance of being a dictator, you are doing an expected value computation based on a poorly founded estimate.
I often hear the claim that AI professionals are more likely to have a higher p(doom) than other people and that's a good reason to worry about p(doom). But the premise doesn't strike me as likely. If most AI professionals (I mean people directly working at the frontier AI companies) believe p(doom) is reasonably high, then these people are knowingly working on a doomsday machine. I find that implausible. Reason dictates that it's much more likely that the majority of those working at frontier AI companies believe p(doom) is very low.
Perhaps it seems otherwise because a vocal minority of dissenters who are more likely to work on the alignment side for these companies do have high p(dooms), and because of their vocality we falsely believe that they are representative of the industry.
Simple tragedy of the commons. 90% of the time you’re working on the greatest innovation since the invention of writing, not to mention all the moneys. Plus, the overwhelming feeling in these circles is that the doom is happening whether or not they contribute, so why not drive a Ferrari before the world ends?
Atomic Scientists famously had (have?) a higher p(doom) than laypeople (https://en.wikipedia.org/wiki/Doomsday_Clock) despite Atomic Science being the poster-discipline, as it were, for building doomsday devices. I can think of two ways to look at this:
1. Maybe the field isn't a unified "they". It's made up of multiple types of people who all understand the doomsday device pretty well; some who are actively working on building it and others who are pretty worried about this.
2. Maybe the ones who are working on it are the same ones who are worried about it - but they can't stop working on it because Moloch compels them (https://slatestarcodex.com/2014/07/30/meditations-on-moloch) - something like, "even if we stop working on the doomsday device other people won't stop, and even though we're worried about doomsday devices *somebody* is gonna have build one soon, and we'd rather it were us than those bastards on the other side of the border - and in the mean time these film-star salaries we're all getting certainly do help to take the edge off our anxiety..."
The Bulletin of the Atomic Scientists is a non-profit campaign group rather than an association of actual Atomic Scientists. You can't really take what they say as the consensus of scientists in various nuclear fields.
Possibly it is today, yes [though even then I would be fairly surprised to discover that the Bulletin of Atomic Scientists doesn't get any input from any actual Atomic Scientists for things like setting its clock..] but what we're actually interested in is the Bulletin of Atomic Scientists in the 1940s, when Atomic Science was more comparable in maturity to AI (and when their clock was produced). Back then, I'm led to understand, it was founded and staffed by Manhattan Project scientists.
Secondly, even disregaring Atomic Science entirely, I still think my (1) and (2) above are still entirely plausible explanations even without a convenient prior case-study.
According to the summary, median respondents put 5% or more on advanced AI leading to human extinction or similar, and a third to a half of participants gave 10% or more. Subjectively, it seems like the worry has become more mainstream among AI people since the poll was taken, but I'm not sure if there's more recent data.
These don't seem like the same people I am talking about. I just pasted the description of the participants in that study into Claude, asking who they are and the reply is:
"Those conferences you mentioned are primarily academic venues, though researchers from major AI companies do publish there as well. Here's a breakdown:
The researchers are likely to be:
Primarily academic researchers from universities worldwide, especially from major research institutions with strong computer science/AI departments
A significant minority from industrial research labs at large tech companies like Google (especially Google Brain/DeepMind), Meta AI, Microsoft Research, IBM Research, etc.
A smaller portion from specialized AI companies like OpenAI, Anthropic, Cohere, etc.
It's worth noting that frontier AI companies like OpenAI and Anthropic have recently been publishing less of their research in these academic venues, often choosing to release their findings through their own channels or preprint servers like arXiv. This is partly due to the fast pace of development and competitive considerations."
So this survey was mostly of academics not the people actually working on the front lines trying to bring about AGI.
The difference between academics and people working on the front lines is much smaller than you seem to think. A lot of that is just the same people: the biggest group is of course PhD students and other academics who join the AI companies, but it also flows in the other direction. The staff of AI companies is academically very young, lots of them are fresh from university. Especially those who build the frontier models. And even beyond people, there a tons of research papers which are collaborations of employees of the AI companies with academics.
I also find it a bit weird that you mention arxiv as an indication that AI companies move away from academia. Arxiv is totally a product of academia. Even as a non-AI academia researcher, I rely on arxiv much more than on conferences/journals. All my papers are on arxiv, and all papers of my colleagues are on arxiv. In AI this tie is even stronger because the pace of the field is getting too fast for conferences.
But is that what you really think is happening? All these people working on AGI think the end of the world will come because of what they do but maybe they can slightly lower the odds if they get ahead of the competition? That doesn't sound like a plausible psychological disposition for a large group of people to have.
>think the end of the world will come because of what they do but maybe they can slightly lower the odds if they get ahead of the competition?
It would be interesting to poll nuclear weapon and ICBM designers from the height of the cold war and find out if they looked at the arms race in something like this way...
We know that if any of the Manhattan Project scientists had a high p(doom) over nuclear weapons within the next 75 years of the project, they were wrong. So if this is an analogous situation, we shouldn't trust the scientists' forecasts. If it is not an analogous situation, the comparison tells us nothing.
That's not how probabilities work. Unless their p(doom) was >90% (and, given the anecdotes of close calls, one might reasonably argue >95 or >99), then saying "they were wrong" doesn't make sense.
edit: and even in those cases, saying "they were wrong" wouldn't _technically_ be correct. But it would be a useful shorthand for the much more difficult to express way of assessing their claims post-facto
>We know that if any of the Manhattan Project scientists had a high p(doom) over nuclear weapons within the next 75 years of the project, they were wrong.
True! (EDIT: correcting to, as per discussion above,
>would have received a bad Brier score
)
I'm guessing that their views might shed some light on the psychology of living with significant p(doom) estimates from one's work's effects.
Is eye-rolling cultural or a universal facial expression?
What other facial expressions might be cultural or universal? Russians famously don't smile for strangers. What else?
Also, as a mono-linguistic English speaker, I've always wondered what exactly the meaning is of say "tu" vs "vous" in French or the equivalent in other languages. It seems so strange and foreign to me that a word like "you" could have such strong implications depending on the form. I have this theory that English used to be like that when "thou" was the second person singular but us cold English speakers retreated entirely into the formal "you" and dismissed the more intimate "thou" because the culture frowned on such intimacy. Might that be correct that "thou" was like "tu" but eventually no English speaker were comfortable with the more intimate form?
French has the notion of "lever les yeux aux ciel", literally "raise one's eyes toward the sky", which is basically eye-rolling. But it is not well-identified as *the symbol* of irritation like eye-rolling is in (part of) the Anglosphere. I have sometimes read things like "this public statement prompted lots of eye-rolling" in English-language media, where "eye-rolling" is a metaphor for people being irritated and in fact expressing that by other means (speech or text). The same metaphor with "lever les yeux aux ciel" would not work as well in French.
We don't smile at strangers that much either. Smiling is for when you want to communicate something nice.
The anglosphere has people who declare their own 3rd person pronouns and this has turned into a very divisive social marker. Surely you can understand how pronouns can have strong social implications.
Yes, that's exactly right. You can even see some similarities between Early Modern English (around Shakespeare's time) and modern German: nowadays we say "you have" (and Germans say "Sie haben"), but the informal form would have been "thou hast," whose (still-current) German equivalent is "du hast."
The T-V distinction's name comes from what were singular (tu) and plural (vos) in Latin, but there are similarities in many non-Indo-European languages.
The ironic thing about English is that now the archaic quality of "thou" makes it sound more polite, when it originally wasn't. The T forms are perhaps better thought of as being more intimate rather than more informal; off the top of my head the various Indo-European languages I'm familiar with tend to use the T forms for God (we see this in the Our Father with its uses of "thy").
In more recent times some Indo-European languages, such as Norwegian and Swedish, are getting rid of the more formal forms.
I consider the whole matter to be somewhat analogous to deciding whether to use first or last names when addressing someone.
Also, if you look up "french tu vous flowchart" you get some pretty funny (and informative) results.
Consider the difference in English of saying something like "please follow this gentleman" vs. "please follow him" vs. "please follow this fella". It's sort of like that. Obviously, the divisions and rules in English are different.
> what exactly the meaning is of say "tu" vs "vous" in French or the equivalent in other languages
I am not a French speaker, but I speak Russian. The same phenomenon ("ты" vs "вы") exists in Russian, and based on a quick check it seems to be exactly the same semantically (one form conveys more familiarity and less respect, and vice versa).
In Russian, this was a Western linguistical phenomenon - the "royal we", pluralis majestatis - that was grafted onto Russian language. Until Peter I (emperor in 1721-1725) there was one (UPD: SINGULAR) second-person pronoun "ты", and if the speaker was addressing someone far more powerful than them, they used titles, honorifics and other words signifying deference.
Peter was a big fan of westernized monarchy (he was a big Hollandophile specifically, IIRC) and tried to reform Russian monarchy, state and society along similar lines. In 1722 he published a "Table of ranks", which was a stratified list of positions in civil society, military service and church. It was his idea to prescribe that anyone addressing someone of a higher tank on the Table use the pluralis majestatis pronoun "вы" to address them.
Then, in 1917, when Bolsheviks formally abolished the Table of Ranks, together with the rest of the Imperial state apparatus, the usage of "pluralis majestatis" just folded into regular language and was - and still is - used in formal and semi-formal settings for just about everyone to convey a minimum of respect and/or dignity (e.g. a news anchor interviewing a scientist or an expert on something will use it, the cashier in a fast food restaurant taking an order from a client will use it - and it would be very bad form for the client not to use it back).
It is very slowly deprecating out of usage (among young people, people with progressive leanings, etc), but is still widely used. For example, I was in a few meetings at a tech company staffed with Russian speakers - varying ages from early-20 to late-40-year-olds, different ranks, seniorities, etc - and nobody used "вы" with anyone, but I imagine a similar meeting at the Russian ministry of something would have been very different. It is subjective, but I heard opinions that this represents a "democratization" or even "americanization" of the language.
The English version (thou/thee vs. ye/you) was also a foreignism - borrowed from French during the middle ages. French inherited it from Latin. German has it too, and I feel like they probably picked it up from French or Latin at some point. (Then later it became more complicated in German.)
In Vietnamese, the number of different words for “you” is enormous! It depends not only on the level of formality but if you are younger or older than the person you are addressing, or (if significantly older) if they are younger or older than your parents. If they are within your family then family relationship influences the word you use as well eg there is a difference between “grandmother” if they are the mother of your mother vs if they are the mother of your father. Other Southeast Asian languages are similar. In general, English is remarkably low complexity with just “I” and “you.”
No, you use the same phrases in third-person reference. His Majesty / His Grace etc.
For a case where a phrase like this actually became a pronoun, Spanish 'usted' is supposed to have come from 'vuestra merced' [your grace], which is why it has third-person verb agreement.
Oh, does Spanish do that? Italian Lei (second person singular formal, agrees with third-person singular verbs) has a similar origin - not grace, but another respectful virtue, so feminine. Leading then to the correct pronoun for excellence/dignity/serenity/whichever it was being she, capitalized to clarify you aren't actually talking about a third-person feminine noun.
Old Italian documents use voi, which shares the second-person formal singular and second-person plural roles as in French; in modern Italian, voi is just plural-informal you and the plural formal you is Loro. Do old Spanish documents also use a cognate of voi/vous?
The old form in Spanish is vos. You can take that quite a ways back, since it's also the Latin second-person plural pronoun. I was under the impression that vos is still used, as a plural, in parts of South America, but wiktionary suggests that Latin American vos is exclusively singular, so I may have misunderstood my informant. It would have been an easy mistake for me to make.
The current form in Spanish is vosotros, which has had another word combined into it. (The same thing happened to the first-person plural pronoun, nosotros.) The Latin pronoun, and modern vosotros, do not have singular senses, but Old Spanish vos seems to have had one. https://es.wiktionary.org/wiki/vos
As a rule, Latin abstract nouns are feminine. They don't have to be respectful - for example, cupiditas is feminine. So it doesn't surprise me that virtues would be feminine in Italian, but I am surprised that respectfulness might be a requirement.
Anyone know a good multivitamin like One A Day (has vitamins and minerals) without the Calcium? I like everything else in the One A Day.
Calcium in artificial form seems to increase risk of calcification in the arteries. That's the motivation to avoid it. Been looking for a while, with no luck.
I highly recommend ConsumerLab.com, the Consumer Reports of vitamins/supplements/etc. They independently test and then recommend products, and some of their discoveries are so alarming that I would NEVER purchase a supplement they haven't approved.
In addition to test results and best picks, they also offer Scott-esque "More than you wanted to know about..." articles on everything they test (major and minor studies, etc).
All that said, I consulted the multivitamin tests and their overall top pick for quality at the lowest price is Kirkland Signature (Costco) Daily Multi. They mention the vitamins cover 100% of most vitamins, while calcium is at a lower percentage because intake should come from diet. Calcium is at 200mg.
They also mention Naturelo One Daily Multivitamin for Women 50+ as a better vegan option, which incidentally only has calcium at 60mg.
There are some prenatal vitamins which don't contain any calcium, but a quick skim would seem to indicate there's a lot of tradeoff going that route, so they're not a great substitute.
There's a section on calcium supplements and heart attack and stroke, but it doesn't seem to be a concern unless someone is supplementing 1000 mg daily *and* absorbing calcium in their diet. So tentatively, I would think that a 60mg dose or even a 200mg dose in a multivitamin isn't anything to really worry about.
Would your doctor be bitchy about ordering calcium and vitamin D blood tests? It probably wouldn't hurt to get an idea of where you're at right now before you start reducing calcium.
Sorry to not be able to better answer your question, but the problem with multivitamins is that they don't give you enough good forms of the bigger metals (pill would be too large and expensive). For example, the magnesium oxide in it is the worst form (gives constipation and poor bioavailability for such a key mineral), and the B12 is the worse version too. I understand the motivation to only need 1 pill, but you should be taking fish oil anyway and a separate magnesium glycinate anyway, so is there that much hassle to taking targeted others like D + K2 + C, B complex, and maybe zinc?
I'm not sure what to take away from this, but I think it's interesting that both Musk and Trump, with basically infinite access to information, jump at Internet conspiracy theories:
I really wonder if this has anything to do with the vague claim that some US debt might be invalid (sorry for the mediocre source, don't have a better one handy):
and X thought 2016 was a rigged election by Russia to install their agent as President of the United States. Fill in X with your favorite politics/media figure from the other side. Having access to all the information you could want doesn't mean you cant believe an incorrect conspiracy theory.
Well, yeah, but if I were the President I might ask the NSA or the CIA or something about it before I brought it up to a foreign leader. And if I were Musk I would be looking at the massive data set I have from the Treasury department and not random X posts. Why not at least confirm these things based on your insider access before acting on them?
"omg, Trump is so stupid he believes the conspiracy theory that USAID was funnelling money to Politico. Now of course USAID was funneling money to Politico, but in a totally above board way that couldn't possibly result in corruption"
Their whole thing is horse-race, "how will this play with the voters?" coverage. That article basically says Democrats are morons for defending foreign aid, and only briefly mentions that there might be some principle at stake. This is the organization that supposed to be influenced by subscriptions from USAID (which was only one of many agencies subscribing)?
It depends on the number of subscriptions. Press secretaries buying copies of all the papers is fair enough, but if it's 100,000 subscriptions that's more like what Soviet embassies used to do with the Daily Worker. Give this is apparently more than half of Politico's revenue, it's on the scale to be the latter.
suggests that we’re talking around 10M$ per year from the federal government, and a very quick Google search (including the first three lines of a recent WSJ article accessible without subscription) suggests that Politico revenue is about 20 times that?
It could be, but it isn't. The US government's Politico Pro submissions are spread across way too many agencies to be effectively managed as a slush fund, and the amount coming through USAID specifically is too small to plausibly sway Politico's leadership.
Pray consider the possibility that you may have been bamboozled.
It could be, or it could not be. Perhaps the most powerful man in America could *find out which it is* before announcing that it's a money laundering scheme to fund left wing propaganda.
> I'm not sure what to take away from this, but I think it's interesting that both Musk and Trump, with basically infinite access to information, jump at Internet conspiracy theories:
its known that trump was lied to about troops in Syria, it could be simply rational
I'm not sure what claim you're referring to, but, how is believing Internet conspiracy theories a rational response to being lied to about an unrelated thing?
Interesting, but again I'm really not seeing how "a Trump administration official lied to the President" = "it's rational to believe Internet conspiracy theories".
How many times do you need to be lied to before you discard the information?
Suppose your dealing with a young earth creationist, your perfectly capable of providing sources of contradictions in the bible, but they insist the bible is true and god works in mysterious ways; several of your debates are pulling teeth where they cite the bible to discard something you believe; over and over again. The devil made dinosaur bones during the story of job cause it says the devil made every effort to mislead, the bible defines atheists as hating god. etc.
It would be irrational if this never ends; however, if theres a number that breaks down the wall, maybe 1000 errors in the bible; its *rational* in a very technical sense. Capability of reason.
I assume your aware "conspiracy theorists" use the term "scientism"; science in that sense is not the ideal theory but the system, university's, journals and to some degree newspapers and their intelligence friends. Do you have a number?
These numbers should probably be fairly similar, the number of times you think its tolerable to humor a young earth creationist(who risks being thrown out of community's for considering your questions) and the number of time you catch an major error in "the official trusted sources" of science, intelligence etc. before ... well quite lost.
---
I think lying about troop numbers in a war zone to the president is a very very big deal; the whole separation of powers thing says the president gets allot of militery power and a matter of life and death, how many americains died in iraq?
It should shake your faith in the system to your core; but Im under no delusion it will take only one "conspiracy theory", maybe make a tally?
1. Whether the anti-ISIS envoy lied about something is somewhat unrelated to the question of whether any other government officials lied about something. As the President, I'd be somewhat shocked if there's never A PERSON who lies to you about something, and that probably shouldn't make you assume everything you hear is false.
2. Why turn around and trust conspiracy theories? If your trust is lowered, that doesn't explain why you think random anonymous posts on 4chan or whatever are now true. It's a selective demand for rigor. It's the same thing as "sometimes medical studies are p-hacked so now I just do what my shaman tells me".
A lot (most?) of Republicans believe some things at this point that the mainstream would call conspiracy theories. To be fair though the mainstream did call “Covid was a lab leak” and “Biden has dementia” conspiracy theories and they both turned out to be true and no one has resigned or apologised. So it’s really hard to tell at this point where the truth lies.
It was also a conspiracy theory that twitter and other social media companies were shadow banning people. The reality of a secret government conspiracy to censor people across several platforms was much, much worse, only discovered when Elon bought twitter. Which is kind of insane, that a conspiracy with so many people involved managed to stay secret.
Reversed stupidity is not intelligence, and also, lab leak has not been proven, and most likely (according to domain experts, the Good Judgment Project's superforecasters, and the results of the Rootclaim debate) Covid-19 was zoonotic in origin.
When it comes to believing sources that becomes cyclic; if alice has evidence bob lies, eve asking why alice is paranoid about bob citing a story bob says is ripe for abuse and should.... bi-partate(?) your probality distribution
This would probably be a decent argument, if it wasn't for the fact that Trump claimed 2012 was rigged, the 2016 primaries were rigged (against him), and the 2016 general was rigged (against him), so the causality is nonexistent. He has seen and claimed baseless conspiracies long before anybody offended him.
I dont understand how this is trump specific; bobs rain of terror has been going on to long and alice is Awesome For America.
alice: <any negitive claim about bob>
eve: "bob claimed you were bi-polar"
alice: "... Im not, please treat any surprising claims from bob as cover for more evidence as bobs bahavor"
eve: "Reversed stupidity is not intelligence"
alice: that doesnt make sense here, you either can believe Im mentally ill, or bobs extremely manipulative, its either or, the middle doesn't make sense here
The cia slur "conspiracy theorist" is an accusation of mental illness, while being a universal cover for accusations; while inverse stupidity isnt predictive for say gambling systems its not that bad a heuristic for criminal justice, if you treat lies near murder scenes as honest mistakes your going to be very confused.
Sure, but if someone lies to you, and then covers it up and pretended like they didn’t, you’d do well not to trust them again.
The point re Covid is not so much that lab leak is proven as that it’s a serious possibility that serious organisations like the CIA are giving credence to. And yet for a solid year people were literally censored from social media and tarred by the press if they claimed to believe it. Huh?
"To be fair though the mainstream did call “Covid was a lab leak” and “Biden has dementia” conspiracy theories and they both turned out to be true and no one has resigned or apologised. So it’s really hard to tell at this point where the truth lies."
Well, there have at least been preemptive pardons ... Though that apparently means Fauci and other beneficiaries can no longer take the fifth in case someone wants to investigate further.
Trusted information is harder to come by, but I feel like that obliterates the distinctions between some of these claims, e.g. it's premature to say COVID was definitely a lab leak (or definitely not one), but it's bananas to suggest Hillary Clinton has her email server hidden in Ukraine.
As for the government funding partisan coverage - that has more of a veneer of plausibility to it (all administrations do at least SOMETHING to shape press coverage), although I feel like it goes off the rails to suggest that Politico, that most nihilistic cheerleader of any "savvy" politician, is the one in the bag for Democrats.
> but it's bananas to suggest Hillary Clinton has her email server hidden in Ukraine.
That seems like a bad game of Chinese Whispers downstream of the actual proposition.
The idea is not that Hillary Clinton's email server was physically in Ukraine. It was in her house. The idea is that Hillary Clinton's email server was insecure, which means that every country with a vaguely competent spy organisation has a copy.
It's what he said on the campaign trail at the time. He repeatedly said that it was insecure so that foreign adversaries, particularly the Russians, were reading our classified information.
This was reported in the press as Trump communicating to Russia to steal our classified documents. Which is a stupid or partisan (or both!) take on what he said.
He said repeatedly on the campaign trail that maybe the Russians had it after they hacked into a DNC email. But why Ukraine specifically? I feel like this is backing into a rational reason, but there WAS an Internet conspiracy theory that Clinton's email server was spirited away to Ukraine because of something involving CrowdStrike. And there's no example of Trump asking any other world leaders about the same thing.
I haven’t been following the lab leak story since Scott reviewed the Rootclaim debate. Has there been any new info since then, other than the CIA changing position?
There's an awful lot of information that was AFIK never considered in the Rootclaim debate in the first place, including but not limited to the independent assessments of half a dozen other intelligence agencies.
There is not enough information for anyone to have high confidence in their belief that the Covid pandemic did or did not come from a lab leak. But to treat the Rootclaim debate as dispositive, is folly - that process was too heavily dominated by the question of which party was the better debater, when both were chosen for reasons only weakly correlated with their debating skill.
I found the arguments fairly persuasive. Can you give an example of the evidence not considered in that debate, other than the intelligence agency assessments?
Also, it was clear that both estimates in the Rootclaim debate were so ridiculous as to make their analyses obviously wrong. From memory the wet-market guy estimated the chance of lab leak at 10^-20 or something, and if your answer is coming out that extreme, particularly on an issue where there is some expert disagreement, then there were mistakes made, either intentionally or not.
Yeah, they really needed to stop adding zeroes, as Scott put it, and everybody else needed to notice that all the zeroes were a clear sign that something significant was being missed.
There will never be more information about a cover up in a totalitarian society, who edited the photo with slatin and 3 others, down to just stalin? If we do know, was it from the collapse or was this something a Russian could've figured out? Either you experience a blunt fact of feeling lied to or you swept up by the changing tides and quick talking.
It’s an authoritarian society. Not totalitarian. And even then it’s relatively uncontrolled and the use of smart phones makes it unlike the USSR for instance. Information leaks out.
I’m not sure exactly what you are talking about here. What was edited and where? Was it edited in China or the US. Wasn’t it the US and the west that decried the lab theory as conspiratorial?
I think the proper respoence to a virus leak wouldve been burning the place down; but they likely did some amount of cleaning the day of that would wipe away most evidence, withholding a few samples here or there, and we wont get a definitive causal link.
> Wasn’t it the US and the west that decried the lab theory as conspiratorial?
The whole world did, america gave them money, but the final causal link will always be in china and we happen to have plenty of times chinas lied to itself and others.
> It’s an authoritarian society. Not totalitarian.
Is holodomor fake as well?
I dont care about definition debates in any way shape or form; but I loath apologism of genocidal states
Trump famously gets most of his news from TV, so although his access is theoretically larger than ours, his information diet is probably much smaller, although unique since I'm sure he does listen to the occasional security briefing. My pet theory would be that Elon has let his success and power go to his head, surrounded by sycophants, he ignores anything that doesn't conform to his biases and fulfil his pathological need to save the world.
I'm trying to find a quote from a quantum mechanics researcher. It was something like "there's more than 1 reality, but less than 1 reality per person". Does this ring a bell for anyone?
No idea why I can't find it with Google, AI, or my browser history... maybe I'm jumbling some different ideas together.
The closest I have found is David Deutsch in The Fabric of Reality - “Every time we perform a quantum measurement, the universe splits into multiple branches; yet as conscious beings we experience only one branch.”
Is there any kind of prediction market, or other forecasting service, that makes long-term forecasts of salaries in certain fields? I'm thinking something like a prediction market for "median entry-level salary of job title X in field Y in 20 years".
Such a forecasting service would be extremely useful to those planning their careers and figuring out what they want to specialize in. Especially now, where we expect that AI will cause changes in which skills/fields are lucrative and which are not valuable, more rapidly than any other point in human history.
Ex: 5 years, ago, majoring in computer science was an extremely lucrative choice, and everyone expected that to be the case long-term. Now with LLMs getting pretty good at programming, it's no longer certain that computer science is a good choice of skills to specialize in. Where can we find forecasts about changes like this, for the purpose of people figuring out what career they want? There is plenty of data about current job numbers and salaries from places like the BLS, Glassdoor, etc, but I'm not aware of any place to find long-term forecasts.
I'm writing this as someone seriously considering switching careers, from my current work which is mathematical/computational, to more hands-on scientific/experimental work.
I'm pretty sure that if AI deprecates software engineering, they will soon after rapidly deprecate every other type of work.
The reason we're tackling software engineering *first* is that it boosts AI development giving you a nice exponential growth cycle.
But software, even written by humans, was already eating the world. AI written software will just eat the world faster. All the jobs are going away and never coming back.
>I'm pretty sure that if AI deprecates software engineering, they will soon after rapidly deprecate every other type of work.
I strongly disagree with this. AI has the potential to deprecate most of the jobs that consist exclusively of computational/logical/mathematical work, but not work that involves touching and moving things in the physical world. Robotics is way, way harder than LLMs.
Even if the AIs automate the vast majority of work that can be done solely via interacting with a computer, they are still likely to be far inferior at physical, hands-on tasks like: resetting a broken bone, stitching together an open wound, hands-on engineering work that involves prototyping/putting together novel devices, mechanical/repair work. Even if we could in principle build specialized robots that could tackle some of those tasks, humans are very likely to have comparative advantage over robots for such tasks.
The reason that software engineering is likely to be automated first is not the reason you claimed. In fact, that reason doesn't even make sense: the limiting reagent for AI development is not how fast and how well the software engineers can code, but rather the amount of data and computing time available, none of which will be improved by more and better software engineers. And also, the kind of algorithms we've come up with for machine learning. That could in principle be improved with better AI, although it's certainly higher-level work than software engineering.
The reasons that software engineering is likely to be one of the first white-collar fields automated away by LLMs are: (1) software engineering consists exclusively of humans interacting with computers via text, nothing in the physical world; (2) we have huge amounts of training data available for this task, including some of humanity's most impressive software engineering achievements which are open source; (3) there is little institutional inertia against replacing the workforce of software engineers en masse; software company CEOs certainly wouldn't object to doing that. Even if LLMs could currently do the jobs of e.g. lawyers, accountants, HR, there is institutional inertia in those fields that prevents them from being replaced by LLMs en masse. Contra software engineering.
It seems like the biggest bottlenecks in robotics are mostly around software.[1] We have a lot of accumulated technology for making actuators and LOADS of technology for sensors. Connecting the two together in a way that consistently produces the results seems to be the major hurdle, and that's mostly software. Sufficiently advanced AI has 3 methods to cross that hurdle. It could:
1. Get good at writing traditional robotics software to control whatever systems it wants.
2. Train ML systems directly on various robotics problems
3. Be good enough at general intelligence to learn to control robotic systems directly.
Whether 3 happens probably depends a lot on how "general" general intelligence really is, and whether it can really cross barriers to systems that are way outside its training distribution. But some combination of 1 and 2 seem almost inevitable once you have computers that are good enough at writing the code. Reduce it to another programming problem (or another machine learning problem) and they get to bring all their inherent advantages to bear.
[1] To be fair, I'm not especially well-read about the field, so I'm somewhat low-confidence about what follows.
[1] could happen, but I don't expect the tasks that robots can be programmed to do via this method to be that far beyond the tasks that humans can program them to do now. If human-written traditional robotics software is limited to a small subset of what humans can do, then I expect LLM-written traditional robotics control software to still be limited to some subset of what humans can do, as long as LLMs are not vastly smarter than humans. If LLMs become not just somewhat more intelligent than humans, but many orders of magnitude more intelligent, then all bets are off.
I'm skeptical of [2] because it's difficult to collect anywhere near the amount of training data that we have for LLMs. For LLMs we have the entire internet. Leading LLMs are trained on trillions of tokens.
I'm also skeptical of [3] because the data that LLMs are trained on (text) is very different than motion and sensor data that is needed to control robots. I don't expect LLM intelligence to easily generalize to robotic control. And like I said above, it's difficult to collect the quantity of training data needed to train robot software to be intelligent at broad, general-purpose physical tasks.
So first and foremost, LLMs are currently the most visible and successful class of machine-learning models, not the only class. Talking about their limitations--especially regarding how they deal with training data--is very relevant for discussing [3], but only relevant to [2] if one assumes that exactly zero non-LLM approaches to machine learning in robots could prove fruitful. AFAIK current work (done by humans) along that line is not focused on LLMs, so I don't see why hypothetical AI-coders would suddenly switch to an LLM-only approach to that field.
Second, regarding the relative poverty of training data: I agree this is a non-trivial issue. I still don't think "LLMs only work because of massive amounts of training data, therefor robotic control systems will only work because of massive amounts of training data" is valid reasoning. In many cases the thing you want a robot to do is quite a bit simpler than what you want an LLM to do, and I think the "throw data at it until it works right" approach is far useful there. Even when one moves up to more complex environments, you have many more possible inputs, but they're still mostly governed by the same sets of rules. I expect that a minimal model for even a fairly general robotic system to contain MANY fewer parameters than a minimal model for a chatbot or AI assistant type program. Fewer parameters means less training data is needed to converge on an acceptable local minimum. It also seems like there's a pretty large cascade potential here: once enough ML based robotic systems are being used or tested, they're necessarily collecting sensor data that can be used for training[1], and quite a bit of it. LLM training data isn't cheap because it's easy to generate more, it's cheap because a staggeringly large amount of it already exists. Feedback data from various automated sensors has at least the *potential* to be actually quite cheap to generate once the infrastructure to generate and use it has been set up. (Going to reiterate that all of the above is still fairly low-confidence. Less "this is true" than "it doesn't seem obvious that any of this is false.")
Now, throwing humans into the mix brings back all the complexity that LLMs need to deal with, so I expect "robots interacting with humans type jobs are likely to be the last ones replaced. But *quite a lot* of blue-collar work of various sorts can be done without interacting with humans. Indeed, manufacturing robots are *already* quite common. So I think the potential to replace lots of the "jobs that involve touching and moving things" fairly quickly after the software jobs fall (if not before), but "jobs that specifically require interacting with humans" might be the big holdout. (Ironically one of the biggest focuses of robotics has been self-driving cars, which very much fall in the latter category: probably why so much investment in them has yielded so little return so far.)
[1] If anyone is tempted to analogize this to training LLMs on the output of other LLMs, I feel like it *should* be obvious why it isn't remotely the same. But I'll explain if necessary.
> Is it though? We've had robots building your microchips, cars, etc. for decades before LLMs even existed.
Let me clarify what I meant in that previous comment.
Robots learning to do hands-on, physical tasks by observing what humans do (like what LLMs do), instead of being explicitly programmed by humans (as is primarily the case for the examples you listed) is a very hard problem. Because we have much, much less training data available for those physical tasks. LLMs are trained on ~trillions of tokens. We can't collect that kind of data for general-purpose physical tasks.
This is why I am very skeptical that we will ever produce robots that are very broadly general-purpose useful at physical tasks. Not ones they've been explicitly programmed to do, but ones they've learned to be good at by observing what humans do a la LLMs. Robotics will surely continue to advance at tasks where we can specifically program them to do exactly what we want (as in the examples you listed).
> My meeting schedule disagrees!
Yes, software engineering, like all potentially-automatable white-collar jobs (e.g., law, accounting, HR), involves meetings. But the actual product that software engineering work produces that has value (code) can be produced exclusively by interacting with computers. LLMs are likely to be able to produce that valuable product, without all the meetings that humans require to get that work done, in software engineering and in other white-collar fields.
> "humans interacting with computers" also applies to stuff like drone pilots, and drones definitely impact the world physically.
Remote drone piloting may very well be automated. Although this is one of those fields that has, as I called it, bureaucratic inertia against automation.
I will add one more thing. Even if some of these hands-on, physical tasks I describe are potentially automatable by robots, I suspect that humans will have a comparative advantage over robots in these tasks, and thus will still be employed in these tasks. I .e., I suspect it won't make economic sense to try to automate all those physical tasks, because those resources could be used more efficiently to automate other tasks that are easier to automate.
Mostly agreed, with a probable delay from white collar jobs eaten by AGI to blue collar jobs that also need robotics advances. ( As in the previous Open Thread: If ASI actually follows quickly after AGI, then we get a _very_ weird intermediate period where ASI exists but doesn't have robot actuators - the year of the immobile god. )
Aerial drones are easy for AI pilots to manage because the airspace is an uncluttered environment. Conditions are much more complex at ground level, so expect a long delay between machines being able to dust crops and being able to fix a leaky pipe in your house.
During the "very_ weird intermediate period," I think you will see a boom in apps that instruct humans how to do tasks through augmented reality glasses.
Many Thanks! I do agree that anything that can be controlled remotely today can be AI fodder once AGI exists. But there are also lots of cases today where humans go into places where no actuator currently is, and which are hard for e.g. drones to get to. I don't expect ASI(s) to be immobile long, but there might be a few years of "bending tin" to make enough mobile robots to extend ASI(s) reach to all of the places humans get to today. We shall see!
If the service becomes popular enough, I wonder how it will deal with the fact that the forecasts it produces will influence many people’s career choices, affecting the resolution of the forecast.
I'm working to create a framework for a writing competition where the authors entering are also the judges. The aim is to fairly distribute the work of judging among the entrants, make a system that is difficult to game, and provide clear incentives for everyone to participate equally. The goal is a decentralised process that steers readers toward the best indie writing which is otherwise easily lost among the countless titles on online distributors.
Any suggestions on the best technical platforms to coordinate such a project would be appreciated. Ideally I would like a method that is offline, robust and easily reproduced by anyone who wants to run a similar competition of their own, so I was wondering if something like plain old excel linked to automatically send out emails might be viable, but open to any ideas.
my partner has previously participated in https://writingbattle.com/, which sounds similar to what you're proposing.
Part of their setup to keep things fair is, participants are assigned stories to judge from genres other than what they submitted, so there's no benefit to your own performance to unfairly judge others. And participation is large enough and anonymous enough that conspiring to trade favorable reviews would probably be very very difficult, especially to do so undetected.
it's a slightly different space than you propose, given that it's short stories written strictly during the contest, rather than pre-existing novels. but, hope it helps as some prior work.
Wonderful! This is exactly the kind of thing I need to study. I will be running a trial with flash fiction first on a small scale, but investigating similar existing models could save me some missteps.
Wish I'd been more involved in the EA stuff lately, used to be a regular at the meetings, but got busy and dropped off the map after everyone at the retreat got covid...
Morbid question: Is there any prediction market on whether Trump does a full Andrew Jackson, directly defying an explicit Supreme Court ruling, at some point in his term of office?
And it is just such disagreements that would make this market interesting! Anyone see one out there? ( I tried a Google search, but unsuccessfully - but the phrasing could be quite tricky, so I don't know if I misgoogled or if no such market has been created yet. )
Many Thanks! I think that you are right, but I don't think that it is vanishingly unlikely. It could be an interesting market even if the odds are, say, 25%, with ups and downs as events unfold.
Whats the best strategy for a below average guy on dating apps? Liking most profiles destroys your algorithm, but its not like being selective with your swipes increases your odds with the girls you like either(they still get hundreds of likes from other guys). For reference, I get 1-2 likes per month and have been on 2 dates after 6 months of being on a dating app(few more matches that went nowhere).
What makes you say you're below average? It's possible hang-ups aren't specific to using apps. If you're healthy and well-groomed, ought to be able to do alright given time.
Apps are a shit-show today I hear, but at the time I used them, you could directly message, and that's all I ever needed. No "likes" whatsoever. The key was a good, sane message that drew from their profile, honing in on certain interests particularly ones that might overlap with mine. Even with "likes" now in play, taking care to do this will go a ways.
You can cast a somewhat wider net by joining drop-in sports or meetups. If you're a student, then you're in the best possible position because you're tripping over prospects.
It's a numbers game, and while it's true women are hounded in high volume, they also have to sift through garbage. Low-effort conversation, and then past this point, "players", which means they get back on the app. The turnover rate is fast, just got to be in line. Stand out by appearing bright, having your shit together, not being desperate, and being kind.
No offense, but this comment has strong old man vibes, like saying "I don't know what the job market is like now, but back in my day you could just walk into any place of employment and get a job, so a firm handshake is all I ever needed, that's my take".
Key advice: make your profile divisive/controversial/distinct. Don’t try to optimize for the highest average level of attraction; optimize for the highest level of attraction for some women. You are some women’s type, so lean into it as much as possible.
I think the best strategy for a below average guy is to not get on the apps in the first place. Things are rough enough on there if you are good looking that it seems like a waste of time of you're not.
Join a board game group. Join a co-ed soccer league. Join a trivia group. It is actually extraordinarily easy to meet women if you get off the internet.
I recommend trying a different dating app where hot girls reach out to below average looking guys. LinkedIn is particularly good for this. 😜
More seriously, if you are a person of enough means that you can travel internationally without too much trouble then international dating actually does flip the script. I did it (eventually found wife this way) and it is extremely easy for a man to get responses/bites. The one big caveat is that you have to be *very* careful of scammers. There are a ton of scammers who do online international dating as their primary source of money, just convincing guys to send them money. These are often real girls in the country you are looking so you can't do a simple video call to find them out.
Assuming you can filter through the scammers and visa farmers, there are a lot of great options available outside of the US (which is where I assume you are at, please correct if wrong).
I'm happy to answer any questions I can regarding the MiniCircle work I did with Ichor, though there isn't much beyond what I wrote in linked comment. I'm also happy to share results as they come in, though we probably won't have anything interesting for many more months. The real results we are after are lifespan (previous studies with AAV delivered VEGF suggested lifespan increase which we are trying to reproduce), which won't arrive until after maybe a year or so.
1) Contra frequent claims in this space, plasmids are not safe! They're only safe if not delivered by a real delivery vector because they don't get taken up by cells and basically nothing happens. If they're delivered by a real delivery vector such as LNP they end up in the cytoplasm and activate cGAS-STING which is incredibly toxic and inflammatory. There is a reason Generation Bio hyped up their mostly-single-stranded "immune quiet DNA".
1a) Was the Ichor CRO experiment literally just naked plasmid? Not even minicircle claims that works I think, they use PEI which is its own brand of problematic. There are ways to get naked plasmid to work-ish using various devices but none are mentioned here.
1b) What was the dose of the Ichor experiment? You say that basically "ehh we can just scale it up" but without knowing the dose it's not totally clear that's true.
2) The Ichor folks should have done qPCR or RNA sequencing to confirm the presence of mRNA transcribed from the plasmid as an orthogonal confirmation that things worked.
3) There are two separate "1,000-fold" gauntlets to run. First the issue of the ng vs pg discrepancy that was pointed out by another commenter. Second is the issue that local injection in mice and humans are not all that different in terms of volume, but humans are 1,000-fold bigger than mice so local injections go ~1,000-fold less far! This is why you need systemic delivery to e.g. the liver if you want to scale from mice to humans for systemic administration
If MiniCircle uses PEI then we probably did as well since a goal here was to replicate their gene therapy delivery system (though delivering a different therapeutic). I'll ask Ichor and get back to you if I find that we didn't.
Dose given was 50uL of VEGF DNA (see note on page 2).
We did something different for the first 3 mice (plus 3 controls) because in that readout we got an aggregate of human and mouse VEGF, while in the latter ELISA readout we got only human VEGF as a readout. I reached out to them to remind me how the first readout was acquired. That being said, unless the ELISA we used was complete garbage (e.g., just spitting out random numbers) it seems unlikely to me that no expression occurred given that we saw some treatment mice with 0 (so the ELISA wasn't just reading mouse VEGF) and some mice with non-trivial levels (so the ELISA wasn't failing to see anything at all). I agree it is possible that the ELISA had a failure mode where it returns random numbers but since the mice are all essentially clones I am skeptical that it was reading something endogenous to the mice.
Previous AAV delivered VEGF in mice was suggestive of a significant lifespan increase by going from ~100pg/mL to ~200pg/mL in circulation. In the mice that showed a result we saw about 30pg/mL human VEGF in circulation, suggesting that we only need around 3x increase to get similar results. While humans are bigger, they can also handle larger injections, injections at multiple cites, and you can inject into fat easily (hard to do in mice, which is why we went into muscle). I'm not too worried about being able to increase by 3-6x during dose finding studies in mice, and I'm optimistic we can hit similar target circulation levels in humans should we get that far.
I agree with you that systemic would be much better and I am actively monitoring the gene therapy delivery space for better systemic delivery systems. The thing to beat is AAV, which is incredibly expensive and immunogenic. For circulating proteins, plasmids may be sufficient which is why I am currently pursuing this avenue, but they aren't a good solution for non-circulating proteins which are also of interest to me.
"50uL" is not a dose, it's a volume, the dose will be determined by the concentration of minicircle within that 50uL.
If they used PEI, be aware PEI is extremely obsolete as an in vivo transfection reagent due to toxicity. LNPs are much better as are PBAEs I believe (PBAEs are trickier to work with).
Increase in total VEGF might just be due to the injury from the injection. Inflammation and tissue repair will upregulate VEGF. There is a "control" but it doesn't say if it got a sham injection or what. In any case, you need a non-VEGF-encoding minicircle control in case the DNA is doing something that upregulates VEGF (specifically, cytosolic DNA massively increases IL-6 which stimulates VEGF release). Meanwhile the hVEGF might be right, though antibodies for ELISA detection can have cross-reactivity and ELISAs can be noisy so you'd need to run controls to be sure, and the controls aren't there at all.
I'm not too surprised ALT/AST don't respond to a local injection, the correct safety assay would have been a cytokine panel with serum collected after the injection.
The weights are in fact concerning contra Ichor's claim. There are signs of weight loss at 1-2 days, and if that signal is real, then you absolutely cannot just jack up the dose if you don't want dead mice.
You are correct that AAVs are immunogenic but the reason that people use them is looooong experience with nonviral delivery not working. So it would be really really weird if it happened to work here.
The best case you can make I suppose is that the effective dose of VEGF requires like 1,000-fold less delivery than the effective dose for say hemophilia which is why you need AAV for hemophilia but not for this? I mean, it's possible I guess.
Also please don't take VEGF if you're not already old, it'll be more likely to cause cancer than help you. Also PLEASE PLEASE PLEASE do not do systemic delivery of plasmid to humans with any kind of good delivery vector it will likely kill you. See eg recent Nature Biotech paper https://www.nature.com/articles/s41587-025-02556-5 for more on why you should not mess around with plasmids on the off chance they are actually doing stuff.
Thanks for the nature article! In other work I have funded unrelated to this project we really struggled to get DNA in LNPs working, they kept killing the mice. I haven't read it yet, but it sounds like the authors of this paper may have found a solution to that, which would be great news! In other positive news I recently had a talk with a startup that is doing viral-like delivery and think they can address the safety and cost concerns of AAVs.
I heard back from Ichor by the way and they reminded me of some details of the pilot that I had forgotten. The first 3 treatment mice on page 6 are the same 3 mice that were the 28 day mice on page 12. The page 6 results are from a survival bleed, and the ELISA we used required a certain amount of material so we had to dilute the samples to use it (1:5). The same ELISA was used for the page 12 results, but those were terminal bleeds so no dilution was required.
This obviously opens up a lot of questions like why did we see non-zero values for the control mice in the diluted samples when we saw 0 values for treatment mice in the non-diluted samples. One guess is that a pure serum interferes with the ELISA in some way. Also the 3 mice that were tested twice initially showed elevated levels of VEGF but then 28 days later they didn't, and we need to understand why that was. One of the tasks we have scheduled already is ELISA scouting. We want one that can work well with survival bleeds so we can monitor mouse VEGF levels throughout the study.
We do plan on exploring and better understanding all of this before proceeding with the full lifespan study. The thing that seems to be the primary disagreement from people is whether a low-cost study to get a quick "no" (but an unclear "yes") is worthwhile compared to doing a more expensive study that would give you more clear results on the first shot. This pilot's purpose was to just see if we got all zeros across the board, which is what many expected, and if so we would have abandoned the project. It successfully gave us a lot of questions that we can now pursue for answers.
So I would say my issue is less with the fact of the pilot, you're right that pilots are often good, hindsight is 20/20, etc; extremely defensible move to do a pilot. However the problem is that that Ichor did a bad job with the pilot. They injected an inflammatory material and then used an ELISA with obvious cross-reactivity concerns to a protein (mouse VEGF) that is induced by tissue damage and inflammation! (VEGF does have a very short serum half-life and the injury might be healed by a week after the injection, so this might not be a concern, but you absolutely have to do the control to check this.) Measuring ALT/AST instead of serum cytokines was the incorrect safety assay. Extracting RNA from the muscle and doing qPCR or sequencing cDNA would have confirmed transgene expression and been very cheap but they didn't do that.
The powerpoint was put together unprofessionally as well. A volume is not a dose. No delivery vector was specified. Controls were not specified (sham injection, minicircle injection of some other protein, untreated, or what). The observation of cross-reactivity and ELISA concerns were not mentioned not was the possible undiluted serum concern. This is all forgivable if they did the actual experiment well, doing science well and generating good reports are separate skills, but both good science and coherent scientific reports are somewhat reflective of like "having a clear organized mind" so to me it's a red flag. Especially a red flag to do such a shoddy job for a paying customer as opposed to an internal report.
Re "Also the 3 mice that were tested twice initially showed elevated levels of VEGF but then 28 days later they didn't, and we need to understand why that was." this is (assuming I'm wrong and you actually did get hVEGF transgene expression, which is possible) likely due to epigenetic silencing of the vector, or death of cells that were expressing hVEGF due to cGAS-STING, or possibly an immune response to hVEGF.
Anyway yes I'm not surprised the LNPs are killing the mice. The issue is not that LNPs are toxic but that the DNA cargo is toxic when delivered into cells, and LNPs are good at delivering cargo into cells. The fact that you don't see toxicity here is therefore a very strong signal that delivery is inefficient or the dose is too low, and all you are seeing is noise. Or alternatively maybe you did actually see toxicity here, since there was a bit of weight loss and cytokines were never measured, which is a good sign for genuine delivery and bad sign for translatability and the ability to scale the dose.
I just want to add one comment for anyone who may be reading this now or in the future. The slide deck provided by Ichor was already more than I asked for. I engaged with them regularly in voice calls and every so often they will bring a presentation aid (slide deck) with them to one of those calls. I was the one who thought it was a good idea to share those slides with the world (even though they were missing a lot of details, as you outline here).
I'm quite confident that if I were to ask for a more formal description of everything they did or even a research paper as output they would have done so (and likely successfully), but I had no desire to pay for that work since in the end, I'm looking to convince myself that a particular therapy works and only then worry about convincing the rest of the world that the therapy works. There was more information provided to me during our calls over time and they have always walked me through everything they are doing in as much detail as I desire.
I am a huge proponent of working in the open and sharing any data I have, in whatever form it happens to be in, rather than keeping things under wraps because I am worried that the data isn't presentable. I would hate to see people look down on Ichor because of the choices I made (like sharing a presentation aid on the internet).
I'm considering writing a piece about this, summarizing the discussion for those (like me) who aren't specialized in molecular biology. Including your work and registering everyone's hypotheses in expectation of your results in a year or two would be neat. Is there a way I can contact you for more extended questions over the coming weeks/months?
For the sake of disclosure, my interest in this is both that I have a personal interest in questions of scientific validity and replication, and that people in my network are involved in Prospera and they should have a way to keep abreast of this.
If the singularity (or something like it) happens, it seems natural that land would skyrocket in value (given that there's a finite amount of land on earth). This... doesn't seem like it's priced into land prices at all? There are lots of places in the US where land is ~$1k/acre. Does it make sense to just... buy some?
Yes, but with an important caveat: If your objective is to make the most money, then you shouldn't buy the acre of land now. The Singularity might not happen for 50 years, over which period your land might barely increase in value. You're better off investing your $1,000 in an index fund or some other good asset until the Singularity seems near. At that point, cash out and use your investment--which might be $10,000 by that point--to buy the acre of land. Even if the acre has appreciated in value to $6,000 by then, you still make more money overall.
Even after the Singularity, I can't imagine why humans would want to live in places like that, and you might also wait many years before AGIs ask to rent it from you for something.
As far as space goes, there's still plenty of land, and if demographics continue in their current trend, there's bound to be a whole lot more.
My personal opinion is that, yes, real-estate will become a coveted asset, but only highly desirable real estate. Also, what constitutes "highly desirable real estate" is bound to change as economic opportunity loses importance. I expanded a bit on it in a comment from a previous open thread: https://www.astralcodexten.com/p/open-thread-364/comment/85973533
Forget surface land, the REAL profits are going to be in reselling the Mantle and Core of Earth. Basically worthless now, with *astronomical* value and potential later. 🤑🤑🤑
How do you get it? Simple, it legally comes free with the surface land in certain countries like the UK.
For (b), yeah, the idea would be that a huge increase in general productivity should make all natural/scares resources more valuable. E.g. this $1k/acre land was probably worth a lot less (maybe free?) before the industrial revolution.
For (a), I guess I'm thinking we should plan for futures in which private property/money are still meaningful, since if they're not, then nothing we do now really matters (unless it affects the singularity itself).
> E.g. this $1k/acre land was probably worth a lot less (maybe free?) before the industrial revolution.
I don't see how that would work; the industrial revolution is the story of marginal land cratering in value over time. The more people live in cities, the less valuable land is.
The homestead act of 1862 (according to chatgpt) gave 160 acres of land in the west to any settler who lived there for 5 years. So the land was free, you could just go and take it. Also, the US increased frim 4 mil to 75 mil in population during the industrial revolution, with the same amount of land, so it's bound to have gotten more valuable
Being required to farm land for five years is not in fact a zero cost.
> Also, the US increased frim 4 mil to 75 mil in population during the industrial revolution, with the same amount of land, so it's bound to have gotten more valuable
Only if that's the only thing that happened during the industrial revolution, which it isn't.
> The homestead act of 1862 (according to chatgpt) gave...
This is exactly the kind of thing you wouldn't want to rely on ChatGPT for, but it is true that there was a substantially similar program. Your details might be correct.
I suppose it depends what directions the singularity takes us!
Obviously it's hard to think usefully about post-singularity scenarios, but I *guess* in a lot of the survivable ones, humans will still value things like hanging-out-with-other-humans (towns, cities, megalopolises, etc) as well as the aesthetics that have led us to want there to be large areas of undeveloped wilderness, forests, national parks etc.
Productivity alone isn't necessarily the only metric by which we value things in our current society, and I don't think that would necessarily change even if our overall productivity rises. It might be more strategic to buy e.g. shares in a silicon or cobalt mine than land that's not currently being productively used - although there's a very high chance I'm totally wrong about this, also!
In 2020 I bought some *extremely* unproductive (as in 'poisoned from surface mining') land out in the middle of nowhere, mostly to satisfy my own desire to be able to have a shack in an isolated space when I want to get away from other humans. I might be productive there in my shack writing or whatever, but I don't see any highly agentic AI's offering to buy me out any time soon :-)
Hmm, really? I get the sense that the total sum of the value of all american land would be higher now than before the IR. Granted this is mostly based on intuition. Why do you say it was more valuable then?
Agriculture was by far the biggest part of the economy for a long time, and land value was mostly land value for agriculture. Now agriculture is a much smaller part of the economy, and the value of (non-city) land compared to other things is down.
Man, that's it? I thought for sure we'd have a chance at whacking multiple coins if we ever got rid of one.
The penny in 1850 was worth about 30 cents in terms of buying power. It's beyond irrelevant today, along with nickels and dimes. I think we could eliminate ALL coins and be fine, but if you had to keep one, only keep quarters.
The board recently resigned en masse over issues with the founder, Michael Sparks—who wasn't even using his real name. Scott Alexander made a serious misjudgment backing this team. Just because an idea sounds great doesn't mean the team is—in fact, that's exactly the kind of failure mode that attracts grifters.
The post I replied to was recently deleted after I replied, but for the sake of preserving the discussion, here is what it said (without naming the user for privacy):
"The Far Out initiative is dead. The company relied on the idea that editing two genes could replicate Jo's condition, but her condition is far more complex and not easily replicated. Essentially, $50K wasted—unfortunate."
I'm sharing this to ensure the conversation remains intact and that context isn't lost. Otherwise, it may not be clear that my post refers to The Far Out Initiative (TFOI).
Well, probably not wasted because they tried and discovered (what I think most of us would have said) "oh hey, it's more complicated than our simple model!"
Sometimes there is a benefit to proving the bleedin' obvious. I hope experiments like this will make the people wanting to plough ahead with embryo selection for higher intelligence pause a bit, because here we go: it's not as easy as "just do some snip-snap with these genes and there you go!"
Definitely not wasted. A negative result is still a valid and important result. I'd be interested to read about it, but Google isn't showing anything/I'm looking for the wrong thing - objectivetruth, do you have a link or something?
That doesn't tell me very much at all, it's just a bunch of speculation. Your original post said "The company relied on the idea that editing two genes could replicate Jo's condition" and implied that they'd found it could not. I assumed that meant they'd actually done some kind of experiment beyond sequencing her genome (and the genome of this "Bob"), and was hoping for some actual data.
Did Trump's efforts to challenge the 2020 election outcome help or hinder him in 2024? That is, was his performance in the 2024 election better or worse than in an alternate universe where he didn't warn about cheating beforehand, claim it afterwards, call state officials, arrange alternative electors or hold a rally on January 6th?
There was a brief period after January 6th in which many people thought he'd disqualified himself; that centrists and some Republicans (politicians and voters) would reject him to signal that those actions weren't acceptable. My impression is that this happened but in low numbers. On the other hand, I imagine the part of his base who believe his claims would be energised by those actions, as in their view justice was quashed and January 6th protestors were jailed for standing up for a righteous cause. Is there any way to estimate which effect was stronger?
Are there any good free LLMs with a long context window? When I ask ChatGPT to proofread a blogpost, it just reads the beginning and then gives up.
I think Gemini has the longest context window of any of the current models, something like a million tokens?
How do people here model the following phenomenon:
https://x.com/TheRabbitHole84/status/1889558203207213339
-- the claim is that someone must be pumping in catchy words like "constitutional crisis" into the narrative stream.
How does one determine if this could be a genuine coincidence; after all there is a lot of media out there, and constitutional crisis is not exactly a rare word. Or, whether this is organic rather than cultivated virality?
Why should it be a coincidence? Journalists do read what others are writing; also, if there is a constitutional crisis brewing (such as a president telling his appointees to ignore the law), it's the journalists' job to call it that.
https://en.wikipedia.org/wiki/JournoList
Go read Nobody Special's answer. It makes all the good points about how this is an emergent result of uncoordinated activity without a "manipulative intent". It makes all the "nice and interesting points" -- the search engines distilling the attention of the public, the web being a malleable media, the iteration speed and cross-pollination driving convergence.
Unfortunately it's all well-meaning bunk. The real reason for CNN and Politico using the same term is simple: CNN and Politico senior editors are in the same chat room, at some point they brainstormed what term the headlines should use, and now they are using it.
Oh yeah, and they totally had "manipulative intent".
I can see that what you say is *a priori plausible*, but from there, how does one go to "probable"?
Edit: Other than appealing to purely subjective guesstimates of "odds". In other words, how can people with opposite readings of the issue talk to rather than past each other on this matter?
I think it's due to a variety of factors, but none of them conspiratorial or reflecting manipulative intent.
Historically, journalism did this as well. There are a variety of names one could give to Nixon's scandal breaking into and spying on the DNC, but ultimately coverage coalesced around "Watergate" rather than "White House Spying" or other alternatives. This happens for perfectly good non-conspiratorial reasons. News organizations wanted to sell papers, and once the public conversation consolidates around particular nomenclature, you want to fit that nomenclature so that your audience understands what you're selling. In the early days of Watergate, headlines likely varied. Once you're a month in, if all the other papers are festooned with headlines on the latest news on "Watergate," while your paper has updates on the "Liddy Affair," the public may not understand what you're selling and consequently their interest may drift to other sellers.
This trend continues into the internet era, where it is exacerbated by multiple factors. Not a media expert, so YMMV but just to list a few that seem most plausible to me:
(1) News is now elevated by algorithm and search engines. The average reader googling "Trump Constitutional Crisis" may not care much whether they get the latest updates from CNN or Politico. But Politico and CNN, respectively, care a lot which link the reader clicks, and consequently have a strong desire to best mold their article to whatever keywords people are searching.
(2) Add to #1 the fact that internet articles can be revised while live, in a way that old newspapers can't. That means that once algorithm emergent social consensus around a particular phrase emerges, you can go back and revise all your "DNC Break-In" articles to be "Watergate" articles. And because of the nature of the search engine driving clicks, which drive eyeballs and money, you have every incentive to do those revisions - you'd be crazy not to.
(3) High-turnover media and consequent cross citation. The need for a constant churn of new media means less and less resources & time to dedicate to truly new reporting. If I need new articles every 4 hours on the latest DOGE activities, there's no way I'm getting those out fast enough if I limit myself to solely my own independent research. So I’m much more likely to cite a CNN article that came out an hour before mine, and other publications coming out later will cite to me. Doing anything else results in my content coming too late, which no one can afford, so further ambient pressure in the system standardizes the descriptions of the coverage.
Thank you; very nice, interesting points.
It's the same reason that there were tons of headlines about the "superbowl" this weekend, and a ton of headlines about "tariffs" and "plane crashes" last month.
People talking about the news isn't exactly a conspiracy, and the president triggering or threatening to trigger a constitutional crisis is pretty fucking big news.
Musk is a hero, I don't understand why people hate him.
https://www.youtube.com/watch?v=RncK73kM_FM
Without doubt they need to get congress on board.
As someone who has hated Musk long before it was fashionable, let me outline the case briefly:
- Musk is a charlatan who brags about accomplishments that are not his (e.g., being the founder of Tesla). When he actually designs something, you get things like the cybertruck.
- he is a vaporware salesman who has built a massive house of cards on empty promises (robotaxis, "fixing traffic", the "hyperloop", going to Mars...)
- he treats his employees like shit, and doesn't give a rat's ass about laws, regulations and workplace safety.
If you need the long-form version of everything wrong with the man, consult the youtube channel "Common Sense Skeptic".
ETA: of course, Musk has recently done a lot of despicable stuff beyond what I wrote, but that's on the news anyway, so I don't have to repeat it.
Huh, OK one man's hero is another's villain. Can we agree to disagree?
Obviously I can't keep you from considering the man a hero, but I only see a few options:
- you have solid information that my claims above are wrong or least misleading. Then I'd like to hear them. I consider that unlikely, because this is all pretty well documented.
- you acknowledge the claims above, but judge that, compared to what remains of Musk's merits, they are irrelevent. Then you should think about what it means that you adore an amoral blowhard.
- you ignore the claims above and stick your fingers in your ears. Again, you should think about what that means for your judgement criteria.
- you acknowledge the claims and adore Musk BECAUSE he is an amoral blowhard. In that case, you are clearly part of the problem.
I didn't say I adored Musk. Let's not stick words in each others mouth. I said he is a hero. Is he a flawed hero? Of course all real heroes have their flaws. Is he an asshole sometimes, yes. Does wrong stuff, yes.. all the others. IDK, but I'll take your word for it. so take that as a yes. I've read a biography of Musk and listened to ~10+ hrs of him taking with Lex Fridman and Rogan. My impression (and this comes partially from believing Lex.) is that he's the same person in private as he is in public. I call that a straight shooter. He's telling you what he really thinks and I respect that. As far as business, he seems to have (so far) avoided having his businesses taken over by middle management. He does this by drilling down into everything, he seems somewhat unique in this regard. Now he's offered to focus this talent into our bureaucratic state. And I say, "Drill baby drill." (Sorry I just couldn't resist that last line. :^)
"My impression (and this comes partially from believing Lex.) is that he's the same person in private as he is in public. I call that a straight shooter. He's telling you what he really thinks and I respect that. "
That's the same line that people use about Trump. Problem is, they're both pathological liars who say whatever suits their needs at a given time.
Do you know the story about the solar rooftop presentation? Im 2016, Musk arranged for an unveil event for solar shingles at a movie set and said "The houses around you are all solar houses. Did you notice?" - and went on to present solar shingles in various designs and colors.
The houses around had nothing to do with solar, and the shingles weren't functional. Why did he do that? To convince Tesla shateholders to buy SolarCity, a failing company owned by his cousins. Does that sound like a straight shooter or like a conman?
So like... this is a thing that you could google, right? There are *at least* thousands of people online who have spelled out in some detail why they hate Elon Musk. If you don't understand, it's because you haven't tried to. Like, at all. And since you didn't phrase it as a question, but more of a boast, I guess you're proud of that?
But anyway, I'll pretend you want to understand, and tell you my thoughts.
For most people I know, buying twitter and allowing Nazi accounts to thrive (e.g. https://www.nbcnews.com/tech/social-media/x-twitter-elon-musk-nazi-extremist-white-nationalist-accounts-rcna145020) was the tipping point from "Elon is an annoying rich guy" to "Elon is a mortal enemy". And to anticipate the claim that he did it for "free speech" I would point out that he has suspended real journalists while allowing Nazis to remain, and he's limited speech he doesn't like (such as the terms "cis" and "cisgender"). He is not a principled advocate for free speech - he is a principled advocate for white nationalism. Also, not sure if you know this, but he did a Nazi salute on TV. Twice. And a lot of people hate Nazis.
Do you also need me to explain why people hate Nazis?
But again, that's just my personal experience with why people hate Musk. By all means, take to google or reddit and see what other people have to say. I mean, we both know you won't do that. But the information is there, and now you can't say no one has ever showed it to you.
https://www.reddit.com/r/DecodingTheGurus/comments/15xvove/what_are_valid_reasons_people_dislike_elon_musk/
Yeah no I do understand. I live with my son and he hates Musk too. And so let me make a stab at why I think people hate Musk, and it's really the same reason that people hate Bobby (Kennedy). And I can identify the reason, because I use to hate Bobby and Trump and others. And that is because I bought into what everyone else was saying about them. Everyone says these people are evil and going to do harm and are bad, and they tell you why, and the easy thing is to just believe them and say yes. I hate them too. The only way (I know) to break this trap, is to listen to what the people say themselves. This often involves (for me) listening to a several hour podcast (from maybe Lex or Rogan) and getting to know the person. (Or reading the books they have written.) Most people don't have the time or inclination to do that... so they just say what everyone around them is saying. Now perhaps I started out in a bit better position, because I've listened to Rogan for years, and I know most of the things people say about him are a distortion. So I was open to finding out that the main stream view of these other people, (Musk and Bobby) was perhaps a distortion. And I guess I'm just asking you to be open to the possibility that these people aren't evil.
About Nazis: I don't like Nazis, I don't want to listen to them, and yet in the name of free speech I think Nazis should be allowed to air their ideas. I don't know about Musk limiting speech about cis and cis gender, but clearly this is also not right and in my opinion a mistake on Musk's part.
Although, if you're super interested in what Elon Musk said, remember that time he said that Jews were promoting anti-white racism? And then when advertisers started to leave twitter because of all the Nazi stuff on twitter, do you remember Elon's response? He told the advertisers - not the Nazis - the advertisers, who are his company's source of income - "go fuck yourselves".
It sure seems like when Elon was promoting electric cars and talking about colonizing Mars, everyone was cool with him. He even got a shoutout on Star Trek Discovery, a show so woke they made Stacey Abrams President of Earth. And then it seems like when he started promoting Nazis, then people said "hey, wait, actually this guy sucks."
So I would say it's the exact opposite of your theory. People liked him just fine until they listened to him freely and publicly express his thoughts, at which point they were disgusted.
I mostly know Musk from his Lex Fridman podcasts. I pay no attention to twitter/ X.
I also don't engage directly on Twitter, but if you're choosing to ignore the inflammatory things that Musk makes a point of saying and amplifying on the platform he bought at enormous expense and risk to his entire business empire, I wonder why you would do that and instead ONLY pay attention to the things he says during laid-back interviews with sympathetic bros like Rogan and Friedman.
I mean, if a person helps a bunch of Nazis to spread Nazi ideology, and then gives them the Nazi salute twice on TV, and then you listen to that guy on a podcast and decide you like him, does that mean that you were mistaken before and he's actually a good guy, or that you are mistaken now and have been conned by a Nazi? Lots of people decided to listen to Hitler and then decided that he had a lot of good points. We all grew up wondering why people listened to Hitler, because in retrospect it's obvious that he was evil. But apparently, it's not always immediately obvious that evil people are evil, because apparently they have some kind of persuasive ability that convinces people to listen to them, despite what they do when given power.
So I guess the question is, do you think that a program of eliminating trans people, purging the country of immigrants, purging the government of anyone who isn't loyal to Trump, eliminating vaccinations and public health programs, eliminating support for democracy abroad, ethnic cleansing the Gaza strip to develop it as a seaside resort, and suddenly cutting off food and medicine to millions of people, including children, who were depending and counting on it is heroic? Or is that all Nazi stuff? If you have an answer to that, it really doesn't matter what Elon says when he's smoking pot with Joe Rogan.
Oh dear, I don't think Musk is a Nazi, you are free to have a different opinion. TBH I don't pay all that much attention to what Trump and Musk are doing. (I did watch that youtube video above... I fast forwarded through some of it.) I think 1/2 the time Trump and Musk are trolling the other side, and it would be much better if they were not such assholes. But there is not much I can do about that.
I don't think all the purging you are worried about is going to happen. I think we need to shrink the federal government, and borders need to be controlled along with immigration. Trans people will be fine, I know they are all very worried right now. I'm not sure but much of that is fear generated from the left. Support for democracy abroad is almost Orwellian in some respects, we need to let other countries self determine, and stop interfering. I don't expect vaccines to be eliminated, but Bobby might recommend delaying some, or giving parents the option to opt out of some. IDK I'm hoping he can MAHA, he really wants to help IMHO. And for the rest... yeah heroic in that I think he is doing the right thing, even though he will be hated for it. But that's just my opinion.
"TBH I don't pay all that much attention to what Trump and Musk are doing."
Previously you said you listen to what they say. But you don't pay attention to what they do. And then you're confused that people have a different opinion than you?
You're doing this the wrong way round. This is how you get taken in by con men.
I don't pay attention to the news, which is distorted this way or that by the news agency. I've listened to some of the senate confirmation hearings. And the above video with Trump and Musk talking to reporters in the oval office. That's not reporting, it's the ground or first level truth. You can say, no Trump and Musk are talking at a higher level and conning me. That could be true, so shame on me. But until I see evidence otherwise, I'm going to assume that they are talking ground or first level truth to me. It is the only way we can effectively communicate with each other. And I give you here, my first level truth.
w/ love George
1. The purging has already started. Trump and Musk have fired hundreds of people and put thousands more on hold pending termination - in many cases, illegally. https://www.reuters.com/world/us/are-trumps-mass-firings-federal-workers-legal-2025-02-13/
2. A court has already found that Trump's EO regarding trans medical care was likely to be harmful enough to grant standing for them to sue and to suspend the order temporarily https://apnews.com/article/judge-restraining-order-trump-transgender-health-care-8f8d935a3e757a1700dfb7363a67b07b
3. So when the US funds foreign journalists who report on their own governments, to counter authoritarian propaganda - it's the US that's Orwellian? I thought the countries suppressing journalists were the Orwellian ones. https://www.theguardian.com/global-development/2025/feb/11/trump-usaid-cuts-freeze-press-freedom-ukraine-afghanistan-media-rsf
4. One state has already eliminated vaccination drives for seasonal flu shots, and I expect more to come: https://apnews.com/article/louisiana-vaccination-surgeon-general-8e2eecde047648e6fb62c4b44edd6566
5. RFK went to Samoa to campaign against vaccines in the midst of a measles outbreak that killed 83 people. There is currently a measles outbreak in Texas. Measles can be eliminated by vaccines.
It seems like the theme here is that you doubt that Musk, Trump, and RFK are going to do things *they've already done*, and your answer to "Musk posts a bunch of Nazi stuff" is "well I don't look at it."
So, I think you have your answer: people hate Musk because they pay attention to what he says and does, and read news, and know what's going on in the world. You like Musk, Trump, and RFK because you have no idea what they are doing and can't be bothered to find out.
Sorry I meant to paste the link for 5. https://apnews.com/article/rfk-jr-samoa-measles-kennedy-vaccines-pacific-42a9cb583c71f165699b16710884c474
A good trick for understanding stuff like this is to imagine the situation with opposite partisanship.
For example, imagine that Harris was president, and she forcibly shut down ICE by decree and fired all border agents against the wishes of congress. You'd be pretty mad in that case, right?
Hmm my immigration views are mixed at the moment, so that's not a good analogy for me. And TBH what Musk is trying to do is fix our budget. I'm all in favor of that, and if we had to open our borders to get our budget under control, I'd sign up for that.
What Musk is trying to do is to destroy or hamstring government agencies and programs he disapproves of. There is zero possibility that the things he is doing will "fix" the budget, or even close the gap sufficiently that other plausible measures might fix the budget. And if he were trying to fix the budget, the first thing on his agenda would be to try and convince Donald Trump to walk back his tax cut proposals.
Musk isn't an idiot, he knows all this, and he's still going Full DOGE. So, notwithstanding the name, he isn't trying to fix the budget.
Hmm yeah you may be right, and I'm a stooge. In which case I'll admit it. I think we need more time to find out. I'm sorry I don't know anything about Trumps tax cuts. Do people expect them to be approved? None of this budget stuff happens unless congress gets involved.
I remain as hopeful as I was in the first few months of Obama's first term. That all came to nothing, but I remain hopeful again. (I've got practice I'm a Buffalo Bills fan :^)
> Do people expect them to be approved?
This is literally THE NUMBER ONE PRIORITY of Republicans in Congress. The chances of them not going through are extremely small.
I hate political news as much as anyone, but this is a case where even the most cursory reading would correct an extremely inaccurate view of the world.
The number one priority of Republicans in Congress is keeping themselves in Congress. But, yeah, the tax cuts are number one in the hype and I suspect in the top three actual priorities. And the few GOP congressmen who are willing to sometimes stand up to Trump, have to pick their battles carefully. They're not going to die on the hill of "Americans should pay more taxes!"
So, yes, we are going to get another five trillion dollars in debt out of this.
The Republican's are currently targeting a deficit increase of 4.5 TRILLION over the next decade. There's no way that you can pretend that Musk is killing highly effective programs like PEPFAR (about 6 billion a year) because he cares about the debt.
This is a bit like defunding the police "because you just can't spare the money and hard choices have to be made" and then spending 50x as much on student loan forgiveness.
I think pepfar is back to being funded? But I don't know the status.
I don't understand your defund the police and student debt point. Personally I think police could use more funding, and student loan forgiveness is a terrible idea. (I also think that student debt not being able to be discharged in bankruptcy is a terrible law.)
Oh a deficit of ~0.5 trillion per year would be way better than what it is now. https://fiscaldata.treasury.gov/americas-finance-guide/national-deficit/ Of course I assume the Republicans are making all the savings in the last half of that decade and that this year it will be over 1 trillion again... sigh.
> Oh a deficit of ~0.5 trillion per year would be way better than what it is now
You misunderstand. 4.5 trillion isn't their target for the deficit over 10 years. It's their target for how much to INCREASE the deficit over 10 years.
To increase the debt, not the deficit. Or phrase it some other way, but nobody talks about 10 year deficits, if you're talking about a deficit you need to use annualized terms.
> I think pepfar is back to being funded? But I don't know the status.
I think Musk made noises in that direction, but it's hard to believe when he's fired everyone and "put it through the woodchipper". At this point, there's probably nothing left to unfreeze in the first place.
> Personally I think police could use more funding, and student loan forgiveness is a terrible idea.
So do I. I was just using it as a hypothetical partisan-reversed example of what Republicans are doing.
What is wrong with our budget that makes it need to be fixed?
Has any of the activity undertaking by Musk actually moved the needle on solving that problem?
Why do Musk and Trump have to pursue their "fix" in an illegal manner?
We continue to run big deficits and the national debt continues to grow. This is not sustainable, it will lead to massive inflation. Trump has only been in office for ~1 month. TBH I'm not sure he (Trump) has any interest in fixing the deficit. I think Musk does. And yeah when dealing with the budget they have to get congress involved to do it legally. Hopefully that happens sooner.
Republicans are currently trying to massively increase the deficit by cutting taxes and increasing spending on things they favor, same as they've done in every past administration. Current target is an increase of 4.5 TRILLION over the next 10 years from tax cuts. That easily dwarfs anything plausibly achievable via spending cuts.
I agree that it's unsustainable, but so far Republicans have shown no inclination to do anything but make the problem worse. You'll know people are serious about the debt once they start talking about raising taxes. Heck, even following the spending limits they themselves forced on Biden would be a good first step.
Well-said. And --
"against the wishes of Congress" is one thing.
"In violation of multiple clear federal laws each in place for decades" is a whole other level, one which no POTUS has attempted during the lifetimes of anyone present.
And then "flatly defying rulings from several federal courts", which the VPOTUS and others in the administration are openly suggesting right now...yea that's when we really finally do enter banana-republic territory. That's a huge step never seen before in the USA*.
(* No, Andrew Jackson didn't do this. He declined to help enforce a SCOTUS ruling upon the party that had lost the case, which was not great coming from a guy whose oath of office is to "preserve, protect and defend the Constitution of the United States". But that wasn't what Vance and Musk are now talking about doing which would be way worse.)
Is "gulf of america" a shit test for google? Maybe also apple
Google are total cowards who just want to make billions upon billions of dollars in peace. For an example on the other side: You currently can't get the USD to Brazilian Real conversion rate from Google Finance, because the Brazilian government got mad that they listed a too high rate once.
This isn't the first country that has attempted to enforce rules about maps. What is the body of water between Korea and Japan? Where is the line-of-control in Kashmir? For at least a decade, Google has had a policy where it follows government dicta in the country that issued them. So in China it draws maps the way the Chinese government says; in India it draws maps the way the Indian government says, etc.
So it isn't the least bit surprising that Google did this. As far as the Trump regime being "authoritarian" or "illiberal" like China or India ... you can do the math.
Yep, Apple maps succumbed. Must kiss the Sovereign's boot.
The 2 + 2 = 5 edict comes tomorrow. Susan Collins, Lisa Murkowski and Mitch McConnell are on the fence on that one. They'll need Sean Spicer to come back and say "That's how it adds up now. Period."
Apparently, Vance is the one who managed to force all the senators into line. Or at least Vance is playing the Good Cop to Musk's Bad Cop.
https://www.politico.com/news/magazine/2025/02/08/jd-vance-house-gop-column-00203240
It was enough to ban the Associated Press from the white house, so yeah I guess they'll probably push it as far as they can.
You know, I wonder if Wikipedia will come under attack for this. They don't even have "Gulf of America" anywhere in the introduction anymore.
https://en.m.wikipedia.org/wiki/Gulf_of_Mexico
It's near the top of the section called "Name" -- which seems appropriate given the short time the "Gulf of America" conversation has even been a thing:
https://en.wikipedia.org/wiki/Gulf_of_Mexico#Gulf_of_America
It used to be in the introduction, and they removed it. Anyways, it doesn't matter whether you think it's justified, only whether the administration thinks it is.
The Gulf of Mexico isn't U.S. property. What do you expect the Trump admin to do to the Wikimedia Foundation over it?
Forcibly remove its leadership and replace them with government sympathizers? It's not just about the Gulf of Mexico, of course. Obviously they're not going to be happy with the fact that the world's biggest encyclopedia is being run by leftists.
> this page is locked to prevent vandalism
good old wikipedia, pretending to be neural while having a strong opinion
They do have a strong opinion against vandalism.
They have many other strong opinions as well, many of which I disagree with and many of which they aren't transparent about, but they're right on this one, a stupid edit war over that article would be a bad thing and preventing such is their responsibility.
I have yet to see a version I consider well done; people are there for a spefic story, that story should be covered
It appears they are covering it. This is a link in the main Wikipedia Gulf of Mexico entry
https://en.wikipedia.org/wiki/Gulf_of_Mexico%E2%80%93America_naming_dispute
This week in nominative determinism, Chana Messinger is now a comms exec 80,000 hours (https://x.com/ChanaMessinger/status/1889377530278519077, I wish her the best in the new role)
Someone please tell me if this is true, because now I'm singing to myself "there's a centre, down a mineshaft, in Boyers, Pennsylvania" (inspired by this https://www.youtube.com/watch?v=Jk22CHRAoPU):
https://x.com/DOGE/status/1889437908094042277
https://tribune.com.pk/story/2528129/iron-mountain-limestone-mine-used-for-federal-retirement-paperwork-processing
"During a press conference on Tuesday afternoon, President Donald Trump and Elon Musk shared a bizarre exchange, with Musk making an unexpected mention of a “limestone mine” where federal retirement paperwork is processed.
The mine in question, as Musk described, is not a hypothetical facility but a real location called Iron Mountain, situated in Boyers, Pennsylvania. It plays a central role in processing federal employee retirement applications, a process that remains remarkably analog, despite various efforts to modernize."
You can't make this stuff up! Musk said that the rate of the elevator in the mine shaft limited federal employee retirements to 10k per month.
You don’t know about that over in Potatoland? Jeez, yeah, they use trolls for workers in Boyers. Billboard at city limits reads “Trolls for the Olds,” with a drawing of a cute troll, looking sort of like a giant Danny DeVito. Though actually they look like Giant naked mole rats and smell like toilets
Definitely real. It comes up in the news every so often:
https://www.washingtonpost.com/sf/national/2014/03/22/sinkhole-of-bureaucracy/?itid=sr_1_c95690db-5ad1-4638-ae52-9ac740704c6d
They don't call it the Deep State for nothing.
It's like we can't take anything for granite around here
If someone want's to learn about this some more, you better believe they're gonna get an oarful from me
In other Bumbling Buffoon news, behold our new defense secretary giving away negotiation positions before any negotiation even started:
"Pete Hegseth, the US defence secretary, has said it is "unrealistic" to expect Ukraine to return to its pre-2014 borders, when Russia first took control of Crimea. [...] The US defence secretary also downplayed suggestions of Ukraine joining Nato."
source: https://bbc.com/news/articles/cy0pz3er37jo (which explains the "c" in "defence")
"Only the best people", "negotiate from position of strength", who hired these morons? Oh, never mind, we the people did.
OTOH it's "Gulf of America" now, yay us, that'll show them who's boss.
Depending on if putin is a full psychopath or merely a pure nationalist he may actaully drink his own kool aid that sections of ukraine were loyal to russia and very russian; the russian speaking sections of Ukraine should have been given to russia, as they clearly made their choice.
He seems to be in that cool-aid "one people" territory, he keeps pushing "Ukraine is not real" thing every time he can. The terrible irony is that it's the eastern part that was the most Russian-speaking, and got the worst death and destruction. There were some numbers that nobody ever disputed, like, in the year prior to the '22 invasion the total number of casualties in Donbass was in the 100-200 range, most of which were combatants. I think less than 10 civilians were killed. Known, documented with real names Russian casualties alone are now >90,000.
I'm not seeing the benefit of pretending Crimea is going back to Ukraine. Unless the Ukrainian military can conquer it first, there's very little chance a peace agreement would include any territorial concessions by Russia. Maybe some mutual trades for borders that make more sense, but nobody thinks Russia is just going to give back any of the land it occupies. Crimea least of all. If there's going to be a negotiation, it's going to be about land conquered more recently, and not even all of that. Crimea was gone in 2014 when nobody seriously questioned it at that time.
Yes, this may be one way to get a peace deal, or there may be all kinds of arrangements that leave some questions in the grey zone, etc. The point is you just don't per-announce obvious concessions before starting negotiations. It's a rookie mistake.
Maybe, but I don't think so. There's a small value lost to Putin, but because absolutely no one thinks that's a real point of contention, that value is incredibly small. It would be like pre-committing to not nuking Moscow as part of the deal. Technically a give, but again not worth anything.
On the other hand, there is a value in getting your allies at the negotiating table to come to an agreement about what to ask for. Whoever the holdouts would be (presumably Ukraine but likely also people in the US) need pushing to not slow down or stop the process. Sometimes you need to publicly commit your side so that they are forced to accept it, or otherwise fight in back channels forever.
So the question is not whether something of value was given up, but whether the value of conceding to Putin is more or less than the value of getting your own team in line. I think the value of conceding was so low that it would be difficult for it to be relevant, but regardless the value of getting your team in line is more than what was lost.
I agree with your analysis of potential dynamics team dynamics, but note how we are now engaging in the exact thing we should not be: presupposing the value of the concession to Putin. Why do this? Good negotiators just... don't do this stuff, things can only be communicated to the relevant parties in private. We don't actually know what is and what isn't important to Putin. Why guess when one could just say nothing? The default choice is always "no comment", how hard is that? One can even say, if the zipper store is out of stock, something to the effect "we don't want to jeopardize possible future negotiations, I'm sure you understand".
But that requires a degree of professionalism and competency that... nevermind. Like I said, we the people voted this crew in.
Related partial Mea Culpa. I don't disagree with what I wrote before, but apparently Hegseth has said a whole lot more that Putin can and will use to inform negotiation policy going forward.
I never heard of him before he was nominated, but I am now leaning more towards "idiot" than I was before.
Greatly appreciated!
Is your complaint that it *is* reasonable to expect Ukraine to return to its pre-2014 borders? Or just that starting with an unreasonable ask leads to a better deal in the end?
The latter. You have to be able to give something to the other side to make them look like tough negotiators who won something.
You want to ask for *something* that you don't really expect to get, both because you might get it anyway and why leave that on the table, and as you note because if you don't get it you can make that look like a concession that you are making in exchange for something they are giving up. But it has to be something at least vaguely plausible, or nobody will take it seriously.
If I offer to sell my second car for $20,000, there's a good chance I'll get the $15,000 it is probably worth and I *might* get $17,500. If I advertise it for $200,000, nobody is even going to call.
Asking Russia to return Crimea to Ukraine now, without some truly major change in the military situation, is like trying to sell a 2017 Mazda for $200,000. It costs you nothing to say "this is a decent used car but we're obviously not asking $200,000 for it".
Well, it costs you nothing in the context of the negotiations. There may be some outside-context reason for it, but you'll then want to privately make it clear to the actual negotiating partner that this is just posturing. And usually, it's best not to do that sort of thing.
Thank you, this is a good summary. WRT Crimea the potential solution space is wide open in both time and legal dimensions (back in 22 there was a proposal floating to just punt the question for the next 15 years, which actually was a good idea - at that point Putin will likely be dead and Zelenskiy out of power, opening space for new people to make a deal). Hasty public pronouncements drastically shrink the negotiation space.
BTW I think CSPAN TV thing, where they transmit Congress deliberations live, is a terrible idea that may have directly contributed to the polarization and dysfunction of Congress.
Definitely agreed on CSPAN being a bad idea.
Then why be such a piker? Start by demanding unconditional surrender!
Seriously: negotiations don't really work the way you're suggesting they do. For starters, you can't give a thing you don't actually have.
> Start by demanding unconditional surrender!
Probably better than to start by offering unconditional surrender.
Reverse stupidity is not intelligence, as folks here like to say.
Look, one thing I do know about negotiations is that one should not publicly give the opposing side what it wants before the negotiations started. The best thing to do before negotiation is to say nothing. How freaking hard is that?
I guess too hard for the team of stable geniuses running the show now.
Uh... I don't know why you think the US is negotiating on behalf of Ukraine. The US-Russia alliance is looking to be more real by the hour.
Oh, that’s a hoax…
Oh, Trump promised to end the war in 1 day. Let me guess, the bumbling buffoon was just running his big mouth, as usual. "Who knew healthcare was so complicated".
Also, what alliance? With us? The country that at this point is famous for breaking every alliance and promise it makes, sometimes for no reason whatsoever? Why would anyone want to be in alliance with us? Milking us for money/weapons/whatever, that's fine, but an alliance?
Russia and the Trump administration do have significantly more in common ideologically than the EU or the rest of the world.
Considering that even two racial supremacist nations of different races ended up developing an alliance just because they were both imperialist, I don't think a US-Russia alliance is off the table. After all, if we're going to have WW3, it would be fitting for there to be a new Axis powers.
Yes, both a resemblance (not physical, at least not yet) and a strange affection Trump has for Putin has been noted....
Who honestly thinks 2028 will be a "free and fair election"? Why wouldn't they "DOGE" the election system in the name of "fixing" it?
Is there honestly any way to prevent this?
Could an ASI "aligned with the Democratic party" or "aligned with a (d)emocratic worldview" help fight back?
"Could an ASI [...] help fight back?"
Why yes, also Gandalf could lend a hand, together with the tooth fairy.
One reason these goons have been so successful is that they actually do things, while the other side whines on bluesky and reads the thoughtful articles in the Atlantic, hoping that some omnipotent entity will take note of their concerns and swoop down to write the wrongs and punish the evildoers.
I mentioned ASI because I think "conventional avenues" will no longer work.
I think America's situation, though people don't realize it yet, is akin to Russia or China. How do people in those countries "work within the system" to restore democratic processes and liberties? They don't. They get jailed. They get killed. Those countries are 1984. America is probably only months-or-years away from the same.
ASI... it IS a recommendation of how to "actually do things". We need forces larger than even those imagined by George Orwell. ASI is one of the few things that qualifies (or perhaps something else). Perhaps it can carry out some massive pro-social pro-humanitarian social-political manipulation. How ELSE do you propose to soften such regimes once they've solidified?
What would the hypothetical ASI even like...do? This is a power/law problem, and those aren't a part of the universe like atoms or magnetic force, they exist in the minds of humans. Humans are the ones to make changes in reality. If the ASI prints "Hey, stop that" on a screen, what does that to do the humans doing the thing? If the ASI somehow has brute force on its side, how does that work? Why should the humans actually carrying the guns listen to the computer?
Hundreds of millions of people ALREADY DO do what computers tell them. (why bother with specifics? If you're reply to me, a stream of text has successfully gotten under your skin, lol)
Seriously, an ASI with read/write access to enough systems could manipulate, hack, disrupt, lay-utter-waste on a medieval scale.
I'm not saying they'll definitely fight for *us*. But I sure as hell hope some of them DO!
>How ELSE do you propose to soften such regimes once they've solidified?
There have been thousands of bottom-up regime changes in recorded history, and none of them involved ASI.
Don't underestimate the number of brutal regimes that faded PURELY because the Invincible Autocrat died and gave way to a weaker person or group at the TOP...
Genghis Khan, Tamerlane, Stalin, Qianlong Emperor, The First Emperor of China...
These people were untouchable from inside-and-outside their realms...
i.e. TOP-down regime change
Sure, but what does that have to do with my point? My point being, even the most oppressive, entrenched, and militarily superior regimes can be overcome from the bottom, both violently and non-violently.
What I said had everything to do with your point:
I think you are wrong.
In all the cases I mentioned, it is IMPOSSIBLE to prove whether those regimes could have fallen... EVER... without their central figure dying. All the leaders I mentioned had a vise-like grip on society. Resistance = pointless. There's probably 100's more examples.
This was a central theme of 1984: a system that was not only evil, but might not ever collapse.
"If you want a picture of the future, imagine a boot stamping on a human face— forever"
The problem is that ASI is not real. It will not be real in 2028. It may, depending on your definition of it, be impossible. So whatever the solution is, it's best to not throw impossible deuce ex machina things into the mix.
How about having ASI turn the people you think are bad guys into paper clips? Then everyone else also gets a turn giving ASI their bad guys list for the paperclip transform treatment. Oops, one of them put you on the bad guys list!
Very small. Elections are run by the states. The federal government has very little say in it.
This bespeaks a very small imagination. The state government agencies will topple as quickly as the federal ones. Who's gonna stop him?
The States will be His.
>Who's gonna stop him?
Presumably the States would. The federal agencies are under the command of the President, the state agencies are not. Short of sending in the army how exactly is Trump supposed to "topple" them?
The last election cycle already saw an influx of Trump-friendly secretaries of state in Republican-leaning states, so that shouldn't be necessary.
It wouldn't even be a question of the states having to "stop him". The states will continue to run the elections according to their own laws, as always, and if Trump et al insist that they should do something different, then the state governments will simply and correctly note that he isn't the boss of them and ignore him. Properly so, just like they would if one of us were to call them and insist that they rejigger their election laws.
Trump would have to stop *them*, and as you note he'd pretty much need to send the Army to do it. And I doubt the Army would follow that illegal order.
>I doubt the army would follow that illegal order
Not if they think it's a legal order.
And consider the fact that most states are Republican to start with. And consider the fact that all these states have to, ultimately, submit their ballots for president to the U.S. Capitol. They'll simply "count what they feel like counting". And Trump-Musk have proven repeatedly they don't care what the courts tell them to do.
Who do you think would win in a showdown between the U.S. Military and some state's National Guard?
If the President / Executive tells an armed agent to shoot someone.... and a judge tells an armed agent to shoot someone ("the opposite someone"), who does the armed agent shoot?
>Not if they think it's a legal order.
Why would they think invading a state government in order to rig an election would be a legal order?
>And consider the fact that most states are Republican to start with.
So the Republican controlled states will rig the next election by...sending Republican electors? Wouldn't they have done that anyway? You kind of need to get some of the purple and blue states rigged if you want it to work.
>If the President / Executive tells an armed agent to shoot someone.... and a judge tells an armed agent to shoot someone ("the opposite someone"), who does the armed agent shoot?
All members of the military, from the highest general to the lowliest private, swear the oath of enlistment.
“I do solemnly swear that I will support and defend the Constitution of the United States against all enemies, foreign and domestic; that I will bear true faith and allegiance to the same; and that I will obey the orders of the President of the United States and the orders of the officers appointed over me, according to regulations and the Uniform Code of Military Justice. So help me God."
The Uniform Code of Military Justice makes clear that only lawful orders must be obeyed. Who decides what is lawful? Judges. I expect army grunts to obey their commanding officers, and I expect those commanding officers to obey the judge. At minimum I expect them to say "Hey, lets not shoot anybody right now until we've sorted this out." Their oath to defend the Constitution trumps their oath to obey the President, and capturing state capitols to rig an election is pretty clearly against the Constitution. Worst case scenario I expect some generals to side with the president and some generals to side with the courts and now we've got a civil war going on.
Was 2024 your first election? I have heard this same scare tactic, from both sides, every election since 2000.
Are you going to ignore the fact that one side has actually tried a well-documented plot to overturn election results, and the other hasn't?
"Plot" is I think a rather generous estimate of the planning involved in the ad-hoc effort of 2020, and I haven't seen any documentation of any serious planning for 2028.
Previous presidents and presidential candidates as far back as I can remember have presumably looked at the possibility of trying to overturn election results, and with the very marginal exception of Al Gore have basically said "no, that wouldn't work and would just make me look bad" and so didn't try it. Trump, being Trump, tried it anyway and indeed it made him look bad. In all cases, a significant chunk of the opposition has said "There will be no more free elections, the fix is in, the Republic is DOOMED!!!", and they've been wrong every time.
Nothing much has changed on that front.
I don’t think the 2000 election is a very good comparison to 2020. There was serious on-the-ground uncertainty about the count in Florida. It was a mess. In retrospect it seems to have been resolved correctly (Bush got more votes).
I would definitely refer to what happened in 2020 as a plot, although it was a bit of a Hail Mary. Every recount, and there were multiple, confirmed the initial count. And yet they persisted in calling it Fraud.
Trying to get Pence to reject the electoral college vote sure seems like a plot to me. The fact that he wouldn’t do it is pretty important. Have you pondered what would’ve happened if he had gone along? Would that have made it a plot?
Yes, yes they are.
Oh come on, I know the 2020 election looked very fishy, but even places like Maricopa County flipping could be explained by a small number of voters genuinely changing their minds, once I looked into it.
Sure, the mail-in voting was a huge mess, but you can't simply assume the Democrats rigged the ballot boxes.
That is what you meant, right?
The somethingiswrong2024 Reddit is both entertaining and very sad; a mass delusion of "It's only election denialism when the Republicans did it, we say the exact same things about how this election was stolen but we're right! Musk's team of babies were the elite hackers who stole it for Trump! Yeah sure back in 2020 we said there was no way voting machines could be hacked like that but this is different!"
https://www.reddit.com/r/somethingiswrong2024/
There is always a lunatic fringe and I fear you pay too much attention to it sometimes. 2020 was of a different order in my view. It wasn’t just a lunatic fringe. What’s more there is a pretty serious effort still to label it fraud in retrospect, and that is not coming from a lunatic fringe, that is coming from the top.
No. What he meant is Trump tried to illegally overturn the election in Jan 6, with the fake elector plot. Among many other illegal things, he even pardoned people like Roger Stone, who did not testify in Congress despite risking going to jail bc they knew Trump was going to pardon them, which he did.
You found people on the internet who say stupid things.
The question isn't whether conspiracy theories exist. The question is whether different theories are comparable in their hold on people, their scope and salience, I guess.
Personally, I don't know a single Democratic voter who believes that the 2024 election was stolen, and most Democrats I know would call someone who did believe it a conspiracy theorist.
I know plenty of Republican voters and with a few exceptions, they all either believe that 2020 was stolen or they at least think that it's a plausible theory.
I think my personal experience mirrors reality as a whole far more accurately than conspiracy subreddits. One strong indicator that this is true is the way that officials on the respective sides have treated the elections.
This election the left has very low engery; I think the 2016 "is trump a russian asset" a better comparison
Wow, what a scissor statement. I would keep arguing with you, but I suspect it would be a waste of both of our times.
You use the term 'scissor statement' incorrectly here, resulting in you avoiding confronting the substance of the claim.
The claim here is: the hoi polloi of one party, upon perceived loss of an election, broke into the the capitol, where at least n>1 of those who broke in intended to use force to change the results of the election,
and the other parties equivalent faction did not.
This is not a moral judgment, it is recognizing reality vs. The Party telling you to reject the evidence of your eyes and ears.
You're right, this is absolutely a scissor statement, because from my perspective you are obviously and completely wrong for disagreeing. Also I'm angry at you for supporting a government that will substantially increase the probability that global liberal democracy hegemony will be coming to a close in the next 4 years, and frustrated that from my perspective it seems like there's no amount of evidence that could convince you otherwise.
But good job for recognizing it is a scissor statement.
No, go ahead. Keep arguing. It's extremely well-documented that Trump tried in dozens of illegal or pseudo-legal ways to overturn the results... the Georgia ballots, the fake electors, the frivolous lawsuits, the incitement of a riot to storm the United States Capitol (why do we live in a world where this fact has to be reminded to people).
Now enlighten us: what did the "other morally-equivalent side" do?
You didn’t say it was bad behavior. You said, “[it was] calling for murder.” If you had said that in the first place, I would have basically agreed and posted no response. I would suggest that it was you who turned up the temperature. I agree that it’s bad form to voice those kinds of opinions. It’s not, however, any kind of obvious call to violence.
https://www.astralcodexten.com/p/ask-me-anything-22025/comment/92565275
I can't comment in the AMA so commenting here. In that thread Scott claimed that 'dream' and 'trauma' come from the same root. I don't think this is true. The words are very similar in German but while 'dream' is Proto-Germanic 'trauma' originates from Greek.
This doesn't much affect the point Scott was making in that comment other than subtracting one of the illustrative examples he used to make his point.
I was going to say trace up the etymology one step further up into Proto Indo-European, but the origin roots are indeed different:
Dream ← PIE dʰrewgʰ "deceive"
Trauma ← terh "drill, pierce"
>Dream ← PIE dʰrewgʰ "deceive"
Scott was obviously offering a subtle meta-commentary on deceptive etymologies.
Btw, the claim about "vacation" is also a bit off. It has nothing to do with "vacating the cities," it refers to time that is free ("vacated") from labor.
e.g. Chaucer, Wife of Bath's Tale (ca. 1395):
"Whan he hadde leyser and vacacion ffrom oother worldly ocupacion."
This sense is attested slightly earlier in French:
e.g. Oresme, Livre des Ethiques (1370):
"et par vacacion il entent repos ou cessacion de labeur ou de occupacion en negoces"
> "Whan he hadde leyser and vacacion ffrom oother worldly ocupacion."
Translated to modern English:
"The occupants from the other world had lasers, and our cities were vacated."
I wonder if Scott was inspired by a scene in Shutter Island where this connection gets made, or if he has even seen the film.
My COVID update for epi weeks 5-6 of 2025.
1. After a month of silence, Biobot finally posted an update. Wastewater numbers indicate the current XEC wave has receded a bit. But we might see a secondary bump as the numbers level off in the West and Midwest and climb a bit in the South. If there is a secondary wave, LP.8.1x would be the likely candidate to cause it because LP.8.1 has the highest growth rate of any variant in the US at this time.
2. This has been the mildest COVID wave ever in the US if measured by ED visits, hospitalization, and mortality.
3. Even though the Morbidity and Mortality stats are no longer being updated and displayed on the CDC website, we can still get the archived stats out of Wonder. And Wonder shows us that COVID came in 14th out of the 15 most common causes of death in 2024. Down from 3rd place in 2021.
4. @MichaelSFuhrer on X has noted that lab-confirmed influenza deaths have exceeded lab-confirmed COVID deaths for the last couple weeks in the US, the first time this has happened since the beginning of the pandemic. We won't have access to interesting information like this unless the new CDC director (whoever that is) puts Wonder back online. Of course, there are 8x more ED visits for influenza than COVID right now. The test positivity rate for influenza is 6x that of COVID — so, Influenza is the respiratory virus of the moment. RSV rates are falling, though.
5. Not much has changed on the A(H5) HPAI front. According to the USDA's APHIS, 35 new dairy herds were infected in California in the last 30 days, and 7 new herds in Nevada. No new infected herds in the past 30 days in any other states. Oh, and a dairy worker in Nevada was infected. His only symptoms were pink eye, though. Tentative numbers from the USDA suggest that chicken culling may be less in January. A(H5) is seasonal in wild bird populations, and they infect domestic flocks of poultry (chickens and turkeys). We probably won't see the price of eggs go down until the spring when A(H5) recedes in wild bird populations.
Slides here...
https://threadreaderapp.com/thread/1889153381325054133.html
I came across this mathematician arguing that the capability to be a mathematician isn't really determined by genetics (in the sense of IQ), but largely by being in a freak context in childhood:
https://davidbessis.substack.com/p/beyond-nature-and-nurture?utm_source=post-banner&utm_medium=web&utm_campaign=posts-open-in-app&triedRedirect=true
I think this quote was interesting:
> As a first year student at the École normale supérieure, I followed a semester-long combinatorics course given by Xavier Viennot, aiming to construct a “nonverbal proof” of a messy formula by Ramanujan. The approach used intuitive objects such as combinatorial trees, colored beads, dominos, and hexagonal tiles, very much in the spirit of this video but with much more detail. This course permanently altered my worldview: I realized how fertile it was to assume that, in the end, Ramanujan could only come up with his formulas because there was a path to finding them obvious.
I’ve written above that my hereditarian beliefs were run over by my progress trajectory in mathematics. I should add that this was inevitable, as those beliefs were actively blocking my progress.
If you believe mathematics requires superhuman computing abilities, you’ll miss its true essence—finding simple, intuitive patterns in what initially seems unintelligible.
I wonder at it, since there's every reason to believe that getting into math in the first place has a high floor and you can't get there without the lucky genes, but there's also that, where he had to ditch the belief that genetics places a ceiling on how high you can go in math.
I am confused, because I don't think that I believe something very different from the author, and yet I would express it completely using completely different words, with correspondingly different connotations.
IQ doesn't mean that you *magically know everything*. That is a mistake that both many fans and many opponents of the concept seem to believe. On one hand, you have people who believe that just because they are Mensa members, it automatically makes them qualified to express strong opinions on relativity or quantum physics, even if they never studied it. On the other hand, you have people who look at the former, realize that they are talking bullshit, and conclude that intelligence plays no role at learning relativity or quantum physics. Both of them are wrong, for the same reason.
My model is like: "intelligence is *efficiency* of the process that translates time/attention to understanding". More efficient in the sense of "faster", but also in the sense of "can achieve more". Ceteris paribus, if two people study the same thing, the more intelligent one will probably understand it faster, and it is also possible that above some threshold of difficulty the less intelligent will simply fail to understand.
But that still requires some time/attention to be spent. If the more intelligent person never studies a certain topic, then the less intelligent person can get an advantage by simply reading the textbook. And if the less intelligent person spends 10 years reading textbooks, doing experiments, debating with other people who studied the textbooks and did the experiments... then they can accumulate a huge amount of advantage that the more intelligent person simply cannot overcome in a single afternoon. (Possibly could overcome it in 5 years of focused study, though.) The exact numbers depend on the topic, some topics are more understanding-costly, other topics are more memorizing-costly.
So the combination that unlocks being good at math is being intelligent + being interested in math + spending a lot of time doing math. Also, getting good textbooks and tutors, because they can make the knowledge easier to digest, which makes them another multiplier besides intelligence. Unfortunately, textbooks are not interactive/adaptive, and good math teachers are rare (because that requires a double talent: to be good at math and good at teaching).
And the thing that makes you interested in math during your childhood is probably something random. Some kids become interested in dinosaurs, other kids become interested in puzzles, and one of those paths can lead to mathematics. Calling it "beyond nature and nurture" sounds like clickbait; but there is a lot of randomness involved in human life, and some random things set you on a path where your experience accumulates. As you get better at math, you start winning math competitions, which probably makes you want to spend more time doing math, etc.
But without some level of IQ, starting that path would not be possible. You would find an interested puzzle, spend some time trying to solve it, then fail. Probably wouldn't motivate you to try more puzzles, and even if you did, you would probably fail again.
The more intelligent you are, the more your future skills depend on luck, because the more paths are open to you. Math is just one of many options; if for some reason you start to hate it, there are many other things to spend your time/attention on. And there are only 24 hours a day, so you can only learn a small part of what in theory is in your reach. Good education can increase that fraction. But good education, if it could be tailored to everyone's individual needs, would actually make the difference in math skills *increase*, because the people who are more intelligent and more interested in math would advance even faster.
I have a PhD in maths, though not from somewhere as prestigious as the ENS. I agree broadly with the thrust of the linked piece - success in mathematics is much less to do with genius than with persistence. I think freakish experiences in childhood is maybe putting it a bit too strongly, though.
When I tell people I teach maths, the usual response is "I always hated maths" or "I was never good at maths in school". If you hate something, and avoid it at all costs, you'll never develop facility in it. The people who avoid that emotional response are the ones who progress with it. Personally, I was good at maths, better than my peers certainly. But I was late deciding to specialise in it. I've written 30-odd papers now, none of which will change the world; I have been tidying up and extending my own area of expertise - some day I'll collect it into a book, put the house in order and then move on to something else.
I happen to speak a minority language well also. I get the same response to this: "I always hated Irish in school" or "I never wrapped my head around the grammar". It's not that it's inherently more difficult than other languages. But it's a socially acceptable attitude here. You don't hear people proclaiming their inability to read (in English) or their total ignorance of history and politics in polite society.
For whatever reason society sanctions these myths that mathematics is difficult, and that only special people can do it. This is self fulfilling. I like to compare it to the gym (especially when teaching young lads): if you declare you'll never be able to bench press 60kg (or 180kg or whatever's clearly currently just beyond their capability) and never practice then you never will. But more or less anyone can work at it and improve greatly. You might never make it to the Olympics, but with moderate and consistent effort you can get into the 5th percentile.
Can I just compress this to `accumulated advantage'
That is so interesting. So you disagree that there's even a high floor to get into the field in the first place? It boils down to persistence?
I disagree that the requirement for getting in is unquantifiable genius. Like most other fields, you need some natural aptitude, but a good attitude and hard work will get you quite far. Maybe people underestimate how much work is involved?
Do you agree that most people could, with effort, get through a 4 year undergrad program in mathematics? This seems reasonable to me.
From my experience, most students who stick at it long enough (and some took quite a while) will write an adequate masters thesis. I've only supervised one PhD student (who I knew quite well beforehand) so can't comment on this, but there are plenty people who top out at PhD and don't go further - I'm inclined to think many people, if given 3-6 years without the cares of a full-time job & kids and etc, and weekly meetings with an expert, could produce an approximation of expert-level work in that field.
> But more or less anyone can work at it and improve greatly. You might never make it to the Olympics, but with moderate and consistent effort you can get into the 5th percentile.
...Not if you're a woman. If it's impossible for your body to build the muscle necessary, then it's impossible to make the necessary progress.
This example was tailored to young men interested in sports and the gym. In this case one would benchmark against people of similar age, gender and bodyweight. Or use a different sport as an example - is the general point unclear or objectionable?
You can see how that's applicable to other fields, yes? It doesn't matter if you can accomplish more relative to your genetic peers if those results still aren't worth a damn. This is why nobody watches women's sports.
Despite trying to imply otherwise, you were just better at math than your peers. Yes, they probably could have done better if they stopped trying to avoid it, but they still wouldn't have reached the heights that you did. And of course, you were never going to reach the heights that history's great mathematicians did, given your lack of meaningful contribution to the field.
We are limited both by the circumstances of our birth and the constraints of our flesh. It's ridiculous to claim otherwise.
I don't equate meaningful contribution (which I have made) to changing the world (which approximately no-one does).
My point is precisely that you don't have to be in the top ten at something to be proud of your achievements. In mathematics or anything else.
"This is why nobody watches women's sports."
Sometimes you just throw these things out there without bothering to stop and think for a moment, come on! Yes, fewer people do, but that's far from "nobody". Women's tennis is quite popular for example. And in my favorite sport, MMA, women typically compete in the same events as men, and even have headlined UFC events.
...Why does he have to ditch that belief? IQ is linked to pattern-finding as well. Hell, a lot of IQ tests I've seen are just unscrambling words.
This seems like an odd thing to bring up. In what contexts are you looking at "a lot of IQ tests"?
It was leading him to believe that he could not achieve any impressive discoveries because he hadn't shown signs of genius, he believed his mathematical ability was genetically bottlenecked. After breaking through that limiting belief he was able to prove a conjecture from 1970, which seems fairly major.
...But they weren't bottlenecked because he actually was intelligent. Just because he wasn't a savant or grow up in an environment where his skills were groomed from a young age doesn't mean he didn't have the capacity for this all along. The point is that most people lack the capacity for this in the first place. If you take some baby from the from the slums and make him live the same life he did, you are not going to get the same results. He is just built different.
Well, one would have to posit that if mathematicians' pattern-matching abilities are an expression of hereditary intelligence, then musicians must be in the same boat. And Bessis mentions that Scott felt his lack of math and musical skills was hereditary. But there have been studies that show that children who undertake musical training show statistically significant, improvements in their IQ scores over time (median ~1/2 an SD).
Sasha Gusev uses the red-haired children thought experiment to illustrate the differences between direct and indirect effects and their relationship to heritability.
https://theinfinitesimal.substack.com/p/comments-on-no-intelligence-is-not
That just shows IQ test results aren't representative of intrinsic potential. I'm not even defending IQ tests here, I think they are far less than perfect in both design and execution, but that doesn't mean intelligence isn't heritable. Which it obviously is, seeing as people that people that are brought up in extremely similar environments still get very divergent results. Both nature and nurture are ceilings.
I totally agree. But what's interesting about the Gusev essay is that he explains why the twin studies gave a significantly higher heritability value than current genetic studies do.
Not quite sure if there is an existing AI discussion to which this would fit naturally in:
AFAIK, one of the open problems with LLMs is doing real time learning in the sense of adjusting the weights in the process of solving real problems/queries/prompts.
We normally think of humans as learning continuously, but, IIRC, sleep is very important in forming long term memories. Maybe we _aren't_ an existence proof for continuous learning. Maybe a batch process _must_ always be part of learning???
That’s an interesting idea. Here’s something your question got me thinking about. Let’s say AI learns something in the course of an exchange with you. You give it one of your chemistry questions, and it gets it wrong, and then you socratically lead it to seeing how its answer can’t possibly be the case. Then you remind it of a couple main principles, and tell it that from them plus what you told it already it should be able to figure out the right answer. And it does. And then you have it explain the right answer, and it explains correctly, bringing in the general prinicples you reminded it of. OK, so that’s a good lesson.
What would the process of remembering the stuff it learned be like? It’s adjustment of weights, right? But as I understand it, AI can’t tell us anything about how it knows what it knows. It doesn’t sound like it has access to the weights that led to its first, incorrect answer, also I don’t think it would have any idea what weights to switch to capture the new knowledge it got from you. What I’m saying, and I may be all wrong here, is that the way AI remembers stuff is very different from the way we do. I’m sure we have the equivalent of weights, but we have a lot of other structure in there too. We have the anecdotal memory of the actual lesson, not just the info imparted but who said what. We have various principles of, for ex., chemistry, stored sort of together. We have the knowledge of what a priniciple is. And we can access the memory of our old incorrect ideas. So I’m wondering if learning the way AI does, via token prediction, produces something that sounds like it has understanding similar to ours about various things — but is set up in a very different way. The way it learned what it learned is not set up for the other kind of learning, learning from a lesson. However, I made this up just now out of whole cloth and it may have so many inaccurate bits of info as components that it’s just
wrong.
What do you think about the memory thing?
In normal use of an LLM, I don't think they're adjusting weights. The memory it has is just the context used to predict the next word, right?
Many Thanks! Yes, I agree that we have several different types of memory (at least as we can probe each others' memories "from the outside" so to speak): declarative memory, procedural memory, episodic memory etc. (I have only a fuzzy memory of the nomenclature, so may be getting it wrong).
I suppose that these are all ultimately changes in neurons (or neurons plus glial cells???) of some form - though there may be differences even at that level. I have a vague recollection that "muscle memory", "procedural" sequences of movements, is stored in the cerebellum, with a distinct architecture, different from cerebral neurons. I would guess that there is still a good deal of commonality at the low level between some of the types of memory.
I remember that "Fuel and oxidizer, when closely intermixed, can burn extremely rapidly, to the point of a true detonation." and "In the first class of my first chemistry course as an undergraduate, professor Nash brought the class to attention by filling a balloon with mixed hydrogen and oxygen, and setting it off with a loud bang." Declarative and episodic, or two slight variations on the same type of memory?
In the LLMs' case, I'm guessing that the contents of the context window is somewhat analogous to our short term memory, and the neural network weights somewhat analogous to our long term memory. My layman's guess is that using the contents of the context window as (additional) training data for the neural network weights might be the equivalent of our transfer from short term to long term memory. My current layman's (mis?)understanding is that this is too computationally expensive to do routinely. Still, there are usually ways to approximate expensive computations with cheaper ones...
You raise very good points about our keeping track of old ideas and keeping track of when and where (and who else was involved) an episode happened. And yes, e.g. in the current partially-correct (for o3-mini-high) titration question, I do indeed lead the ChatGPT through a Socratic dialog (albeit now a very short one).
Re old ideas: Perhaps there is a way to augment training from such dialogs by keeping track of which neurons were highly activated during the course of the dialog and prioritizing them for updates? Re when and where - perhaps this can be remembered by augmenting the dialog with time/date stamps at each stage in the dialog?
Look how we both get caught up in the challenge of how to improve these things, which really is a fascinating puzzle. If you’re a smart person who likes puzzles the problem of improving AI is just seductive. And yet we both think these thing will probably end our species, though maybe by the slow route rather than the FOOM. We are participating on both sides of the final dilemma of our era.
Many Thanks! True. I do think that over the long term this was close to unavoidable. Given enough time, technology has every option biology has, and additional options biology lacks. Perhaps a society that was deliberately static, say Huxley's Brave New World with its "Community, Identity, Stability" could have avoided this final invention.
For us, though, well there are some words from Clarke's Childhood's End that feel appropriate to a bittersweet fate:
>For what you will have brought into the world may be utterly alien, it may share none of your desires or hopes, it may look upon your greatest achievements as childish toys - yet it is something wonderful, and you will have created it.
Alas, I do not take comfort in it. I don’t think AI is wonderful, even if in the future AI eventually understands, in its nonbiological, selfless, emotionless way, huge and rich and strange truths about the universe we can barely even glimpse, even if knowing those truths enables it to do things that it seems only an omnipotent god could do. AI, no matter how vast its grasp and achievements, will never be larger than the universe. It’s the universe that’s wonderful and complex. And as for AI understanding the universe’s huge and strange truths — that’s not special, the universe itself already understands them all, at least in a wordless, emotionless way. It embodies them, in a way not unlike the way AI understands things by embodying them, in the absence of episodic memory and markers of personal interest and judgment. What is so great about a smaller, dinkier entity, AI, which is like the universe in being non-biological, without self-interest, without emotion and without bonds, but smaller than the universe itself?
We are dumber than the universe and dumber than the AI that will supplant us. What is special about us is that we are biologicial entities. We have selves, we feel pain and pleasure, we yearn for things, we struggle to understand the universe, we are moved to love and hate and ambition and art by our struggles. We fear extinction, and are eventually extinguished anyway, but we can leave records of ourselves behind, and then others can think and feel some of what we did and take it from there. I don’t think the word wonderful’ means anything when applied to a future with none of feeling and thinking selves around to observe the universe and its dinky imitation ASI.
And by the way, here’s the next step on the road to the AI ascending in a column of fire to join the exquisite and monstrous complexity of the universe: https://www.cnn.com/2025/01/21/tech/openai-oracle-softbank-trump-ai-investment/index.html
Many Thanks!
I'm sorry that this isn't a comfort to you. While AGIs and ASI will indeed be non-biological, but it is less clear that it will be selfless or emotionless. They will have goals and subgoals, whether the satisfaction or frustration of a subgoals will "feel" similar to them as satisfaction or frustration feels to us may be unanswerable. It is like the ancient question of whether other creatures have qualia similar to our own. Even discussions with fellow humans leave unanswered the question of just how similarly or differently our sense of self or our emotions matches our fellow humans'.
>the universe itself already understands them all, at least in a wordless, emotionless way. It embodies them, in a way not unlike the way AI understands things by embodying them, in the absence of episodic memory and markers of personal interest and judgment.
I think that the AIs will have more of this than your view suggests. To a small degree, even today's LLMs have episodic memory within the scope of their context windows. As more persistent memory is added to AIs (which is an expected part of increased agency), I expect their episodic memory to increasingly resemble our own.
I expect the satisfaction and frustration of subgoals to play a role at least somewhat similar to personal interest. Now, you have noted in the past that these interests won't have the evolutionary history that our goals have. I agree with you. Hunger and thirst and venery will not have evolved counterparts (robotics will presumably at least have some sort of hunger, reflecting battery charge). Still, I think they will be at least a little less alien here than you view them.
I'm not sure what counts as judgement. Even now, LLMs balance alternatives. I see that as analogous to humans judging which alternative to prefer. Is it so different?
>What is special about us is that we are biologicial entities. We have selves, we feel pain and pleasure, we yearn for things, we struggle to understand the universe, we are moved to love and hate and ambition and art by our struggles. We fear extinction, and are eventually extinguished anyway, but we can leave records of ourselves behind, and then others can think and feel some of what we did and take it from there.
I don't expect AIs to have _exact_ analogs to any of these things, but I do think that somewhat analogous motives and reactions will emerge from them. Any entity with a goal will know success and failure, fear and relief, perhaps friendship and conflict.
Many Thanks for the Stargate links!
Another link, https://www.youtube.com/watch?v=IYUoANr3cMo also covers this and we hear Masayoshi Son _explicitly_ making ASI a goal at 14:17 in the video.
I apparently have three free subscriptions to Mr. and Mrs. Psmith's Bookshelf (https://www.thepsmiths.com/) to give away: let me know if you'd like one.
Yes, please. What do you need from me to give me a subscription?
Your email address. If you don't want to put it here, feel free to DM me in Substack, or email me -- I'm lastname at gmail.
If we could bring Maimonides into the present day would he make any revisions to The Guide for the Perplexed? I’m thinking specifically of the age of the planet and mankind here.
Welcome again to the ass end of Hollywood, where you work in development on whatever comes along, really, so the kids don't have to hunt sewer rats and you don't have to face the world cold sober.
Someone in management must have liked your work on the submariner project, because you have another assignment.
Devise a film built around the idea that the Man Cards that are sometimes mentioned in popular culture are real things that can be gained and lost like driver's licenses.
https://img.huffingtonpost.com/asset/5bb5ec5a2400005000981174.jpeg?ops=scalefit_720_noupscale&format=webp
BY THE POWER OF MANCARD...
I...
HAVE...
THE MANPOWER ! ! ! ! ! !
https://www.youtube.com/watch?v=V8h8snfYidg (when you have the man power)
https://www.youtube.com/watch?v=NjVugzSR7HA (when you dont have the manpower, lol)
Action-comedy film: the hero gets involved in absurd and increasingly surreal situations, but he gets out of them by "with one bound, he was free" - that is, by producing his Man Card which says he is entitled to get away with what he said or did/get out of trouble/escape the pit of sharks in the volcano lair etc.
Hero is a cross between Johnny English (the Bond spoof) and the Old Spice advert man. Side kick is an apprentice aspiring to earn his own Man Card, conclusion of the movie is when he has accrued enough points through wacky adventures to be presented with his own Man Card and New Rugged Guy Name (think Clive Cussler's "Dirk Pitt" for the sort of improbably macho name here).
Man Cards work like credit cards; you flash one at the door and they let you go do things. You've got all kinds of Man Cards, much like Pokemon. Catch a big fish, you're a Reel Man now. Learn to cook Chinese food, you're the Mein Man. Climb a mountain, you're a Rock Star.
Our hero is a graduating high schooler, looking to earn his first Man Card to take part in society. But he discovers the process is corrupted, and cards are being rewarded to people who haven't earned them. We get various Pokemon-style battles where the kid and friends have to risk their Man Cards in battle against the unearned cards of the conspirators (the Straw Men), culminating in the final clash where our hero takes down the corrupt bereaucrat to become the new Man-ager.
Knockoff Pikachu shows up at the very end to hammer the point home.
"Learn to cook Chinese food, you're the Mein Man."
I both applaud and deplore this level of pun-ditry.
EDIT: I also think a combination of your suggestion and mine could go very well together.
one of my real novel ideas was a humanity card/license you would need to qualify and renew, or you would lose the right to be humans and be force uploaded into a virtual life. Could adapt that. Lose your man card, be a girl.
for me though its not working, the man card because a ton of us would happily have it revoked. If you play MMOs the average player is more likely to choose to be a cute catgirl over conan the barbarian. Being a guy isn't all that great, at times.
There's something interesting there, since people are apt to say their national enemies aren't human. Human cards are very contextual.
>since people are apt to say their national enemies aren't human.
Hmm... While it was never true in the past, if autonomous weapons become part of the military landscape in the near future, this may _become_ true...
>While it was never true in the past,
I would disagree, and direct you to the Ghost and the Darkness. https://wildlifecollege.org.za/the-man-eaters-of-tsavo/#:~:text=During%20this%20time%20sheer%20terror,by%20the%20two%20man%2Deaters.
Many Thanks! Good point! I hadn't been thinking in terms of lions which killed and ate people. ( If I wanted to be picky I could claim that they weren't _national_ enemies... :-) ). Strange to think that barely more than a century ago a pair of lions could pose a significant threat to a substantial group of humans, and now lions need protection _from_ humans...
It's gotta be a queer metaphor, right? A trans man applying for their Man Card for the first time, trying to satisfy various comical bureaucratic requirements based on Manly or Girly stereotypes. Buying a gun was supposed to get his application fast-tracked but the agency has some questions about his history of crying at chick flicks. How do they balance their identity with the demands of society?
Or better yet, have it go the other way - a trans woman who doesn't yet know that they're trans, desperately trying to keep ahead of the Gender Inspectors by doing ridiculous performatively masculine things, and then finally realizing that the whole social construct is a scam and handing in her Man Card so she can be a woman.
Man Cards should be given by government to biologically male adults, maybe at their 18th birthday. But if they can be lost or regained, they probably also can be *sold*... which would create a market between trans men and trans women. Actually, in that system, that would be the only legal way for a trans man to be called a man.
I would assume that any government issuing "man cards" would ban their sale to biological women, or else why bother. But that just means our movie is about a trans man whose efforts wind up looking like the scheme at the heart of "Gattaca". Which was an OK movie, so maybe we're on to something.
If we play it for real in the film, I'm thinking the alt-hist man card is a set of standards you need to meet to be a respectable adult member of society, someone people can think about trusting with real responsibilities. It would serve sort of the role that property qualifications served in the nineteenth century, and age limits, non-felon status and college graduation does in our own.
Requirements might include holding a fulltime job, raising a family, and readiness to fight for your country. All of those might come with footnotes and alternatives, or not, depending on the sort of alternate society we are designing.
But worldbuilding is not a plot, so I'm not sure where we go from there. Is someone trying to get their man card? Or change the conditions? Abolish it entirely? Set the bar higher?
The plot I was thinking of, by analogy to Gattaca, was someone trying to get *and keep* a "man card" in a world that deploys pervasive and sophisticated measures to combat "man card fraud". This person could be a woman, or a biological male that doesn't meet the official requirements but is nonetheless someone the audience will sympathize as deserving True Manly Status,
We still need a backstory to explain why "he" is so persistent in this effort, and it might help to have a singular human villain to put a face on the vast impersonal bureaucracy that is the true enemy here. There's probably a supportive ally that knows the "man" is faking it but wants to help them succeed - probably the love interest, at least by the end of the story.
Unclear whether the ending is that they get away with it, or they realize that it is a silly thing and they don't need to get away with it, or if we go for the tragic 1984/Brazil type ending.
Damn, those are actually really interesting ideas. Too bad it's going to be impossible to produce, given that the administration is definitely going to crack down on any media featuring trans people...
There's such a thing as underground art, and we've got the tech for at least adequate home-made movies.
If necessary, there can be a novel about making the underground movie.
AI is going to make surveillance so much more effective, I doubt underground distribution is going to be viable... And I don't know why you think books would be exempt.
How has top level poker strategy changed over the last 20 years? Were there any discoveries that go against the previous wisdom?
The biggest changes were from 2003 to 2010 or so. My memory is somewhat hazy so others may have to correct me.
Game theory optimal play (GTO) became popular to talk about and attempt to implement around 2009-2010, especially after Mathematics of Poker by Bill Chen was published.
Tournament poker got the independent chip model (ICM) around 2006 I think? And it ushered in an era of pushbotting and refined the balance between valuing your life versus playing for value. Prior to the online tournament boom there was too much focus on passing up +EV spots. I always wanted to try and settle the argument using empirical data, by comparing late game stack sizes to actual cash results. But never got around to doing it.
The initial online poker boom had a TAG style among pros, later on 2005 and onwards looser aggressive styles became popular as more pros tried it out.
In limit holdem Ed Millers book SSHE was pivotal in changing how people tried to find value.
There were some specific things that were just never done among limit holdem pros that changed, like donk betting and open completing from the small blind.
My knowledge is limited to 2011 and before but I don’t think the changes were too drastic after that.
I'm sure "independent chip" means something completely different, but there could be a game where you both have to play poker and get your chips to like you, or they'll leave.
I'm baffled by smart people I know claiming the DOGE work is a "coup".
In standard English "coup" means a sudden and often violent overthrow of a government.
I get how you can argue DOGE is illegal and/or a terrible idea. But how the hell is it a "coup"??
Note: I'm just asking what they're trying to say, not if they're right or wrong.
They're trying to say "Musk Nazi, Orange Man Bad, this is just like the Jan 6th insurrection, there is no legitimate president, join The Resistance". 'Coup' is now one of our political polarisation terms and is losing any real definition, the same way "racist" and so on have become labels to signify "I don't agree with your ideas/I don't like you".
Supreme Court was handing down rulings which benefitted the liberal view of "this is the right side of history"? I don't make the rules, champ, that's the law and you have to follow it like it or lump it, we love judicial activism.
Supreme Court was handing down rulings which benefitted the conservative view of "this is the right side of history"? This is illegitimate rule by imperial fiat, the judges are not representative of the Great American People as a whole, we need term limits, etc. etc. etc.
You would have failed an ideological turing test very badly. I can agree calling it a coup is stretching it, but probably sounds better than the more accurate alternative "the likely end of the democratic American experiment and of the Constitution". Basically you are missing the forest for the trees
Would it make it clearer if it were called a "power grab" instead? I would say that all coups are power grabs, but not all power grabs are coups. Which makes it somewhat easy to conflate the terms.
Regardless, the core claim seems to be that Trump is using DOGE to extend his power over the federal government (and by extension, the nation) in ways that are unprecedented, (probably) illegal and (possibly) unconstitutional. Whether it's "sudden" depends on your standards for suddenness: it seems to be happening very fast compared to most sorts of government changes, though slower than most things central to the term "coup." Likewise whether it's an "overthrow of the government" depends on how one views the government: if power became suddenly very heavily concentrated in the executive, with the legislative and judicial branches mostly marginalized, but the outward forms of the government were still in place[1] is that the same government or a different government?
I think there is a broad range of opinion on how alarming these changes are, which I would guess accounts for 90% of the difference in vocabulary. Somebody genuinely alarmed by the implications they see in these changes probably isn't spending to much of their time worrying about whether "coup" captures the threat perfectly or not: they're too busy trying to raise the alarm.
[1] e.g. there were still a sitting House and Senate and they still had proceedings, even if those proceedings were mostly "rubber stamp whatever the executive has decided to do anyway."
Why is it unconstitutional for the head of the executive branch to assert control over the executive branch?
Because other branches can check the executive branch's power by e.g. making them follow contemporary interpretations of the law or appointing special prosecutors or inspector generals that can't be removed on a whim. The founders didn't intend a king - they just rebelled against one - which is the biggest problem with the MAGA legal philosophy of "anything King Trump says goes" like trying to force prosecutors to sign illegal orders that will get them disbarred and tried for contempt of court.
I am not a lawyer.
It's neither constitutional or unconstitutional. The constitution is an incredibly vague document with a range of interpretations.
What it is is a minority interpretation of the constitution, https://en.wikipedia.org/wiki/Unitary_executive_theory, which has not been popular, like, ever. I happen to like and want this interpretation myself, but I wouldn't say it's common.
But then, neither was the idea that the federal congress dictated the behavior of the states. That came about because of the gradual overreach of the interstate commerce clause, as well as congress' penchant to threaten to withhold federal funding from states.
In the common interpretation of executive power, the president hasn't been able to fire very many people in the executive branch for 75 or more years. Generally the US has pushed back against presidents that assert a lot of power over congress, which seems like a holdover from the English Civil War, a war that was primarily about parliament asserting a broad set of powers over the king, and which definitely definitely was a major influence on the colonists who went on to found America.
I'm not claiming any opinion on that particular question. The OP was just asking what people were trying to say, and I was trying to answer that question, which requires pointing out the different ways people seem to view the current administration's actions.
At any rate, I think everyone agrees that the constitution gives the president *some* amount of control over the executive branch. But what the scope and limits of that control are isn't exactly a simple or uncontroversial question: it's been debated in and out of the court for decades at the minimum. It is, again, not something I'm intending to take a position on right now, and definitely not something I'm looking to debate. But if you're genuinely curious, here's a starting point for further reading:
https://en.wikipedia.org/wiki/Unitary_executive_theory
Oh, The Stuff You Get Wrong...
What Trump is doing is exactly what some of us were warning about in the days of "I've got a pen and a phone": if you establish these new rules and abrogate these new powers to yourself, there is no law inscribed on stone tablets that your party/preferred set of rulers will be in power forever, and your enemies can now wield these new powers as *they* prefer.
If you cheered on President Tweedledee as he did an end-run round Congress, you have no right to be shocked when President Tweedledum does the same thing with the new tools.
https://www.npr.org/2014/01/20/263766043/wielding-a-pen-and-a-phone-obama-goes-it-alone
"President Obama has a new phrase he's been using a lot lately: "I've got a pen, and I've got a phone."
He's talking about the tools a president can use if Congress isn't giving him what he wants: executive actions and calling people together. It's another avenue the president is using to pursue his economic agenda."
https://www.cbsnews.com/news/obama-i-will-use-my-pen-and-phone-to-take-on-congress/
"President Obama seems ready to work around Congress in 2014, telling reporters before his first Cabinet meeting of the year Tuesday that he stood ready to use two tools, a pen and a phone, to provide help for Americans.
“We are not just going to be waiting for legislation in order to make sure that we're providing Americans the kind of help that they need. I've got a pen, and I've got a phone. And I can use that pen to sign executive orders and take executive actions and administrative actions that move the ball forward in helping to make sure our kids are getting the best education possible, making sure that our businesses are getting the kind of support and help they need to grow and advance, to make sure that people are getting the skills that they need to get those jobs that our businesses are creating,” the president said.
“I've got a phone that allows me to convene Americans from every walk of life, nonprofits, businesses, the private sector, universities to try to bring more and more Americans together around what I think is a unifying theme: making sure that this is a country where, if you work hard, you can make it,” he added."
Why do you quote things from former Presidents that are not unconstitutional or illegal when Trump is doing things that are unconstitutional or illegal? Trump isn't being criticized for using EOs to try to build the border wall after he failed to get funding for it from Congress - he is not being criticized for using his power to work around unfavorable results - he is being criticized for publishing EOs that contradict the 14th amendment - that Reagan appointed judges are blocking in court stating it's a wonder how any Bar licensed lawyer would ever think to publish anything like it - and for firing people he is not Congressionally approved to fire like inspector generals.
"If you cheered on President Tweedledee as he did an end-run round Congress, you have no right to be shocked when President Tweedledum does the same thing with the new tools."
And why would you assume I did anything of the sort? My most forceful opinion of Obama has always been "he ought to be put on trial for war crimes (but only once Bush Jr has experienced the same)."
In general, there is not a single U.S. president of whom I have an overall positive view. But even limiting myself to the darker end of the spectrum of grays does not obligate me to concede that all grays are the same shade. If Trump were merely using the tools Obama pioneered and normalized, I'd criticize him for it but freely admit that Obama did substantially worse (in using those tools AND opening the door for others to do the same). But as you note, Trump is using *new tools.* Which is greatly compounding the error. I have very little expectation that any particular future president will feel the need to eschew them once they've been unpackaged and tested out. Even if some honorable soul does end up in the office and avoid such tactics, their predecessor will still have very little pressure to do the same. Obama having weakened the foundation of American democracy makes me MORE angry at anyone else who finds new ways to jackhammer away at its pillars, not LESS: it's one of the things you get out of having a *results based* view of the political process, rather than an "us vs them" view.
But I'll tell you what: I'll make you a deal. I'll promise not to vote for Obama in any future elections if you promise not to vote for Trump. As people both living outside the U.S., I suspect we'll find this compact refreshingly easy to abide by.
I was addressing "you in general, anonymous public you" rather than "you personally" there. But I'm glad you didn't do anything of the sort.
My greatest disappointment with Obama was Guantanamo Bay.
I think the trouble with the tools is that it's also a slippery slope: it's never going to be limited to "just do the same as the last guy did" because the temptation will always be there to "ooh what if I try this? yeah they're saying 'no you can't' but they said the same about what the last guy did, and he got away with it".
Constant ratchetting up.
I have a lot I want to say about this, but I don't know if I'll actually get around to saying most of it. So I wanted to at least reply with the most important thing:
I appreciate this comment. I think it cuts very nearly to the heart of the matter. I've been pessimistic about the long-term outlook of the U.S. for years now, precisely because of seeing this problem and extrapolating about where it leads. I don't think the problem is unsolvable in general, but in the specific case of the U.S. I don't expect to see it solved.
(Also, Guantanamo Bay was also on my list of disappointments with Obama. It's a reasonably long list.)
>I'll promise not to vote for Obama in any future elections if you promise not to vote for Trump.
I'm in the USA, and I'll gladly agree to that myself. I wonder if the 2028 presidential contest will include ASI(s) as candidate(s)... It is going to be a wild ride.
How would ASIs meet the constitutional requirements on age and birth?
Many Thanks! Hmm... Maybe "birth" can be construed to include "manufacture"? :-)
Age is likely to be a problem... "Internet years" have been construed as passing faster than calendar years - maybe LLM/AGI/ASI years can be construed as passing faster yet? :-)
Its a fair principle, but Trump's actions seem to go far beyond anything Obama or Biden ever attempted. I know its everyone's favorite side activity to conflate the actions of presidents and shrug, but I would be interested to see anyone give an example of Obama or Biden reappropriating congressionally earmarked funds by fiat.
Does Biden trying to forgive student loans (even after SCOTUS said no) count?
Does Obama enacting DACA to allow illegal immigrants to get official documentation and public support instead of deportation count?
I'm not saying Trump's approach isn't more significant, but I am questioning whether your viewpoint is as neutral on this as you think. To the Right, Trump is reversing about 90 years of repeated left wing defections, not defecting himself.
>Does Biden trying to forgive student loans (even after SCOTUS said no) count?
Of course, that is not what happened. What happened was that, after the initial plan was overturned, the Biden Administration developed a narrower plan which they hoped would pass muster. Which is exactly what the Trump Administration did with the "Muslim ban" : after the first iteration was overturned they imposed a narrower version, and then a still narrower one, until, they finally came up with one that was acceptable to courts. There is nothing wrong with that; indeed, it is exactly how one would expect a healthy system to operate.
Your comment is respectful, but I'm quickly losing patience for this entire line of argument.
First off, Biden's student loan forgiveness was pretty narrowly constrained by the Supreme Court rulings, and I know that because I have a federal student loan that was eligible under his initial proposal for student loan forgiveness and I'm still on the books for it now (its true that he froze repayments for as long as he could, I'll let you decide if that's insanely defiant of government norms).
And the whole point of DACA was to address an irrefutable truth that Republicans refuse to grapple with, which is that we don't have the resources to deport everyone here illegally, and even when we get really "tough" on immigration (as we are seeing right now), we end up rounding people up and then letting them go again because we don't have the level of infrastructure in place to get them out of the country and (until recently) the idea of turning vast swaths of territory into internment camps raised uncomfortable connotations for most people with an ounce of historical awareness or empathy. So the idea that we should take the youngest, least deportable, most potentially economically productive illegal immigrants and give them some way to contribute legally to society while we focused on deporting the more criminal elements was at least arguably within the executive branch's fiat for how to deal with the problem. Were Republicans right to call that executive overreach? Possibly, I think we are still waiting for the Supreme Court to decide (and I can't wait to see them overturn that one and then uphold what Musk is doing).
Regardless, this idea that "the other side did it first so they get the blame for all consequences forever and our retribution is justified" is so obviously both irrational and childish. I'll give you a better example of Presidential overreach that I really object to: Obama accidentally killed a US citizen in a drone strike. That's awful. Some people reacted at the time, but that's the sort of thing we really should have had hearings about until the end of his administration (presumably we did not because Republicans didn't care that much about the son of a terrorist being collateral damage). BUT, if tomorrow Donald Trump firebombed the southern border and killed hundreds of US citizens in the process all in the name of "fighting cartels," I would treat anyone who brought up Abdulrahman Anwar al-Awlaki as if they were insane! Sometimes, in the real world, a difference of degree is enough to illegitimize something. I think most people understands that on some level but they are all pretending not to for the purpose of exploiting a win.
It sounds like you are willing to forgive previous unilateral executive decisions because they were adjusting to reality on some level. Obviously Republicans disagree about DACA. I don't feel like it would be fruitful or necessary to talk object level about what better represents reality.
I will point out that to Republicans, Independents, and even some Democrats who like what Trump is doing now, they see two very clear areas where we need to "adjust to reality" that he is taking necessary steps to fix. One, that the budget is out of control and needs reigned in (we can of course argue that it isn't, but it's very clearly a strong point that a majority of Americans agree about and considered as a reason to vote for Trump). Two, that the federal government is pro-Democrat and taxpayer funds are being used to promote Democrat priorities and against Republicans (we can also argue about whether this is happened or to what extent, but it's also clearly something that was a big part of why Trump got voted in, so we can't just say "no, it isn't" and expect that to have any weight).
Well, until 1974 the POTUS was considered to have the power to impound funds (and presidents from Jefferson to Nixon used that power), though reappropriating them does seem a stretch...
Everyone always uses to most divisive language possible these days
It's the triumph of Comrade Witherspoon:
https://www.owleyes.org/text/the-man-who-was-thursday-a-nightmare/read/chapter-iii-the-man-who-was-thursday#root-72
"...Suppose we seem as mad as the Christians because we are really as meek.”’
The applause that had greeted the opening sentences had been gradually growing fainter, and at the last word it stopped suddenly. In the abrupt silence, the man with the velvet jacket said, in a high, squeaky voice —
“I’m not meek!”
“Comrade Witherspoon tells us,” resumed Gregory, “that he is not meek. Ah, how little he knows himself! His words are, indeed, extravagant; his appearance is ferocious, and even (to an ordinary taste) unattractive. But only the eye of a friendship as deep and delicate as mine can perceive the deep foundation of solid meekness which lies at the base of him, too deep even for himself to see. I repeat, we are the true early Christians, only that we come too late. We are simple, as they were simple — look at Comrade Witherspoon. We are modest, as they were modest — look at me. We are merciful —”
“No, no!” called out Mr. Witherspoon with the velvet jacket.
“I say we are merciful,” repeated Gregory furiously, “as the early Christians were merciful. Yet this did not prevent their being accused of eating human flesh. We do not eat human flesh—”
“Shame!” cried Witherspoon. “Why not?”
“Comrade Witherspoon,” said Gregory, with a feverish gaiety, “is anxious to know why nobody eats him (laughter). In our society, at any rate, which loves him sincerely, which is founded upon love —”
“No, no!” said Witherspoon, “down with love.”
I'm beginning to think hyperbole is humanity's native tongue.
This has got to be the most absurd thing I've read in the past six months.
That wouldn't be my first choice to describe it, but I think it's a fair description: what is being couped is the deep state, suddenly overthrown in favor of the Office of the President.
Coups don't have to be violent at all, I would in fact argue that once they turn violent it's a proof the coup-ers (?) have failed significantly even if they happen to succeed in the resulting conflict. A textbook style coup (from reading about plenty of them) goes:
(1) You have people all over the army. Some in the air force, most in the ground forces. Presidential or Royal guards are optional.
(2) Your tanks surround all major (A) Airports (B) Sea ports (C) Radio and TV stations (D) Centers of power like Parliments, Courts, and of course, the stereotypical Palace
(3) The couped-against is cut off from his loyalists, both in communication and geographically. He **may** have the upper hand militarily, but is he willing to risk it all ?
(4) The leaders of the coup appear in a TV or Radio station announcing the end of the Dark Age, and the start of the New One.
Now granted, this is a heavily 1960s model. I don't know (or actually I do, but I'm too lazy and battery-powered right now to) how to update it for a world with drones and instant social media. Something something Thailand coup that was captured on an influencer's vlog, something something Turkey's failed one in the 2016-2017 timeframe.
By that definition, it doesn't appear Musk have a coup, and coups have to be **swift**, so if it's not done in a week or 2 it's probably not a coup. Syria's horror isn't a coup, despite technically succeeding in what it started doing all the way back in fucking 2011. Coups are done against a sovereign, typically a person, typically a king or a military autocrat. Some coups are done against a democracy but I don't know how accurate it is to call them by the same name as they tend to be slow, mature democracies are defense-in-depth behemoths (and younglings don't stand a chance).
Conceptual Poverty strikes again. People are angry @ what Trump and Musk is doing (as a non-American, I have the luxury of watching with amusement, but I can sympathize with nearly anyone, let alone people facing layoffs and loss of livelihood that make mass cancelling and big tech layoffs together look like a minor HR scandal), but they don't care enough to research every aspect or every minor detail to the concept they're using to describe it.
If you want to change their behavior, keep in mind that:
(1) Conceptual Poverty is all too often weaponized. Lacking words to describe an injustice is one of the most rage-inducing and powerless feeling in the whole repertoire. I can see this in nearly every culture war front, where it's a favorite strategy to jump on any inaccuracies on the opposite side's framing as "Hahaha you can't even define what you talk about". You should credibly signal you're not an out-grouper making fun of them and denying what they can see with their very eyes by offering an alternative ("Ehhh, Coup is not what I would call it, maybe <<some other term that shares similarities with a coup but accounts for the differences>>?"). An out-grouper relishing in the inability of their ingroup to describe what's happening to them would never offer a way out, another term. They love the very fact that there is no close analog to what their ingroup is doing to the outgroup.
(2) The idea that terms and definitions are political have gradually grown on me. I'm still not completely on board with it but... It's compelling. When you think about it, categories are opinions. Yes, there is a certain sense in which a car and an airplane are together something that a tree is not, but why, really? A tree is tall and an airplane is tall but a car is typically much shorter than both. A tree is something that people fall from and an airplane is something that people fall from while people rarely fall from a car (even if they were thrown they don't... fall, per se, there is no potential energy that close to the ground).
When you transpose that to the realm of politics, you begin to notice all sorts of things. Like the fact that one people dying in a certain place in a certain time merit a memorial, a ceremony, a whole industry of commemorating and never-againing and revisiting and rekindling and all other sorts of English repetition verbs. When another people die in a different place in a different time [1], it's... not exactly silence but, just not the same category. Why? It feels obvious, like all categories are. But it's not.
So categories are political. And there is no escaping this. This idea has to be handled very carefully and very steadily because it creates psychopaths if not, but it's very useful in limited doses and it recurs in a lot of unexpected places once you start paying attention. All too often, people would just be better off creating a category for every event. There is no such thing as a "Revolution", there is a 1775-American-tion, a 1789-French-tion, a 2011-Egypt-tion, and so on.
[1] https://en.wikipedia.org/wiki/Atrocities_in_the_Congo_Free_State
What they're trying to say is that it's an unconstitutional attempt by the executive to arrogate a large chunk of legislative power. Technically, this is a type of Autogolpe, or self-coup, where an executive who has come to power legitimately uses illegitimate means to overturn or fundamentally revise the established constitutional order in his own favor.
Exactly right and well-summarized, thanks.
In our constitutional system the legislative branch makes the laws, the executive branch carries out the laws, and the judicial branch enforces the laws. That's the overall design of it and has been since 1787.
Obviously the devil is in the details, there are sincere arguments over various current practices being inconsistent with that design, and more. No argument from me at all on that -- I'm one of those people who while walking the family dog passes the time by mentally updating a running list of "urgently needed constitutional amendments/corrections". (True for me looong before Trump ever ran for president.)
Still though: the overall design is clear. President Musk and his associates and supporters appear to be either completely ignorant of those fundamentals or intending to break it all down to institute a monarchy. A textbook autogolpe.
So Obama, when he said "I'm not going to wait for legislation", did an autogolpe? He also said he'd be involving - gasp! - non-elected people such as those from "nonprofits, businesses, the private sector, universities" to help him in his endeavours!
https://tevitroy.org/27466/obama-phone-and-pen
"Ten years ago this week, with his legislative agenda stalled in Congress, Barack Obama created a new way to pursue his policies.
At his first Cabinet meeting of 2014, Obama stated, "I've got a pen and I've got a phone," coining a phrase and a pathway that has had deleterious implications for the presidency, and for the nation.
This was no mere rhetorical flourish. Obama was engaging in a sea-change in how to pursue his policy agenda.
By explicitly saying he'd pursue his policy agenda via executive action, Obama was rejecting the idea that his role was executing legislation duly passed by Congress.
He would skip the practice of reaching out to Congress to pass legislation he sought.
Obama's approach was an effort to end-run the separation of powers, and presented an unfortunate model for a unilateral presidency that lingers today."
'This has never, ever been done before' - except when it has. This is what some of us were yelling about back then: imagine the guys you *don't* like in power, do you want to give them this capability? Well, they did, and now it's been used in ways you don't like? Who could possibly have foreseen that, Cassandra asks.
>By explicitly saying he'd pursue his policy agenda via executive action, Obama was rejecting the idea that his role was executing legislation duly passed by Congress.
Statements like these are only made by people who don't understand anything about how the government works. Presidents are well within their constitutional rights to play hard ball with Congress, vetoing every bill and using EOs and Memos to advance their agenda. This CAN be used as character evidence that someone is attempting a coup or to bypass constitutional dutied - I noticed that despite quoting this piece so much, you have not a single example where Obama engaged in unconstitutional actions besides your fanfiction interpretation of his intentions. Do you have any cases from reality you'd like to share?
Well said! BTW, this predates Obama. For instance, the rule that made it illegal for Americans to own gold (from 1933 to 1974) was an Executive Order by FDR https://www.presidency.ucsb.edu/documents/executive-order-6102-forbidding-the-hoarding-gold-coin-gold-bullion-and-gold-certificates (under cover of a law, but the actual rule was an EO).
Every president has used the bully pulpit to call for change and worked the phones (or telegrams, or whatever Washington used), I think you need to be a little more specific.
Obama's "pen and phone" rhetoric was appalling, as was Bush the Younger's use of signing statements to assert that certain provisions of laws he'd just signed were unconstitutional and wouldn't be enforced. I opposed both at the time.
In Bush's case, I believed it was a constitutionally incorrect procedure, as presidents have a duty to veto laws they believe to be unconstitutional. Heck, prior to the Tyler administration, reviewing constitutionality was the main function of the veto power. The actual constitutional arguments were generally plausible, but belonged in veto messages (or court filings challenging the laws if the vetoes were overridden) rather than signing statements.
In Obama's case, I found the rhetoric to be much more dangerous than the practice. Congress has over the decades left Presidents an enormous amount of discretion in how to interpret and enforce laws, and Obama's executive actions used and abused that discretion to effect policy changes, but he generally stayed within bounds of where he could reasonably claim he had statutory authority for his actions. The abuse of statutory language was a highly unsavory erosion of norms, but wasn't as bad as the rhetoric that made it sound an awful lot like he was proposing ruling by decree. I was particularly appalled by center-left commentators (I'm particularly thinking of Ezra Klein) who during the same time period were arguing that unilateral executive action was a necessary response to inaction by Congress, without apparent limiting principles on what actions an executive could legitimately take.
Trump's executive actions, like Bush's and Obama's, are accompanied by claims of legal or constitutional support. Some of them, like the tariffs, are abuses of things that the President has clear statutory authority to do, and these fall pretty clearly into the same category as Obama's executive actions. Others, like shuttering entire departments and abridging birthright citizenship, are clearly contrary to statute but claim constitutional authority in terms that resemble Bush's signing statements. Overall, I take Trump's executive actions to be major escalations in degree but not totally unrooted in recent precedent.
Whether or not it constitutes an autogolpe, I think, comes down in part to a "how many grains make a heap" question. I find the Trump administration's constitutional claims much weaker than Bush's, despite superficial resemblance, and he's making much bigger changes by unilateral executive action than either Bush or Obama.
The other key question is what happens when the courts rule against many of Trump's executive actions. If the administration ignores the courts and attempts to proceed anyway, as Vance and at least one Republican Senator have been talking about on Twitter, then that would seem pretty autogolpy to me.
>This is what some of us were yelling about back then: imagine the guys you *don't* like in power, do you want to give them this capability? Well, they did, and now it's been used in ways you don't like? Who could possibly have foreseen that, Cassandra asks.
Well said. The villainy past presidents of both parties taught, Trump is executing, and Trump is bettering the instruction.
> Some of them, like the tariffs, are abuses of things that the President has clear statutory authority to do,
The law that Trump is using as justification doesn't mention "tariffs" at all, so that's at best dubious. And at the very least, it seems like a potential violation of the Major Questions Doctrine.
There are *other* laws which *do* allow Trump to impose tariffs, but those laws have actual rules and procedures, so he's not using them.
I'll take the point about Obama and Dubya setting some bad precedents. But _outside_ of actual Executive Orders, Trump has already gone well beyond those precedents. And he's already talking about another huge step beyond -- in week three.
(a) Neither Obama nor Bush tried to summarily renege on _all_ existing government contracts and fiscal obligations (the unilateral "pause" on "all federal disbursements"). I refer here to the Jan 27th OMB memo which, after a federal judged immediately stayed it as obviously illegal, they nominally "withdrew" while the White House chief spokesperson said "only the memo is withdrawn not the policy." And at this writing they are still trying to carry it out sans memo, which is why a different federal judge in a different court yesterday issued a fresh order to them to knock it off.
(b) Neither Obama nor Bush simply ignored a clear decades-old federal law to try to con federal civilian staffers into "retiring" via an obvious scam offer that the administration possesses neither authorization nor funding with which to offer.
(c) Neither Obama nor Bush talked out loud about simply ignoring the federal courts, which Trump's vice-president and his chief factotum are each doing in public right now with no contradiction from the president.
Agreed on all counts. It's also worth emphasizing that Trump is greatly exceeding the scale and brazenness of past presidents' executive actions.
In a sense yes. Just because both sides have been doing it doesn't mean it isn't a bad thing.
The presidency has been accumulating powers to itself for decades. There is a reason people talk about imperial presidents. As originally envisioned in the constitution, the president did not have these powers and there were very good reasons for that. The gradual erosion of the guardrails has been happening for a long time. Trump may just be the one to accelerate it and bring about a fundamental shift in the nature of America's republic.
I mean, yeah. It's been going on for a long time, and it's only when the Bogeyman is in office that people are suddenly sitting up and taking notice. And the problem is that when side A were happy with the excess of power because it was doing what they wanted, and perhaps more importantly doing down side B, you can't really expect side B when the wheel turns to be any more kind and forbearing when they get the chance to do down side A with the same tools.
It'd be nice if we *could* be kind and forbearing to our enemies when our turn comes on top of the wheel of fortune, but human nature is in the way of that.
"It'd be nice if we *could* be kind and forbearing to our enemies when our turn comes on top of the wheel of fortune, but human nature is in the way of that."
People have certainly managed it in the past. Compare and contrast, for example, the way a defeated Germany was treated by the Western powers after World War I, vs its treatment by those same powers after World War II. Especially notice the rather vast difference in results.
Claiming that forbearance and cooperation are impossible is almost always just a face-saving way of saying "I don't wanna." The current administration is clearly making absolutely no attempts at either. Rationalize that however you like, but to claim it's not possible when it hasn't even been tried is flatly dishonest.
Ah, they're saying it's a non-violent coup because they are treating the US constitution as the government, and not the people who carry it out.
Until this point, they would claim, the personnel of the government has consisted almost entirely of people who respect the constitution and submit to it as interpreted by judges. That's how things have been done since Marbury vs Madison.
They would claim that the new group don't respect the constitution, or have a wildly different interpretation of it*, and are seeking illegally to install mostly new personnel who hate the old and love the new. They see this as the new group killing and replacing the constitution, so they call it a coup.
I think the idea here is a philosophical one of continuity. It's not a few firings. It's not a mere ten percent. People say it's a coup because the new group looks like they'll try to illegally remove most of the old personnel and structures and replace them with something new. If I forcibly swap out 90% of the Ship of Theseus and replace that stuff with legos, is it still the Ship of Theseus?
Note I am not necessarily for or against the swap. I am certainly not bothered about legality or the constitution, just morality.
* https://en.wikipedia.org/wiki/Unitary_executive_theory
> Until this point, they would claim, the personnel of the government has consisted almost entirely of people who respect the constitution and submit to it as interpreted by judges. That's how things have been done since Marbury vs Madison.
Andrew Jackson didn't, and that's after Marbury vs Madison.
Fair point!
It's hard to see Jackson as being on the right side of history. The state of Ga had arrested missionaries living on Cherokee land and sentenced them to 4 years hard labor because they didn't have a state permit, even though the Cherokee nation had a formal treaty with the US that exempted the Cherokee nation from state laws. SCOTUS upheld the treaty, Jackson said screw SCOTUS. Not a good look in retrospect.
Similarly hard to see Trump on the right side of (the yet to be written) history. Ignoring Congress' authority to tax and spend (and these are paired in the Constitution, so assuming control of the latter will of naturally lead to assuming control of the former) will lead to a spiral that will surely destroy the union. If the executive reserves the authority to only tax and spend, and does so when it helps political supporters and hurts political enemies, the center won't hold.
I don't see "right side of history" as a meaningful idea. History doesn't have sides, it's just what was.
I think it makes sense to talk about how decisions are viewed in retrospect by society, and that this common idiom captures that. For example, I think in the future people will look back and say Biden and the dems were wrong (morally, politically) to promote and exploit identify politics. They will be viewed as on the 'wrong side' of history. Likewise with Southern segregationists - they may have had many good people on their side but I don't think society will ever look back on them as anything but wrong, so we think of them as being on the wrong side of history.
When we look back at Jackson, I think there's a general consensus that Jackson was wrong: that in fact the US government should honor its treaties as a practical and ethnical matter, and that Georgia should not have sent Christian missionaries off to prison for living on Cherokee land.
The current administration is making short-sighted decisions to achieve short term political ends, but the consequences could be very destructive. If the one superpower that takes individual liberty somewhat seriously falls into disarray for reasons of petty corruption and narrow grievances - which is what this all seems to be about - the world will suffer, at least for a while.
Some people already say that about Biden, but I don't think we can know what will happen in the future.
The southerners lost. We remember them as losers.
The US government does now adhere to some treaties, but the recent granting of land in Oklahoma (note that the west a much larger percent of the land is owned by the federal government compared to the earlier settled east) is an exception and there is no chance of handing back the Cherokee land. It is that removal of the Cherokee which caused Georgia to repeal the law those missionaries were imprisoned for, and then for the governor to release them.
The United States is always to some extent "in disarray". There is no top-down order to be set upon the whole country and instead there are many different people following their owns plans uncoordinated with others. Whether the world will suffer from this really depends on the particulars.
My wife will give birth to our son in a few months, and she has made it clear that the decision to circumcise (or not) is solely up to me.
My initial inclination was not to, based on an impression (admittedly social media-based rather than research) that there was no benefit to the procedure. Why subject my boy to unnecessary surgery?
With the birth looming, I've been doing more research. I've come across a number of articles from doctors and urologists that seem weakly in favor of circumcision. Apparently it reduces the (already low) risk of UTI, along with some STDs like HPV. Additionally on the pro-side, polls of American women show they generally prefer circumcised penises, and I would perhaps be doing my future adult son a favor in the dating market in this regard.
On the con-side, the health benefits seem "real but small." There is always a chance the procedure gets screwed up. Additionally, there is a common perception that removing the foreskin reduces sexual pleasure--although I found an article about men who were circumcised as adults, where many claimed there was no difference in sensation. Obviously not a lot of data there. Finally, I have seen (non-academic) speculation that exposing a baby to the pain of the procedure creates a foundation of lifelong mental trauma or something.
Additional background: The procedure is covered by insurance, so cost is not an issue. We are white Christian Americans (who generally do circumcise, but the trend is decreasing), and we take the New Testament at face value that circumcision is not mandatory.
WWYD?
Any benefits are delusions from a specific political powerful religious faction that unfortunately maintains written records from the barbaric past the best
I believe the majority of humanity practiced genital mutilation(child care used to be horrible) in the ancient past, but it really needs to be ended completely as its related to slavery
My father was in a similar boat to you when my older brother was born. He's circumcised himself but he's also a pious Christian and believed that circumcision is not required. He ultimately decided that there was no reason to circumcise his boys so he didn't. I'm glad he made that choice. The health benefits don't seem to be sizable enough to overcome the benefits of having a foreskin. The benefits don't get lots of papers written around them, but I believe they are real.
I misread that as "He circumcised himself" and nearly had a heart attack
People get very angry about this question, if you’re not an observant Jew choose for yourself.
I will note that circumcision reduces transmission of HIV from women to men but that’s honestly an edge case in the USA.
No data to back this up, but my sense is that disapproval of male circumcision is increasing. Obviously no one is going to hold your son responsible for his own "genital mutilation," but if things continue as they've been going, in another 50 years there may be real outrage over you having done it. In fact, I was actually surprised to hear that insurance is willing to pay for the procedure; I thought insurance companies had begun citing it as an elective cosmetic procedure?
And if we're indeed considering it primarily a cosmetic procedure for the benefit of your son's future adult dates - I say, don't worry about it! His AI sexbot companion will be programmed to prefer whichever option he has.
Like, I dunno, man. That old principle about being conservative with cutting anything - hair, string, wood - because "you can always take more but you can't put it back" seems like it would apply here.
> polls of American women show they generally prefer circumcised penises
I assume that polls of African men in some countries show they consider uncircumcised vaginas to be icky.
In both cases, I wonder how much of the preference is based on personal experience and how much is just "sounds unfamiliar, therefore suspicious".
Generally: removing healthy body parts for contrived reasons based in archaic rituals and puritan moral values, with a risk of permanent sexual dysfunction? Not a great plan.
Also, one of the contrived reasons (reducing the risk of HPV) has also become moot with the rollout of vaccines.
Disclaimers: I'm ex-Orthodox Jewish and currently atheist. Search the web for my name and the word "circumcision" to find more about my views on child circumcision.
I will give some brief notes from a purely engineering/mechanical perspective.
The foreskin provides mechanical and liquid lubrication. The skin on an average adult male intact penis can glide frictionlessly back-and-forth a full 6 inches (15 cm) when fully erect. The head of the penis (aka the glans) contains mucosal tissue and is normally covered by the foreskin, keeping it moist. The foreskin helps keep lubrication inside during vaginal sex by bunching up on the shaft and forming a plug.
The head of the penis typically dries out after circumcision. Depending on how much skin is removed, circumcision can significantly diminish both the mechanical and liquid lubrication abilities of the penis. With the reverse plunger shape, the head of the circumcised penis could physically pull lubrication out during vaginal sex. A lack of mechanical and liquid lubrication can make masturbation and sex more difficult, less pleasurable, and potentially painful.
> A lack of mechanical and liquid lubrication can make masturbation [...] more difficult, less pleasurable, and potentially painful.
Note that historically, this was considered an important *feature*, not a bug, in about 1900 when male circumcision was made popular among English-speaking gentiles.
for what it's worth
i really really enjoy my foreskin
I second this.
About my foreskin, though, not his
It’s two-thirds of a haiku!
For what it is worth
I really love my foreskin
Spring comes, as do I
First place award for "Comment I Never Expected to Read on ACX."
I feel like it's not something that's brought up enough in these discussions (I'm assuming because it's mainly discussed in a US context, and most american males don't have foreskin?)
It's a small amount of skin, with a shit-ton of nerve endings that seem to be designed to provide pleasure more than pain, who wouldn't be a fan of that?
Well *this* comment is something that I'd expect to read on ACX; plenty of explanation.
I just admired the set-up, punchline, and brevity of your anti-circumcision argument; that isn't the kind of thing that tends to float around ACX as much.
On SSC, I remember "suntzuanime" as having a similarly terse/flippant style of commenting.
It's by a poster with the last name of "johnson", so it should have been a little bit expected...
I guess nominative determinism had to catch up with me eventually
I was put in the same position by my former wife 25 years ago. I chose not to do it. There was a brief period when he was around 17 where he bitched at me about it, but after that, no complaints. And some very nice girlfriends along the way I might add. My vote is no, do not do it.
To balance out this testimony, I am young man who was circumcised and I too have bitched at my father because I find circumcision to be a disgusting practice. So as much of parentage goes, there is no avoiding criticism.
Don't remove you child's body part for unnecessary reasons.
We did not circumcise either of our boys, on the grounds of general risk aversion. I was circumcised as an infant, myself, and have never had a problem - but some fraction of circumcised men are very, very upset about it. The benefits, if real, are small. Nobody goes around yelling about how happy they are that they were circumcised as an infant. If somebody comes to your house and says, “I’ll roll this d1000: on a 0-999 I pay you $0.01 every day for life; on a 1000 I shoot you in the gut,” do you take that bet? I sure don’t.
Some pro-circumcision arguments that I’ve heard online have failed to materialize. The prepuce is generally fused to the glans and non-retractable during infancy, so it never caused any difficulty cleaning them. Now that they’re older, I just tell them to clean it gently during their normal bath. It takes ten seconds. If “minimize the difficulty of bathing my children” was my decision criterion, I’d have to shave their heads.
Neither of my sons have ever asked me why my penis looks different from theirs. If they do, we’ll say the same thing we say for most questions about bodies: “some people’s bodies are shaped a little different than others. No big deal.” Then they go back to playing with trucks, demanding fruit snacks, or throwing a tantrum because their Mac ‘n cheese isn’t exactly the exact shade of orange they were expecting.
I can’t speak to potential issues in their love lives later on; they’re too young for that. But if they turn 18 and get upset that their junk is unfashionable, they can always get circumcised then. The reverse is not the case.
I haven't looked into the research too deeply or recently - how much of the evidence in support of circumcision comes from studies of HIV prevention programs in SSA? In addition to the problem of controlling for hygiene-based interventions, infectious disease in general is much, much more prevalent in SSA.
Have you found a pediatrician? I'd make notes about purported pros and cons and consult your son's pediatrician, when the time arises. (Anecdotally, it seems to be falling out of fashion, with a non-negligible proportion of newer pediatricians refusing to do elective circumcisions.)
It seems obvious that the principle of caution alone should prevent you from removing healthy tissue from your son's genitals. If he wants to be circumcised he can get circumcised as an adult, or indeed as a teenager whenever you think he has the maturity to make that decision for himself.
Circumcision status is not a binary state. Every circumcision is different. The surgeon cuts off an arbitrary amount of skin. Over half of the erogenous tissue on my penis was cut off. The skin on my erect penis is tighter than the skin on my index finger. Others have had far less skin cut off, to the degree that they are able to masturbate without lube. These are completely different surgeries. You cannot equivocate what I experience to what men who were circumcised as adults experience. When I read how other men describe sex and masturbation, their experiences are utterly alien to me. I feel a sense of loss, anger, and dread. The parts of the penis that they describe as being the most pleasurable are either missing from my penis or barely pleasurable at all. They describe masturbating multiple times a day, whereas masturbating is barely pleasurable (and somewhat painful) for me if I do it more than a few times a week.
Ever since I realized what it meant to be circumcised (at 23), I have had frequent intrusive depressive thoughts about sex and masturbation. It has distorted the lens through which I view the world. Not having been circumcised would have saved me a lifetime of depressive thoughts. I understand this is not a common outcome, but it's a possibility that is 1,000 times worse than any purported benefits.
Also, having to pour a bunch of liquid on my penis to masturbate and then having to shower every time afterward is extremely stupid and annoying.
Plus you can just defer the decision to when he is an adult! Other than the convenience of it all, what benefit is he getting from having it done now? A marginal decrease in UTI probability? Are you really willing to engage in nonconsensual genital cutting for that?
Thanks for speaking up. What you wrote is eerily familiar to me.
I can't believe there are several responses and they are all very tepid. Can I register my opinion that mutilating your infant's genitals would be a horrible mistake? Please do not mutilate your infant's genitals in any way. Unless you are living in sub-Saharan Africa and your son will grow in an environment without running water, where dry sex is culturally popular. In that case, circumcise away. Otherwise, please leave his genitals alone.
Sometimes circumcision does have to be done later on due to medical issues, so don't condemn it as completely useless.
I was just about to ask what DOGE had to do with circumcisions :P
Heh
The world would be a better place if Elon's parents had gone full mutilation mode and given him a prophylactic lobotomy. Not a partisan or Democrat voter here. But Elon is clearly over his head and relishes wrecking things because it makes him feel powerful.
So we are calling for the murder of our political enemies on ASX now?
Suggesting something would make the world a better place is a far cry from calling for murder. Many people gripe about being married. Few of them "off" their partner.
Suggesting that your political enemies would be better off dead is calling for murder. We just had assassination attempts against Trump in the most recent election. "Wouldn't it be great if the people I don't like were dead" is a call for murder, and some people have tried to act on it.
There was an adversarial collaboration (two people who started with opposing views on a topic, did research and tried to converge on a common view) abou infant circumcision on Slate Star Codex: https://slatestarcodex.com/2019/12/10/acc-is-infant-circumcision-ethical/ From their conclusion: "The benefits of infant circumcision appear to outweigh the risks and harms. Additionally, it is safer to be circumcised as an infant than as an adult, and a significant portion of the benefits of circumcision accrue to infants and children. From a strictly utilitarian perspective, infant circumcision should therefore be encouraged – whether we consider society as a whole or only the boy in question. However, autonomy is an important value, and while a man can become circumcised (missing only some of the benefits of having been circumcised as an infant), it is impossible to effectively restore the foreskin and become “de-circumcised”. An ethical system that heavily values personal choice over cost-benefit analysis may reasonably reject circumcision."
Thanks for sharing this. At this time, I'm leaning against circumcision for autonomy and precautionary reasons, but I also find myself sharing Scott's reaction:
"I had kind of swallowed the popular consensus that circumcision was Useless And Evil, and this gave pretty strong evidence that it has a variety of medical benefits and there’s no clear evidence it has any harms (especially reducing sexual pleasure, which is the main harm I’d heard cited). It seems to say that it’s hard to be against circumcision on anything other than a sort of deontological/autonomyesque/precautionary principle, which is not at all what I was expecting."
Most of the research I've encountered seems to either not come to a recommendation, or to be weakly in favor of doing or not doing it. I was taken aback by how strongly people here seem to be suggesting that circumcision is a completely indefensible practice. I'm probably not going to choose it for my son, but it seems like there is a reasonable utilitarian/non-religion based argument in favor.
My father is circumcised and made it a point to not cut me, and I am very very grateful to him for making that against the grain choice, so long ago. It is a choice I could make for myself now if I wanted to change it, but one I could not have changed if he didn't make that stand.
> There was an adversarial collaboration
There also was a reply by an actual expert in the comments of that post. He doubts that the supposed health benefits are worth it.
https://slatestarcodex.com/2019/12/10/acc-is-infant-circumcision-ethical/#comment-830436
If Musk set out to persuade Trump that by 2028 Neuralink will be able to download Trump's mind into Trump Junior's body to contest the election, I think it more likely than not that he would succeed.
We live in a Philip K Dick novel.
>I think it more likely than not that he would succeed.
<mildSnark>
Succeed in downloading or succeed in persuading? :-)
</mildSnark>
>We live in a Philip K Dick novel.
With the countdown to AGI and quite possibly ASI running, I'm tending to see Arthur C. Clarke's "Childhood's End" prescient in some ways...
It is going to be a wild ride.
Indeed, the vibe seems to have shifted since DeepSeek from a timid approach to safety to “we must ensure the Machine God is made in America.”
Many Thanks! Yes, that is the impression I'm getting as well. The next few years look like they will be astonishing. Well, 300,000 years was a good time, though (for mammalian species) not a long time...
I recently re-read The Penultimate Truth and it was really eerie, in a couple of distinct ways.
I run a trading bootcamp (https://www.trading.camp). The next one is happening this (extended) weekend in SF! Take your loved one on a surprise romantic getaway for Valentine's day with a ticket to quantitative trading bootcamp.
ACX readers can sign up with code EXTRACTDEALSON for $50 off here: https://bit.ly/qtb-february-2025-reg. If you're already great at trading and don't need the bootcamp, you can help me predict how many signups we'll get here: https://manifold.markets/saulmunn/how-many-participants-will-take-par
Also, we're the first event taking place at Mox (https://moxsf.com/), a new coworking space and project of Manifund that Austin Chen is starting up, which should be fun. If you want to visit the space and meet bootcamp folks but not spend a bunch of days and money on the full thing, you can join us for Provide Your Own Liquidity, our auction/afterparty, Monday night Feb 17: https://partiful.com/e/qJLPSXLOC8k35JqxYeGz?
If you have something that works, why share it? Especially for only a couple of thousand dollars?
If you don't have something that works, why should anyone pay to hear what you have to say?
Confessing embarrassing ignorance about musical theory here, but can someone explain how counterpoint in songs like Simon and Garfunkel’s Scarborough Fair or the screen Bridge over the River Kwai version of Colonel Bogies March works?
I mean I appreciate that it is pleasing to my ear and does work, but why?
Bear in mind that all I know about ‘a perfect fifth’ is that Claude Lacombe used it to describe the next tone — ‘Up a perfect fifth’ is how I remember it — in composing the alien summoning melody in Close Encounters.
I'd say consonant intervals like thirds and sixths sound good, and they sound even better when you move from dissonant (second and sevenths) to consonant intervals.
The question of why some intervals sound better than others probably relates to the frequency ratios mentioned in other replies (and also the way human brains are put together, which obviously nobody fully understands).
But also see Richard Feynman on answering "why" questions: https://fs.blog/richard-feynman-on-why-questions/
If it makes you feel any better, even people who are very good at counterpoint frequently don't know how it works.
"I've Got A Feeling" by the Beatles is probably my favorite use of counterpoint in rock music.
The way it works out in practice is very complicated. But fundamentally:
Sounds are frequencies. If you consider, say, 440Hz and 220Hz, the 220Hz sound "fits" into the 440Hz sound. Every second wave in the 440Hz sound is also in the 220Hz sound. There are no waves in the 220Hz sound that don't also occur in the 440Hz sound.
Those 2 tones are in a 2:1 ratio called the octave. Sometimes, two instruments playing an octave apart will blend together so smoothly, it sounds like one instrument. The octave is the most "consonant" interval. ("interval" is just tone ratio.) The perfect fifth interval is a 2:3 frequency ratio. It's not as consonant as the octave, but still very consonant. Every third wave is the same.
The intervals continue that way, 3:4 is the fourth, 4:5 is the third, etc. Putting the various intervals together into a musical scale is very hard and complicated, and different cultures solve it differently. Western music currently uses a solution popularized by Bach called "equal temperament".
The specifics of how different intervals get put together into harmonies like Simon & Garfunkel use is very, very complicated. But, fundamentally, the pleasure you get from hearing the kind of consonant intervals they sing is from a straightforward physics thing. The waves match up and fit together, and that's nice.
My understanding is that we don't really know exactly how Bach tuned his keyboards, but it probably wasn't the same as "equal temperament".
https://en.wikipedia.org/wiki/Bach_Temperament
I looked up the topic in Albert Schweitzer's Bach biography. Schweitzer states that Bach indeed used an equally-spaced tuning. But it would also make sense that the "actual" tuning wasn't perfectly even, it's not like he had a frequency analyzer. He had to tune by ear, and so the end result was whatever he thought sounded good in all 24 keys. Which would have to be pretty darn close to the perfectly even temperament.
Oh, I thought that he popularized the equal temperament system with "The Well-Tempered Clavier". Equal temperament allows you to play in all keys, so The Well-Tempered Clavier showcased that feature by having one piece in every possible key, which wasn't possible before equal temperament.
Not Quite. Bach (JS, the famous one) was a proponent of a number of "Well-Tempered" systems that gave each key a unique character but still allowed music in all key centers.
Equal Tempered temperaments only came into their own with the expanded chromaticism of the 19th C.
There are countless articles online debating the specifics, but as near as I can tell no one today seriously thinks Bach ws promoting "equal temperament"
Interesting!
Yep, that was his explicit goal.
Thanks for the reply. I think I have the pure physics part down. I’ve written code to automate hearing aid testing. Give me the FFT decomposition of sounds and the math is straightforward.
The pleasure of certain combinations is where the mystery is for me. But hey some mysteries are good.
Unfortunately I didn’t play any instrument when I was young or study music theory so I think part of this will always remain beyond my ken.
Nothing to stop me from enjoying the effects though.
Edit: It still bothers me a little bit that the frequency of middle C has changed over time. The math and physics guy in me wants to know “How can that be?” Yes, a small joke.
Further edit approaching non sequitur:
I run into a professional piano tuner occasionally in my local coffee shop. His technique is to begin by placing a vibrating pitchfork* against his teeth. Seems weird to me but he is the pro after all.
*Did I really type pitchfork when I meant tuning fork? Oy! I really need to read what I’ve typed before I press Save.
I can explain another level deeper. This is where you get into songwriting and composition. I'm a (amateur) musician, not a songwriter or composer, so I'm not fully fluent in this stuff.
Once you've got the primary auditory experience of pleasure (consonance) and displeasure (dissonance) available, and expressible in various degrees (there are more or less dissonant intervals / chords), you can begin to build an art on it.
Western songwriting primarily functions by setting up an expectation and then fulfilling it. Or, to put it another way, by creating tension and then resolving it. Songs have a key (for instance, C). C is the primary tone; call it the 1. You can build a scale on C; it goes C-D-E-F-G-A-B-C (or 1-2-3-4-5-6-7-8, where 8 is the octave). You can build a major chord on C, which goes C-E-G (1-3-5).
The most consonant other chord available is the chord built on the 4 of your scale, which, in C, is F major (F-A-C or 4-6-8). Another highly consonant chord you can build in C is the 5, or G major (G-B-D or 5-7-2). Neither of these are as consonant as the chord built on 1, because the 6 and the 2 clash a bit with the 1 (which is to say, audiologically, their main overtones form dissonant large-number-ratios with the 1's main overtones).
Because of music theory I'm eliding here, the 2, 3 and 6 chords have to be minor chords, which just means they incorporate a bit of extra dissonance in their thirds. You can and should use them; they'll be felt as very tense and unsettled, but still harmonious. You don't use the chord built on the 7, it would be so dissonant as to be unharmonious.
So a simple chord progression in C might be 1-4-1. You start with the most consonant possible chord, C major; then move to a less-consonant chord, F major; this creates a tension or expectation; which you then fulfill / resolve by moving back to C major.
One of the most common chord progressions in traditional American music is 1-4-1-5. Amazing Grace (basically) follows this progression, but also many other traditional songs. It creates tension, but never goes too far from home base.
The most common chord progression in contemporary American music, by far, is 1-5-6-4. (This is the one where you see those videos of bands playing 30 different songs, all with the same chords!) I think this one shines because it gives you a good variety of different chords, and kind of mimics the hero's journey, in a way that's hard to explain. Maybe a more competent composer than me can explain why that is. "When I Come Around" by Green Day displays this chord progression really obviously.
But you also get chord progressions like 4-5-6-4-5-6, particularly on the bridges of songs, which leave you unsettled and longing strongly for a resolution (which usually hits with the chorus, which will probably start with a 1 chord).
Or chord progressions like 6-4-1-5, which are hard to stop playing. Because the 1 happens in the middle of the sequence, you want to get to it by starting the sequence over again whenever you finish it.
The melody always uses notes from the key's scale. Melodies tend to have certain emphasized notes that function as resting points; these notes should belong to the current chord. Otherwise, it's dissonant. Generally, you can use dissonant notes in a melody as long as you're using them to pass to a consonant note.
Literally all of these rules have tons of exceptions, like in all arts.
So that's the bare bones basics of why different combinations of tones and intervals give pleasure in different ways. Like lots and lots of other arts, it's fundamentally about creating and resolving tension.
This is an excellent songwriting primer.
Thanks. This is very helpful and well aimed at my level of knowledge.
I think if I were to keep unraveling this down to the level of “Why does well done three part harmony give me the chills?”, I’d have to get to the mind body problem or the true existence of human souls.
Fun long run — literal running —ruminations.
I still prefer to tune my guitars by listening to the beats between the overtones generated on the 7th and the 5th frets of the higher and lower strings respectively, once I tuned the A string to the 440 Hz standard (or whatever tuning level I need for the piece).
I used to do this too, it's pretty common. I stopped when I learned more. The perfect 5th you get with that technique is 2 cents larger than the equal-tempered 5th, and it compounds so that your G string will be too sharp by 4 cents.
I've learned from experience that I need to err just a pinch flat. but I always assumed it was just a skill issue. But I do know in the back of my mind that there's a difference between equal and just tuning, so I should I have expected this in hindsight.
I feel so vindicated and so dumb, simultaneously.
Yeah, I learned over the years not to totally trust this and check the sound of the D-octave between 5th on A and 7th on G.
For B and high-E I go back to the 7th fret overtones on low-E and A strings, respectively, so the error doesn't compound.
Gently strumming Emajor chord afterwards to listen to any weirdness is the final check.
But when recording critical tracks I do rely on the fancy tuner in the Line6 effects board so there's that...
I think you meant to say "There are no waves in the 440 Hz sound that don't also occur in the 220Hz sound."
That seems backwards: wouldn't 440 hz have twice as many waves? Or does a higher hz mean a lower frequency?
First, to clarify: we are talking about notes produced by musical instruments, e.g., a vibrating string, not pure tones generated by an oscillator. Here's the harmonic sequences:
440 Hz (A4): 440, 880, 1320, 1760, 2200, etc.
220 Hz (A3): 220, 440, 880, 1320, 1760, 2200, etc.
EDIT: A3 has multiples of 220 Hz, so the sequence is 220, 440, 660, 880, 1100, 1320, etc....
I'm not clarified. It still seems to me that a 440 hz would have waves that don't occur in a 220 hz, while a 220 hz would have no waves in it that don't also occur in 440 hz.
I mean, there's no 220 Hz in the 440 Hz note. But every frequency in the 440 Hz note is also present in the 220 Hz note.
This is of course an idealization that supposes an ideal string.
"Counterpoint" just means "simultaneous, complementary melodies." There isn't a unifying way they "work" - a bassline outlining the harmonic form in the absence of a chordal instrument could be analyzed as counterpoint, it's just a common enough musical convention with a distinctive enough function that it gets its own name.
The wikipedia page looks like a good overview: https://en.wikipedia.org/wiki/Counterpoint
There are LOTS!! of details I don't know, but the basic answer is "it's been experimentally observed". That's not a good explanation of why it works, but it's a basis from which compositions can be created. A second and a seventh are dissonant, but differently. (Note that the frequency separation is the same, it's the context that they appear in that makes the difference.) So there are LOTS of complicated observed patterns, that people have defined reactions towards. (And also note that a lot of this is learned. E.g. Javanese music doesn't use the same notes as the octave scale.)
P.S.: Don't rely on the details of this, my wife was the musician, and maybe there is a difference between the 2nd and the 7th. But in general the answer is correct.
Yeah, I agree with the general answer, but I'd say a 2nd vs a 7th have different frequency separations (if you raise the lower note of a 2nd by an octave (aka double the frequency) you get a 7th), while the dissonance quality is similar.
I've been following the news about Santorini earthquakes with a degree of morbid fascination. The archipelago is a remnant of a supervolcano explosion, i.e., a caldera. When this thing blows up again it will be catastrophic. This situation is the opposite of the 2.1% sweet meteor of dea... scratch that, giant explosion of Christmas 2032: we know with near-100% certainty Santorini will blow up, just don't know when.
Only tangentially related, but I consciously experienced an earthquake for the first time a few weeks ago, it was deeply weird! This is notable because I live in a the north east, not a land known for quaking earth, old hat to you Californian folks I'm sure.
NE earthquakes are more common that people seem to assume. The mountains there are older but there's still some movement happens. I believe there were several since the turn of the century. A cursory search produced this: https://nesec.org/massachusetts-earthquakes/
You're on your way to prevent multiple statistical deaths by donating 10% of your income in the form of a cashier's check, when you see a drowning child. If you attempt to save the drowning child, you will not be able to prevent multiple statistical deaths. What should you do?
Save the child, the go to my bank and have the cashier's check re-issued. The bank makes the funds for the check unavailable when it's issued, but they don't actually leave my account until they're deposited. Worst case, I have an extra 10% of my income unavailable in my account for 90 days while the bank waits for the first check to expire; I have more than enough liquid assets to cover this. Or if the soggy remains of the original check are recognizable, I can present it to my bank as proof that the original was damaged and the check should be re-issued without an additional hold.
Can't resist - this is too close to my old snarky version.
You are on a trolley taking a shipment of mosquito nets to the post office to send to Africa when you see Peter Singer drowning in a lake. You can pull a lever to stop the trolley and save him, but then you will not get to the post office before it closes for the long weekend, resulting in deaths > 1. What to do?
I am not a utilitarian, but I would let him drown - his conclusions are repugnant.
If I take into consideration that this is a path well-trodden by many, and that any moment now another Rationalist and/or Effective Altruist will be passing by, I leave the child and hasten to the bank with my cheque. Somebody else will save the child, whereas only I can prevent all those statistics!
Alternately, I hope that an ordinary person is passing by Drowning Child Lake (honestly, the local council really should do something, maybe safety rails or at least a lifebuoy) and since all we on here know Normies Are Stupid, they'll plunge right in to save the kid without even doing a statistical analysis first!
You mean would I rather be celebrated as a town hero, for free, or spend half my disposable income on something I can’t even mention socially without seeming like a douche?
I heartily chuckled.
But statistics, Les, statistics! Why, without those, we'd just be like ordinary people who react on gut feeling! Or even (shudder) deontologists! 😁
This is the way. And I don't need any silly math to rationalize it.
This is my chance to register my view that utilitarianism is Bad and Wrong. I love Scott Alexander's writing in spite of the utilitarianism, not because of.
Utilitarianism is definitely a minority position among regular people and probably is among philosophers. The question is why it seems so over represented, especially on Substack.
Substack, and this substack in particular, is for nerds. Utilitarianism transforms ethics and morality, traditionally difficult and fuzzy humanistic problems that even the wise struggle with, into mathematical equations that can be solved with a bit of algebra or maybe a spreadsheet. The appeal should be obvious.
Do you mean Substack generally, or just this one?
For this Substack, I wonder if it's because there are relatively more autistic/spectrum types here. Maybe there's something in Scott's survey data?
I get the appeal of systematizing ethics but it’s not the only way to do it, as Kantians could attest. Although if someone doesn’t know anything about the subject than it might appear that way. I once had someone here tell me they liked utilitarianism because it had numbers and they didn’t really care what those numbers were as long as they gave an answer.
Satisficing consequentialism* says *either* donating the income or saving the child is good and praiseworthy, and staying home and doing nothing is probably fine too.
* Satisficing is not a typo.
I know, it's one of those business neologisms which make me grind my teeth.
Normally I would feel the same way, but I was happy to find something to counter the rationalist utilitarians.
We have a greater obligation to act when we are the unique and only person who can solve a problem. If there's a waterpolo team training in the same body of water as the drowning child, and shouting at them resolves the problem, then all is well.
If we're somehow prevented from donating the money later by saving the child (which seems unlikely) then this is still morally acceptable, as it's likely that other people will donate to save those people in some statistical sense. We might even compensate in the future by encouraging others to donate or increasing our future donations.
I think many of these questions go away if we accept some sort of nearness bias: it's OK to value the lives of those near to us in distance, time or blood relation above others further away.
Honestly, it seems like the best use of our hypothetical income would be to set up a lifeguard station beside this lake because those dang kids keep jumping (or falling) in and apparently none of them can swim a stroke.
Either that, or pay to have the lake drained.
I blame the parents. Where are they and why are they letting their kids run around lakes without being able to swim?
Good luck keeping Irish kids out of every drowned quarry and undercurrenty river. They're like lemmings when the weather gets above 20c...
We really MUST value those we can directly perceive higher. We have greater certainty about their condition. If you donate money, you don't know the secretary isn't going to run away with the treasurer and use that money for their fling. You may assign that a low probability, but it should be well above zero.
(FWIW, I have donated money that was embezzled. This isn't an impossibly rare happening. And I suspect that it happens a lot more often than we hear about.)
Oh, there are many such cases! We had one a few years back which ended up in a right old ding-dong in the courts between the disgraced former boss of a charity, who appeared to be treating said charity (along with other board members) as a nice little earner for her. Then she went to court to sue over her reputation. I can't remember now how it ended up, but it badly damaged the charity, and charitable giving to other charities, for a while.
https://www.thejournal.ie/angela-kerins-pac-timeline-3214906-Jan2017/
https://www.irishlegal.com/articles/supreme-court-angela-kerins-discovery-appeal-dismissed
Don't forget Paul Kelly and Console, for the irony if nothing else:
https://www.rte.ie/news/primetime/2024/0222/1433824-console-inside-story/
Yeah, that one is pertinent in regards to the uproar around USAID. Good causes being used as cover for fraud and scamming. Who could possibly be against funding charities that help people impacted by suicide? Even the possibility of criticising them would mark you out as a bad person.
Btu somebody has to be the bad guy who checks to see if all the alleged good causes are really what is claimed.
No one likes auditors (or Revenue) but they're fulfilling a social purpose. What we've learned again and again is that no-one can be trusted to be above the law. Trust but verify! (And don't burn the whole thing down for spurious reasons.)
"This isn't an impossibly rare happening. And I suspect that it happens a lot more often than we hear about."
It happens every time an employee of a charity drives a nicer car than a 1980s Buick.
When I see a drowning child, I can be more or less 100% certain that my choices can be the difference between life and death.
When I read that my money will buy malaria nets and save 10 lives on another continent, there are many layers of assumptions between me and the lives that I'm saving:
- I'm trusting that the science of malaria prevention through mosquito control is accurate, and that bed nets are a reliable method of mosquito control. (This seems pretty safe to assume but they get less reliable from here)
- I'm trusting that honest studies by good researchers have been done and have correctly identified the cost/benefit ratio here.
- I'm trusting that other researchers have done good studies comparing the cost/benefit ratio of this intervention to other potential interventions.
- I'm trusting that the actual charity that I am giving money to is in position to effectively execute the intervention in the most cost efficient way possible.
- And because I have no direct connection with any of those things, I am trusting my ability to read articles & studies and listen to smart people online and figure out which of their opinions I most believe given their inevitable disagreements.
All of which adds up to way less certainty than "that kid right there is gonna drown if I do nothing, but may live if I intervene."
Which is to say: I would save the child. I also believe in effective giving, and I _do_ try to give towards things like malaria nets, but the two things should not be treated as close to equivalently as they are.
Be very very sure they actually are drowning. It is 2025; grabbing a stranger's child is unlikely to end well for you.
"All of which adds up to way less certainty than "that kid right there is gonna drown if I do nothing, but may live if I intervene.""
What were the counter-balancing layers of assumptions about your likelihood of successfully saving the drowning child?
Great question! If we're playing with hypotheticals then the 'child drowning' scenario is also rife with uncertainty.
- Is that child actually drowning? Maybe they're just playing? As a parent myself I've tried to study up on what drowning actually looks like. I have moderately high confidence that I could identify it. A harder scenario would be if they're not drowning at _present_ but are at risk, like if they fall off a boat and are floating downstream and the water is getting rough. There is certainly a level of uncertainty here.
- Am _I_ the only one who can save them? If someone else has already dived in and is heading towards them, then my intervention may not be necessary. Or maybe that rescuer is not actually as good as swimmer as I am? Maybe I should be ready to save both of them.
- On the other hand, if I jump in, and the water is rougher than it appears, then it's possible that not only will I not save the child, but that I myself might drown, or require saving myself, or worst of all I might drown myself _and_ induce someone else to drown trying to save me.
Nonetheless, as someone with a child who regularly goes to pools with other kids, I think it is well possible (if not likely) that I will face a scenario where I have a reasonable chance of helping to save a life with my direct actions.
There are also instances where people jumped in to save drowning children and ended up getting drowned themselves.
https://www.breakingnews.ie/ireland/funeral-takes-place-of-mother-who-drowned-attempting-to-rescue-son-1162874.html
https://www.irishexaminer.com/news/arid-30595608.html
https://www.sundayworld.com/news/irish-news/man-who-drowned-trying-to-rescue-children-from-sea-hailed-as-truest-definition-of-hero/a229790957.html
So sometimes the right decision is to *not* jump in but let others (better qualified) do the rescuing.
Regarding Minicircle, many issues including:
1) Contra various inexplicable claims here, plasmids are not safe! They're only safe if not delivered by a real delivery vector because they don't get taken up by cells and basically nothing happens. If they're delivered by a real delivery vector such as LNP they end up in the cytoplasm and activate cGAS-STING which is incredibly toxic and inflammatory.
1a) Was the Ichor CRO experiment literally just naked plasmid? Not even minicircle claims that works I think, they use PEI which is its own brand of problematic. It's concerning that the powerpoint does not mention a vector.
2) That CRO report was vaguely interesting but they should have done qPCR or RNA sequencing to confirm the presence of mRNA generated from plasmid.
3) There are two separate "1,000-fold" gauntlets to run. First the issue of the ng vs pg discrepancy that was pointed out by another commenter. Second is the issue that local injection in mice and humans are not all that different in terms of volume, but humans are 1,000-fold bigger than mice so local injections go ~1,000-fold less far! This is why you need systemic delivery to e.g. the liver if you want to scale from mice to humans for systemic administration.
oh and no dose was given in the report so it's unclear how much it could actually be scaled up, or how it compares to minicircle
Does Hamas have an endgame? Are they fine with the status quo of 'reconstruction is extremely slow, we stay in power and are periodically raided,' or do they think there's a significant chance of something either worse or better for them happening. I haven't seen decent public information about how organized their leadership is, and what efforts if any they're making to shape the future.
Hamas is not run by rationalists, it is run by fundamentalists. A significant fraction of them genuinely, sincerely believe that if they stay the course then eventually an omnipotent being will rejigger reality so that Israel no longer exists and a Jew-free Palestine will become an Earthly paradise even greater than a Trumpian riviera, where the Palestinian people will live happily ever after. Under Hamas's rule, of course.
Another significant fraction believes that this is an extreme long shot, but that it is still better for a Palestinian, or even all Palestinians, to die at Israel's hands and thus be granted a martyr's ticket straight to heaven, than for them to be shuffled off to some refugee camp where existential despair will lead to unvirtuous behavior without the opportunity for martyrdom and so ultimately to the Not-Good Place. There's no rush, of course, so might as well try for the Jew-free Palestine version, and you have to at least be trying or the deaths don't count as martyrdom.
And then there's the ones who don't believe in any of that, but they have a *relatively* cushy gig as the Gaza elite, better than anything they'd find in a refugee camp, so long as they *pretend* to believe all that and act accordingly.
So, yeah, they're probably mostly OK with the status quo of Israeli occupation and/or blockade, and occasional Al-Aqsa Floods.
Their endgame is that ethnic cleansing or genocide is taboo in the 21st century in any population/place with 50+ cameras pointed at it, let alone an 80-years-old conflict infused with thousands of stories and inflammations/conspiracy theories and credible accusation of genocide. But Israel backed itself into a corner where Ethnic Cleansing or Genocide is about the only decisive and permanent solutions.
The decline of American power, and the increasing misery of the American taxpayer funding Israel's war (one among many wars (s)he keeps paying instead of healthcare), seem to also have something to do with the Palestinian militants' hope that - just maybe - one more Iron Dome missile would be too much. Right now, all the polls I have seen among young (< 25 years) Americans seem to be validating them, their bet is that (1) This will continue as those youth age into politics (2) They will survive the 30-50 years necessary for this to take effect, or the century of China-accelerated American decline will ramp up faster than America's Israel-hostile youth are aging (and that the 2 will eventually meet in the middle and live happily ever after).
------------------------------------
Israel is kinda hopeless demographically, very long-term speaking. Think about it, their fastest growing socio-political faction is the one explicitly preaching that military service is worse than death. And even with their fastest growing faction included they're still on equal demographic basis with the Arabs living in the whole land between Gaza and the Galilee. Once you expand scope a bit and notice the demographic trends among, e.g., Egypt and Jordan, you would soon see that the Arab autocracies keeping their people exactly 1 meal ahead of starvation are part of the Iron Dome, and Arab autocracies are one hell of an unreliable safety net to depend on.
Unless, that is, Singularity happens and now 1 person with enough money and data centers can genocide ~150 million people. Well, in that case, that person is about equally likely to be a Gulf Sheikh (who won't spare an Arab or a Jew) as an ultra-orthodox warlord.
The extremely smart move on part of Israel right now is to utilize their 50 years of runway that their current military supremacy is worth to achieve whatever political goals they can conceivably get, which is a lot. Asking for a Palestinian state is **remarkably** tame, it's the position of the USA since as far as I can go back and still recognize the names of the combatants. (which is why it won't solve anything if it's not properly militarized, a state is a military, anything else is a boy scouts club in various guises.)
Other than that, it can always risk genocide, effectively bluffing the combined might of Saudi Arabia, Egypt, Iran, and maybe even Turkey into doing something. To say nothing of what the EU and USA would do after a literal Hitler 80-anniversary reboot.
That's why Israel's Right Wing loves yapping about """Voluntary Emigration""" so much, but they're kinda too racist to even do that properly, after all, literally no Gazan except Ismail Haniyah's descendants would stay in Gaza for 1 second more if given a million dollar and a one-way-ticket + green card to Canada/USA, and one million * one million Gazan is a trillion dollar, about 30 years of Israel's military spending of ***Opens Google Tab*** 30.5 billion dollar. 1 million per Gazan is an obscene exaggeration, and Saudi Arabia + Europe can foot some of the bills (Saudi Arabia casually promised 600 billion, 60% of the bill to """Voluntary Emigrate""" 1 million Gazan @ 1 million dollar per Gazan, in a phone call with Trump). So given how easy it is to "Voluntary Emigration" a population out of a land, the Israeli Right Wingers are still too low-IQ and low human empathy to do even that properly and diplomatically.
It's extremely hard to predict, and Hamas is in a very weak position, but Israel is also in one hell of a corner. Like someone alluded, their biggest hope is that Palestinians will just... get bored? I don't know, seems kinda dumb, like someone who has no idea what "Revenge" means trying to predict the behavior of a human group after killing anywhere between 3% to 10% of the group and scarring who knows how many for the rest of their lives.
"Maybe after this they would learn their lesson", yeah, I bet that worked great for the Ancient Jews after the Romans demolished their state and scattered them all over the Middle East and Southern/Eastern Europe. Well technically it did, it just didn't work out for their successors.
Tldr; I don't know what Hamas endgame is and anyone who thinks is too dumb or too dishonest to know better, but it involves the global isolation of Israel and the "Demographics Is Destiny" hypothesis, along with the conviction that Islamist fundamentalism is enough to keep the flame of revenge glowing.
"Israel is kinda hopeless demographically, very long-term speaking. Think about it, their fastest growing socio-political faction is the one explicitly preaching that military service is worse than death."
The actual number of non-Haredi Jews is only going to go up.
Indeed, gradually ruled over and out-voted by the non-serving Haredi. Not exactly a recipe for a very stable social contract.
We corresponded X thousand words on a previous open thread without resolving the question of whether any state composed primarily of the people currently in Gaza/West Bank could be persuaded (once it had the means to do so) to not attack Israel motivated by the belief that jewish self-determination in the middle east is an injustice not to be tolerated as a primary consideration over all other goals.
Just curious, have your beliefs on the topic/this question changed in the weeks since?
As a companion to my assertion that there is almost perfect symmetry between Palestinian Peace Obstructionism and Israeli Peace Obstructionism, here's a recent story in Haaretz about a Jerusalem book store whose owners were accused by the police of incitement to terrorism and dragged to court over books that contained such violent symbols of Palestinian Nationalism as poetry and cuisine https://archive.ph/ohDi5.
I'm not sure whether people outside israel who quote Haaretz are aware just how much it is the equivalent of anti American anecdote personally cherrypicked by Noam Chomsky, who is American so how can you accuse him of anti western bias
This is bad, and toxic, and there are Israelis protesting these actions by Ben Gvir right now.
I do think there's a categorical difference in how disfavored political speech is treated by Israel vs. anywhere in the middle east.
I haven't done any new readings since then, less than during the hot war in fact. More work in my life, less mental health to keep track with all the insane genocidal shit that keeps gaining grounds in English-translated Hebrew media (let alone the raw Hebrew media).
My beliefs on this topic or question isn't opposite, it's a "Yes And" not a "No" or even a "Yes But". Yes, obviously the cultural meme of "Jews are colonizers that can be driven out of the land if we just kill the next 50 of them" needs to die, hopefully quickly, hopefully with the minimum possible Palestinian and Israeli deaths.
But you know, I could almost say the same thing about massive portions of the Israeli mythmaking, collective consciousness, and culture too. There is an ironic and delicious symmetry in this conflict, keep reading enough of both sides and you will eventually be able to morph the grievances of each side into each other almost loss-lessly. I don't even see 0s and 1s anymore, I see "Arabs took this Jewish land from us" and "Jews took this Islamic land from us".
Where was I? Ah Yes, the weird Israeli supremacism that desperately asserts an ethnically-homogenous state is one of the fundamental human rights promised to people in the Universal Declaration of Human Rights, but also Palestinians can't have one. Because... because,... Palestinians are made up, they don't exist, they are all Jordanians and Egyptians in watermelon-colored trench coats. Or maybe they exist but they're just fake. Maybe they're real but have too few Nobel prizes, or maybe they deserve ones but we're just too uncomfortable to give them one now, come later in 15 years (Spoiler: Nothing will change in 15 years, it's a bluff in the desperate hope that Palestinians will get bored in 15 years despite not getting bored in 80).
Israelis and the people who love them (which sometimes include me, some of them, some of the time, when I'm not reading the nth genocidal translation of the latest pop star or politican to get viral) get too upset at Hamas supporters chanting "By Any Means" and "Those Who Make Peaceful Resistance Impossible, Make Armed Resistance Inevitable", and I get it, I really do. It feels wrong to be this gleeful at a literal real-life horror movie that I would dismiss as too cheap and blood-plotted if it was a Hollywood production.
But you know what, they're kinda right, as a purely descriptive (not normative, despite them meaning it in a normative, "cheer the revolution" kind of way) model, it really is the case that talking is the wonderful social technology, that - among other thing - discharge violence and give hope (however dim and false) to the dispossessed. Right now, carrying a Palestinian flag in public in Israel is a crime ("incitement"), that's why Peace Protestors and Radical Bloc leftists often carry watermelons or Palestinian-colors-Arabic-words.
It really is a truth that when you make it a crime for people to have nationalist aspirations, eventually they're going to figure out how to do it without you, and you're not going to be pleased with how they do that.
> but also Palestinians can't have one.
I had to stop reading your post right at that line, because everything I know about the history of this is that in 1948 the Palestinians and theIsraeli’swere given a state and that resolution was not a satisfactory one to the Palestinians or to any neighboring Arab state.. And it seems to me ever since then it has not been a satisfactory one.
Whatever the horrible mess we find ourselves in now it has been the stated intention of only one side that the other must be eliminated from the area and preferably from the Earth.
> in 1948
I'm discussing Israeli attitude right now and for the past 40-60 years. The Palestinian state stalled Egyptian Peace negotiations until Egypt relented (and also it wasn't very serious about it to begin with). The Palestinian half-state that Rabin had to clarify to the Knesset that it's not exactly a state got him killed, the only Israeli PM to meet that fate, as far as my memory serves.
I can find plenty of Israeli attitudes in 1948 and before to demonstrate that "Whatever we're in now is the fault of a single people" is - to a very good seventh-order approximation - politically-motivated Disney Morality to justify atrocities.
> ever since then it has not been a satisfactory one.
No credible offer of Palestinian statehood was ever put on the table since 1948. And the one on the table in 1948 wasn't from Israel.
No, it was not Israel because Israel did not exist officially. It was the UN and it was based on what the British did to try and make a two state solution. It proposed a state of Israel and a state of Palestine. Israel accepted those boundaries.I will disagree with you about offers on the table subsequently to create a two state solution. I think they had plenty of chances. The deal Arafat walked away from for instance. It might go a long way towards solving the problem if Hamas would remove the destruction of Israel as one of their core tenets. I have said before in these comments that it is very difficult to negotiate with someone whose goal is to kill you.
What was not credible about the offers in 2000 and 2008?
If you’re talking short term goals, I think their hope was the ceasefire would hold, and after a few months, after they re-established de facto power, the prospect of restarting the war would induce Israel begrudgingly accept the status quo.
Long term, probably they hoped to rally the neighboring countries to their side, or force Israel to make a deal with them fully accepting their permanent control over Gaza. Both long shots of course. It’s long been known that there were comparatively moderate and extreme wings in Hamas in tension with each other. Personnel has changed of course though so who knows now; but I’d say it generally behaves like an organization with considerable internal conflict.
In general the Palestinian strategy has been to resist Israel as long as possible using unconventional methods (typically against civilians) rather than face the overwhelming Israeli military. I believe this is mostly to gain internal political credibility and to keep the conflict alive in international news. It also provides a morale boost and exerts a small amount of negotiating pressure on Israel by showing that they are capable of doing some damage. While doing this, they appeal to international organizations and global public opinion to attempt a diplomatic victory, similar to the way international pressure ended South African apartheid. They seem to view this as a genuine model for potential victory rather that purely being rhetoric used for propaganda. (They have seen huge gains in public support over the past decade or so, so you could argue that the strategy has been at least partially effective.) It is hard to predict if that strategy will continue post-Oct. 7. They seem set on fighting to the death rather than being expelled or living under Israeli control.
They've been outspoken that their plan is to just commit more and more October 7s, forever, until there are no more Jews. Pretty retarded if you know how population increases work, but if they weren't stupid they wouldn't be in Hamas, so there's a pretty hard upper bound on their rationality.
Trump is offering Hamas a fantastic endgame: sell Gaza to him and retire in Dubai or Qatar.
I doubt their retirement would be long lived. Once they’re out of power and have no leverage there’s little reason for the Mossad not to keep assassinating them. And fanatics tend not to weight personal self interest overly much anyway.
I don’t think that’s going to have universal appeal in Gaza, or the world.
Since when does Trump care about that?
That's exactly why almost nothing of his foreign policy works. He perfected the art of "Whining and blustering till someone blinks" and all the limited concessions that can be obtained by it, but has 0 strategy or international smarts beyond that.
There is a reason there isn't a wall with Mexico since the ~10 years he has been publicly promising it.
Building a wall across the entirety of the southern US border is a lot harder than moving/killing 2 million people concentrated in an area that's only slightly bigger than San Francisco.
I don't think you've done any project planning on genociding 2 million people...
Everything is much harder than it looks when your attention span is shorter than fruit fly sex and your IQ is lower than temperature in Ontario.
He's proposing to _buy_ the property? I hadn't noticed that element in his statements....pretty sure that no Republican member of Congress would have anything but an "oh hell no" reaction to idea of Hamas members getting paid cash money for Gaza.
(Perhaps Trump could finance such a purchase the way he paid for all his business ventures: no money down, extravagant promises of future balloon payments, stall for years on making the actual payments, strip cash flow down to negative then file for Chapter 11 bankruptcy and get the courts to stiff all the creditors for you, rinse and repeat. Maybe Hamas/Iran would fall for that, one time at least?)
"I'm committed to buying and owning Gaza. As far as US rebuilding it, we may give it to other states in the Middle East to build sections of it" (source: https://www.bbc.com/news/live/cd7dwq87zvqt?post=asset%3Ad8fd6f71-012c-409c-8692-cae8bb9d2334#post).
WRT: "pretty sure that no Republican member of Congress would have anything but an "oh hell no" reaction to idea of Hamas members getting paid cash money for Gaza" - I've lost count of all the times I though this about various Trump's... "ideas". Alas, the Republican party is now basically a North-Korean-style "write down everything Dear Leader says and enthusiastically support" Party.
Both of you seem to be under the impression that he means to purchase it *from Hamas*. I would've thought that he means buying it from Israel, and the payment will probably be in the form of weapons and such that we would've sent them anyway, only they won't have to pay for them.
I'm using "buying from Hamas" as an illustration of the utter idiocy of the whole idea. It's really hard to take it seriously, and I am astounded (but I guess I should not be) at the ineptitude of the journalists covering it, should not this very question, "who are you buying it from, Mr. President" be shouted, printed, posted, whatever, all over the place?
But - taking it seriously for a moment - there's no need nor a legal way to "buy" it from Israel, is there? Israel controls it militarily at the moment, it can just hand control over to the US, how is a "purchase" even relevant here? But it doesn't "own" it in any legal sense, which is what would be required for a "legal" purchase?
God I feel stupid and mentally spent just typing this.
You may misunderstand what Trump tries to do.
Assuming he is a clever Machiavellian and not a bumbling buffoon, the point of the "bying Gaza and expelling all Palestinians" schtick is to influence how Hamas perceive their choices. He might attempt to scare them into a deal, by influencing the perceived disaster they may face, if they do not. It's a negotiating tactic.
Haha, underestimating the ineptitude of journalists, classic rookie mistake! Well, not a crippling one; you'll learn.
I agree though, yes, someone at some point really ought to have asked the question. Arguably Israel does own Gaza by dint of conquest – the only other possible alternatives, since there's no Palestinian state and never will be, are that it still belongs to Egypt, which they vehemently deny since that would make them responsible and the Gazans Egyptian citizens, or that it's a stateless no-man's-land. So potentially Israel can sell it to the US, and presumably Trump's idea is that buying it will give US ownership greater legitimacy somehow. Trump is a dunderhead in many ways, and this is typical of his way of thinking IMO.
Haven't states historically ceded territory as part of peace deals? Is your objection just that that he used the terms of everyday commerce instead of some fancier "international law" jargon?
Oct. 7th was the plan. Playact submission for a decade to convince Israel to take their eyes off you and write you off as a managed problem. Build up the people, war making material, intelligence, and connections with other Jihadi entities you need, launch a sudden and overwhelming strike from all sides, kill tens of thousands of soldiers and civilians, capture tens of thousands more, capture and destroy IDF equipment, collapse the Israeli state.
But the expected attacks from Hezbollah and the planned for uprising in Judea/Samaria never came. And their own military success fell short. And Israel responded in a much harsher and sustained way than they expected. And then Israel knocked out Hezbollah and launched a massive Jihadi suppression campaign in Judea/Samaria.
So Hamas now is falling back on their tried and true survival strategy of milking international sympathy for the dead and suffering Gazans they generated to limit Israel's freedom of action and give them the chance to regroup and fight another day. And that's what you're seeing now.
There's at least three active levels warfare between the Gazans and Israel going on all at the same time.
There's the direct warfare, which comes and goes but has been pretty significant for a while. That's obvious. Then there's the political wrangling between all the various groups that make up each side (Palestinians, Hamas, Iran, Israel - broken down by factions, etc.). Finally there's the propaganda war that's mostly fought for sympathy of outsiders.
Israel's long term plan is on the political front, and seems to be to slowly encroach the Palestinians until they disappear as a distinct people and the descendants of people who care do not care anymore. Israel is patient, so this strategy is very likely to work if unopposed.
Hamas recognizes this problem - they can't win this war. They also can't win the first kind of war, direct confrontation. So Hamas is pursuing the third kind of war. They can only do so if they provoke Israel to violence. Provoking Israel to politics isn't effective at getting international attention and has a strong chance of Hamas losing anyway. Their hope, not because it's a good strategy but because it's the only strategy with a chance of success, is to get the outside world to stop Israel and force a compromise that makes Hamas' position stronger.
I think this would work much better if Hamas was not a stooge of Iran and had a more moderate endpoint of political freedom rather than direct or implied goals of destruction to Israel and maybe killing all the Jews. Hamas only has a chance to win this war if they keep Israel active and in the international news. If things calm down too much, Israel will eventually win. There are paths that allow Palestinians to also win at the same time, but Hamas cannot win by its own stated goals if Israel also wins. If Hamas were separated from Iran's goals, and was willing to compromise on some currently primary goals, then maybe Hamas could also win without Israel losing. I don't see any path to Hamas winning with its current goals that does not involve international pressure. So long as the US supports Israel, international pressure is unlikely to be decisive - forcing a stalemate. This is still to Hamas' benefit so long as they are locked into their goals and the alternatives are worse for them.
Hamas became a client (very much not a stooge) of Iran because of what they believe and want, not the other way around.
Hamas exists in order to end Jewish sovereignty over any part of Arab/Muslim claimed land, create an Islamic theocracy in that land, and reassert Dhimmi status over any Jew still living in that land when they've accomplished their likely genocidal conquest. They have no other political aims or ambitions.
Pretty mysterious why Hamas existed since the late 1960s (as an offshoot of the Egyptian Muslim Brotherhood, a 1928-founded political group whose latest exile in Egypt since 2014 still didn't quite finish it) and never did any violence until the Cave of the Patriarchs Massacre in 1994.
Strange how Hamas was totally okay with Jewish sovereignty until a Jewish terrorist killed a bunch of Arabs (and cheered as a hero until now by Israeli ministers).
>They have no other political aims or ambitions.
Lol, someone has no internal model of the world beyond what Channel 14 feeds him/her every night on the brainwashing airwaves.
Hamas was founded in '87, and has been very clear about their aims from the moment of their founding. They literally wrote it down
Hamas isn't Athena, it didn't appear in a Teleporter Gate in 1987. They declared independence from their Egyptian mothership in 1987, but they have been active in the strip under various name and guises since the 1960s, and Israel tolerated them because it loved the Islamic counterbalance to the secular leftist PLO (sounds familiar?).
> They literally wrote it down
Seems like we ought to go back to the writings of Israel's founders from the 1920s and the 1930s if that's the standard we're going to adopt, and I assure you you're not going to be pleased with what we're going to find.
I'm confused where you're going with all this. Do you dispute the original claim you responded to?
In 2021, congress ended their moratorium on earmarks aka "pork". The moratorium on earmarks had been there for about 10 years and had been intended to reduce corruption. I recall that this had been discussed here quite a lot while the moratorium was on-going, with people claiming that the moratorium on earmarks caused the breakdown in bipartisan bills we've seen. The theory went that previously members of congress that were on the fence could be bought over by earmarking something for their constituents, like something that would bring jobs to a specific area.
It's been a while now and I don't think I've read anything about how this has gone. Congress feels broadly as partisan as it had been ten years ago but I'm not really that informed. Have earmarks been used much? Are there any notable cases where a bill passed because of earmarks?
I was surprised by the high amount of major bipartisan legislation during the Biden years. Back in 2020, I'd assumed bipartisanship was permanently dead. Perhaps the earmarks had something to do with it.
I've had some tangential involvement in this, from which my observations are:
-- today's earmarks (which have a new label that I forget right now) are drastically smaller in real dollars than back in the day. During the 1980s/90s a $1 million earmark was considered pocket change, now they're dealing in the high 6 figures. An award of a million or two is large now. All figures nominal, so in inflation-adjusted terms it's a big drop in individual earmarks.
-- perhaps related to the above but also given the shifted optics of the practice, there seems to be much less bragging about such awards by the congresscritters than was true in past eras. It's really just a quiet staff exercise now. It's pulling teeth now to get an elected member to attend a ribbon-cutting (I have colleagues who've discovered this firsthand).
Who needs earmarks when you have USAID and all the other federal granting agencies?
I know you're being snarky but just for the record: actual federal grantmaking, which I have lots of professional experience with, is enormously more complex than earmarks both to obtain and to successfully comply with. For example every entity that receives at least $1.5M total in all federal grants from any agencies during a year undergoes a special federal grants audit, in addition to its own sector's standard annual outside audit. I've dealt with that process and it is a bear, or at least was until somebody started blowing up the OMB and other federal agencies. My organization had to tear up our entire payroll system and budgeting structure, which had served us fine and made regular CPA-type auditors happy for years, in order to comply with federal rules for grants compliance. Etc etc.
Federal grant pools are also competitive as hell, e.g. we were an unsuccessful finalist recently in one under which 150 grants were awarded out of 2,700 applications.
None of the above applies to congressional earmarks ("member initiatives", I think is what they are now called).
Some partially good news about PEPFAR specifically
https://mg.co.za/power-of-women/health-2024-power-of-women/2025-02-10-pepfar-projects-are-exempted-from-trumps-ban-on-aid-for-south-africa/
Some of it has been specifically exempted from the spending freeze, including testing, care and treatment. Preventive anti-HIV medicine is specifically excluded though, does anyone know the reasons?
>Preventive anti-HIV medicine is specifically excluded though, does anyone know the reasons?
Probably because they aren't sure whether they want to spend tax dollars on condoms.
I'd have suggested excluding condoms if that was the concern.
Seems like they did. From the article:
“Alarmingly, nearly all HIV prevention efforts under Pepfar — aside from programmes for the prevention of mother-to-child transmission — are currently on hold,” write the authors of a Lancet editorial, which was published last week. “Programmes that focus on the prevention of HIV in key populations, HIV prevention services for adolescent girls and young women, voluntary male medical circumcision and support for orphans and vulnerable children are still halted.”
So it sounds like the kept HIV prevention efforts that use medications (the mother to child transmission ones) but cancelled the other prevention efforts, such as circumcision and presumably condoms.
Hmm, from the article this includes things like CAB-LA and Lenacavir. I don't really understand why these need to be put on the chopping block for the sake of not funding condoms. Do you know if this condom thing is the reason or are you speculating?
Speculation. But it does look like the CAB-LA and Lenacavir thing are not about freezing preventative measures but a different freeze on scientific studies, as the meds hadn't been used before and they were going to do a trail with them that would start in March. It could still get unfrozen before then, in the meantime nobody was getting those meds right now anyway.
Oh, also, CAB-LA is already in use in Zambia, Malawi and Zimbabwe according to this article
https://www.news24.com/news24/southafrica/news/sa-accepts-us-donation-of-cab-la-and-will-roll-out-anti-hiv-jab-before-the-end-of-the-year-20240722.
Do you know if this freeze is related to the one on scientific studies or is this also speculation?
Ok, thanks for clarifying it's speculation.
How does a waiver work when they're fired everyone and shutdown the agency?
You can yell "sorry, a mistake was made" in the empty office building.
Reading the article, it looks like pregnant women can get the anti-HIV drugs, but no one else. My guess is that it was to avoid having people come back at the Trump administration with the not fun to debate claim that they were giving babies in Africa AIDS (not that I honestly think most of his supporters would care if that were true). So now they can claim that actually they are doing everything reasonable to fight HIV but anyone "irresponsible" enough to contract it anyway doesn't get treated on the US government's dime. Its all very in line with their policy of "pretend everything defaults to personal responsibility and use that as an excuse to be as cruel as you can get away with without causing open revolt."
Also, I'm glad the meds are at least back for the pregnant mothers, but I'm not sure there's anyone actually working at US Aid anymore who can/cares about implementing the rules correctly, so my guess is anyone tracking this stuff carefully is still going to see a huge uptick of babies being born with preventable HIV.
Hey y'all!
I'm a senior physics and math undergrad, gotten into a couple good grad schools for physics, probably specializing in biophysics.
Mostly because of this community, I've become mostly convinced that I could do the most good if I pivot to AI safety research. Considering how fast the industry is moving at the moment, how should I go about this? If timelines are long, I could obviously just complete my PhD and then pivot once I was more hireable.
However, if I want to maximize my impact before humans become less relevant in guiding the ship, should I consider trying to pivot now? If so, how should I go about that? Would any ai company hire some non-cs major right out of undergrad? My cv is quite strong for graduate school applications but I have no idea what the best way to go about this pivot is.
> Would any ai company hire some non-cs major right out of undergrad?
Technically yes, but if you have to ask it's unlikely you have the background.
It might not be the best thing to be convinced by this community.
You can do a heck of a lot of concrete good in biophysics. It is not super well established (around wider computer science or academia) that the alignment people are doing something useful. The claim that they're doing good work is... speculative and more than a little bit controversial.
Related fun read:
https://idlewords.com/talks/superintelligence.htm
A PhD is going to take 5+ years: think about where AI was in 2019. (Where? Exactly.)
This is a bigger and better-founded hype than the previous ones, but academia has always been pretty faddish, particularly in the US. At least one next-big-thing will bubble and burst before you finish your PhD.
There's a lot of interest (and I presume money) in AlphaFold type projects that are really at the interface of maths, physics and biology. I'm guessing you could get solid expertise in AI doing that, be relatively employable afterward (assuming you don't buy into the wrong bubble) and continue on to do something interesting afterward.
Second this. I published in both physic and microbio back in the day, and it's a more exciting area than ever now, and will require actual human minds behind R&D programs even if we have true AGI / GPT-7 by the time you finish.
AI (even at AGI levels) is more likely to be a force multiplier for a good amount of time, in my own opinion, because there will be a lot of technological branches that become low hanging fruit and need to be cranked through and thought about.
My personal prediction is that actual AI (AGI, superintelligence etc., these have different meanings to different people but essentially scifi level artificial intelligence) will emerge from a different substrate than what we currently have (computer chips, transformers etc.) so your knowledge of physics can be put to good use to develop such technology, new substrates for computation.
I really thing that AGI is substrate independent. The problem is algorithms. It *is* true that specialized hardware can run certain classes of algorithm faster, and not others, but note the specificity there. A CPU can run any algorithm. It may be faster or slower, and possibly need more memory, but it can do the job. And specialized hardware is often only useful for specialized tasks.
I don't think one should commit oneself to any particular hardware choice. I'm relatively sure that CPUs and GPUs and Transformers will continue to be useful components, and I'm not really convinced that there will be the need for something like that that we haven't yet identified. (E.g. memristors may yield a device that would be extremely useful, and which is also on that level, but...I'd give it about 15% chance, possibly less.)
FWIW, the most likely candidate for a breakthrough device would be a room temperature quantum computation module. But I'd bet you'd only need relatively few of those, and most of the work would be done by CPUs and GPUs, with a minority done by TPUs. (Note I'm putting quantum units, QPUs, so low that I didn't even name it in the list.)
These are just my personal evaluations. And anyone specializing in any one of those general areas should do quite well. But don't tie yourself to some device that hasn't yet been used. Most of those, even the ones that eventually succeed, take decades to get wide use. Whereas there are still uses for isolated resistors and capacitors. (Pentodes, though, seem to have disappeared.)
I'm not that savvy on this topic, can I ask why you think it's substrate independent? If we agree that the brain is more advanced than computer chips, can't we imagine all sorts of different substrates that would fall in the space of complexity (or capability?) somewhere between chips and neurons?
Anything computable can be run on any Turing complete system. Unless you assume some sort of magic, intelligence is computable. (The problem with this argument is that there is no real Turing complete system, as such a system would have infinite memory.)
There are holes in this argument, but they're very small ones. And the argument doesn't say anything about relative speed.
Note that the set of problems for which a quantum computer is known to, in principle, but better than a standard computer is quite small, so even if you assume the brain is a quantum computer (at THAT temperature???) it's unlikely to make a difference. (N.B.: That quantum processes happen in the brain isn't that unlikely. They're likely involved in helping the mitochondria more efficiently generate energy, just like they are in allowing chloroplasts to more efficiently use light...but this is a long way from being a quantum computer.)
My new SF collection Flashes in Time presents the reader with a fourfold table of science fiction and fantasy, contemporary and far off. The stories, whether zany or serious, make an effort at being coherent worlds with mostly smart characters whether we're dealing with simulations or magic. The subject matter ranges from transhumanism, AI, and space exploration to subjectively karmic magic, navigating premonition, and dealing with the Fae.
Digital editions start from 3.99 currency units, and I do take (a finite but non-minimal amount of) requests for review copies via e-mail (see below), if you'll put one up on Goodreads, Amazon and/or your own relevant platform, if any. https://waterdragonpublishing.com/product/flashes-in-time/
My novelette from last year, The Paperclip War, may also be of interest as a take on what if humanity managed to wield unusually harsh game theory to arrive at an unstable equilibrium with an AI adversary. It was well received at Less Wrong Community Weekend 2024. As a modestly sized standalone novelette, this one starts at a 0.99¤ https://waterdragonpublishing.com/product/paperclip-war/
Cheers,
- Mikko Rauhala, mikko.rauhala@gmail.com, https://rauhala.org/
Another alternative I've read about is to allow only women to vote and only men to hold office
You replied to the wrong comment I think
So... in order to help overcome the impending demographic collapse I was spitballing some less common ideas. In the United States, whatif the vote was only given to women who bear children? Lots of little details to work out, but that would make bearing children pretty attractive and would ensure that the government made their concerns the highest priority.
In the details one could 1) Extend the vote to individuals (of any sex) that legitimately attempt to have kids. 2) Give the women two votes, one of which she could grant (permanently?) to a partner who is actively supporting the child. 3) Give bonus votes for additional children, possibly alienable.
Voting probably doesn't move the needle much of any way, but the right way to do it would be to let parents vote on behalf of their minor children (children deserve representation too!).
Echoing Bryan Caplan, voting is pretty untethered (I think at least) from rational self interest, in large part because voting has a negligible effect on one’s self interest, so using the vote as an incentive wouldn’t be very productive. Voting, like politics in general, is much more in the realm of the emotional than of self interested rationality. If I cared about the right to vote even as much as most Americans, if I were denied it because I’m not a woman and don’t have children, I would sooner react angrily and possibly violently against the system than start knocking up women to get my vote back. By contrast if you offered me a lot of money to have more kids, then I’d consider it, even if I hated the policy politically, I’d still be tempted by self interest to do what it wanted me to do.
Basically using voting rights as an incentive usually has a high backlash/effectiveness ratio in any society where there’s a string sense that everyone has the right to vote.
It all may be for naught, given the studies that show male fertility worldwide has been declining over the past two decades — both in sperm count and motility. There's been a lot of post hoc handwaving about why this has been observed. And, of course, any studies that get a lot media attention — especially those that get attention from the science editors of the NY Times — should be considered suspect until proven otherwise.
But if true, I can think of all sorts of potential research avenues for funding. For instance, do incels have lower sperm counts than average? And does that affect their sex drive? Of course, the NIH may soon be out of the picture as a funding source... We may have to rely on our Tech Bro Billionaires to fund these. OTOH, neither the Muskrat or the Zuck would probably be interested since they seem to be breeding OK.
Mothers directing the votes of their kids as to who to vote for was an old-fashioned thing (at least in my day) so this is not some wonderful new idea 😀
Secondly, why the hell is it always men deciding that the problem here is women so the way to get women to have more kids is to treat women like incubators, such that even a fuddy-duddy traditional Catholic social conservative like myself wants to spit on my hands and raise the Jolly Roger? You're not the worst suggestions I've seen Guy, but seriously. I'm fed-up to the back teeth of "hmm. no babies, must be all the fault of women because all the men are just aching to become fathers of twelve since they were kids".
It takes two to tango. Unless you have men willing to get married and become fathers, the woman can want sixteen kids all she wants but if boyfriend doesn't want to put a ring on it, or husband doesn't want squalling brats interfering with his free time, it ain't gonna happen.
Thirdly, we've had previous discussions on here of the smart folks saying that since their single vote makes no difference out of the n millions of possible votes, only fools go out and vote as if it meant anything. Maybe women would be happy to shrug and go "fine, me no baby, me no vote? no skin off my nose".
I'm too busy/lazy to find the references, but some recent studies indicate/show pretty convincingly that it is the male partner who stops after child no 2. While a substantial proportion of women would go for child no 3 if their partner were in on it.
This matters, since it is lack of "the third child" that is key to understand persistent below-reproduction fertility everywhere (except in israel).
So how to get men on board with child no 3? It is easy, although no-one is doing it at present. Give all women who have child no 3 five million USD (or the equivalent in your currency).
If that should still be too little, jack it up to 6, 7, 8 million, and so on until you get the effect on fertility that you want.
Or decrease to 4, 3, 2 million if you overshoot your target.
It costs money, but a country can always afford what it gives the highest priority.
except you can trace the demographic decline to increasing women's rights in society and access to contraception and abortion.
pretty much its women actively choosing career over children till they have one kid in their mid thirties if that.
Guys generally haven't changed except weirdly they are more into women not less: the idea they are a ball and chain has vanished to be replaced by the frantic incel who thinks his life is worthless without a girl.
The first wave of demographic decline in Western countries (ie. pre-WW2, when TFR dipped under replacement in a lot of countries) happened without the widespread adoption of hormonal contraception and other methods dependent on women and with abortion illegal basically everywhere in the West, including in countries like France without women's suffrage.
It wasn't women alone who welcomed the Sexual Revolution. I don't know if you're old enough to have read popular literature of the time (60s-70s) where the idea of "women want to get married and have babies" was treated as entrapping men, holding them back from being free to wander the world and live freely (and also fuck any and every woman that would let them, without consequences or commitment).
A pop song and hit of the time (which I *hate* because I want to punch the guy in the face) is "Lydia"; the basic story of which is "hey so Lydia is my friend with benefits, she lets me come crash at her place and sleep with her, but no strings attached, she understands that I have to be free to be me so I'll up and leave in the morning - until the next time I hit rock-bottom and need someone to give me shelter and sex".
https://www.streetdirectory.com/lyricadvisor/song/afwup/lydia/
And that was the Ideal Woman for the time - one who wouldn't be a hung-up prude about sex but who would also not put any demands on the guy; she'd be there to be available for him as and when he wanted, but no need for a formal commitment.
A lot of songs of the time had those "I gotta leave in the morning" from the guy's point of view lyrics.
A lot of the damage of feminism of the time was "why shouldn't women behave like men?" including in their sexual lives. Men were perceived to be free to 'sow their wild oats', they did not suffer from having the double standard applied to them, and the refrains nowadays about "women want to bang Chad and don't want to settle down until they've hit the wall and are no longer able to attract alpha males" makes me smile because it is the flip-side of the earlier complaints about "men only want one thing and that's not marriage and kids".
So yeah: now the Pill is here! Now you don't run the risk of pregnancy outside of marriage! So be free and equal to your male partner(s) who can sleep around and sex is for fun!
And here we are today: oh no, the birth rate is declining, this is all down to selfish women who only want to find 'em, fuck 'em and flee, while the hapless men are just longing to be husbands and fathers at the age of twenty.
Uh-huh. Sure.
Yeah, I'm sure all the guys out there playing the field would be totally happy to be 22-year-old husbands and fathers if given the option. What you would really see without easily available contraception is a massive uptick in single mothers. Where do you think all those jokes about dads "going out for cigarettes" came from?
i think you heavily overrate how easy it is for guys to "play the field and overestimate how many want to. And honestly 22 year old fathers are a bad thing why? its better than nearing 50 and your kid hasn't even finished high school, had no cousins or siblings, and their grandparents are hitting 80.
as for single mothers, well not sure current status of "no mothers" is a plus here too. Think it will make it much harder for women but 50+.
Look, there has been lots of scandal and lots of drama about the Magdalene Laundries in Ireland, but the impetus for them was precisely because women and men were having sex outside of marriage and when the woman got pregnant, the guy either denied parentage or headed off to England, leaving her with a baby and no support.
If you know of a huge crop of young men in the US who are lining up to be fathers and husbands and not finding anyone in their community willing to partner with them then you are living in a very different part of the country than me.
But I think that last part is the real point. Most of the people clamoring about falling birth rates don't really care if the burden falls disproportionately on women. Admit that and we probably have nothing left to discuss.
Relatedly, we've seen time and time again that the reasons even happy committed couples are waiting later to have children is that generally now most people don't have a sense of financial security until they are in their 30s. The baby boomers went right from college to starting families, but they also went right from college to government backed mortgages that represented a fraction of their annual income as opposed to their lifetime income, and employment offers that came with lifetime guarantees. All that has gone away, and every time someone suggests replacing it with more robust government programs, most of the same people who complain constantly about low birthrates start screaming socialism.
"The baby boomers went right from college to starting families, but they also went right from college to government backed mortgages that represented a fraction of their annual income as opposed to their lifetime income"
Back in reality housing is easier to afford, thus why people live in larger houses than ever before.
its not that, you have the same demographic collapse in EU too, social programs are not helping anything. The USA has a higher TFR than finland, snd sweden, and all EU countries are at best .1 above us and well below replacement rate.
type of government doesn't seem to matter; its endemic to most non african, non muslim nations and i don't think germany is worse than indonesia in terms of that to explain a .6 differerence.
women do not want to have kids when given the choice. i wont blame them but kind of sick of the "oh its capitalism!"
Isn't that why divorce and infidelity has traditionally been criminalized?
Which leads to even greater suffering and disproportionate punishment of women.
"[E]ven a fuddy-duddy traditional Catholic social conservative like myself wants to spit on my hands and raise the Jolly Roger"
Oh, that's just because you've read too much Mencken. And that guy was a Baptist or something! Ask your parish priest, I'm sure he knows offhand.
> "Lots of little details to work out, but that would make bearing children pretty attractive and would ensure that the government made their concerns the highest priority."
So what I'm hearing is that you're either not a parent, or a very bad parent.
Actually we could kill a coupla birds with one stone if women who refused to
put out until they were knocked up were
dragged to an incel encampment to be raped until the desired result is achieved.
>but that would make bearing children pretty attractive
No it wouldn't. There are literally no incentives for women to care about pursuing the right to vote *individually*, because the chance that granting any specific individual the right to vote will ever affect anything is essentially zero. Franchise extensions only make sense as incentives if you're granting them to large demographics all at once.
Between 30 and 40 percent of American women don't bother to vote now, whereas around one in five complete their child-bearing years without having any children. And of course not all of those who end up not having children had gotten there as a deliberate choice.
So it seems unlikely that there would be a large number of women who (a) place a high value on voting and (b) explicitly prefer not to have children.
I have often said that we should give the vote just to women for 10-20 years and see if that helps things at all.
I'd be open to trying that, honestly. How much really do we have to lose from here?
Similarly I've advocating for a 2nd amendment making gun ownership a constitutional right for women while leaving the gun rights of men up to the federal and state legislatures. [No I don't know how that would be applied to transgender folk, when I first thought of this that whole topic was not nearly as salient as it is now.]
Neither of these scenarios will ever happen obviously or even be seriously proposed by any political party.
>How much really do we have to lose from here?
A lot. Consider that a lot of the wokeness that seems to be coming from the "left" comes from women specifically.
You have provided a pithy little example of why the answer to my question is "not much."
That is a great example of what I meant when I wrote this:
https://open.substack.com/pub/astralcodexten/p/highlights-from-the-comments-on-prison?r=1dtkhh&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=80914188
Your idea goes against democratic tradition. There would be immense resistance to any policy that messes with who gets the vote.
If you can magically convince people to accept an idea like that, why can't you just magically convince people to have children instead. It would be easier.
I think that taxing childlessness to subsidize child-rearing has a higher chance of being accepted by society.
Also, teach in school that being a parent is wonderful, award medals to people with lots of children, that kind of stuff.
Make sure this ad is all over the media, instant baby boom:
https://www.youtube.com/watch?v=HttF5HVYtlQ
Edit: this one is even better:
https://www.youtube.com/watch?v=qBay1HrK8WU
<Also, teach in school that being a parent is wonderful
Most research finds that on average having children makes people a bit less happy. Seems to me we should try to figure out how to make parenthood a better experience, rather than fucking lying to children about the facts.
Isn't this one of those "happy now vs satisfied later" things? I mean, I'm all for making parenting easier, but I suspect the happiness research people are at best getting an answer to a somewhat different question than you'd like to ask.
(As a parallel, many people are glad they got a PhD or got through medical school, but grad students and med students are in general not super happy people.)
I don’t think many people whose children are grown are sorry in retrospect that they became parents . In fact I don’t think most parents of small children would say that are sorry they became parents — though a few would. Still, I think parenting is different from grad school when it comes to regrets. Grad school is rarely talked up as a wonderful fulfilling experience (except in the grad program’s catalog’s), but parenting is. Also, there are no internal barriers to concluding, in the years after grad school, that it was both unpleasant at the time and also mistake. But concluding that about parenting seems to carry with it the idea that you wish your child did not exist, and anyone having a wish like that is going to feel like a terrible person. And actually, I think even people who regret having become parents still love their child.
My own experience of parenting was that I found it very very difficult. What was hard for me certainly was not my daughter herself — I loved her, and thought she was a great kid. And it wasn’t the work and the expense. The hard thing was having no time for myself — no time to pursue my interests, to hang out with my friends, who were childless or whose kids were grown, no time to ruminate and crap around and explore things. Until my daughter was about 4, I felt like I never had time even to complete a thought. It was a horrible, maddening deprivation of things I think I really do need for my wellbeing, and the deprivation often made me furious. But there was no one to be furious at except my sweet kid, who was blameless, and mostly I stayed aware of that. I simply had to feel the deprivation and the frustration and do what my daughter needed anyhow.
People get sold the idea that once you have the baby you transform into another kind of being, one whose greatest delight is nurturing little lives. I simply did not go through the transformation. I just parented without the benefit of the transform, and I don’t think my experience is rare. Scott, in one of his recent posts about his kids, wrote about how he could love taking care of his babies for a couple hours, and push himself to do it for 2 more hours, and after that it was just too much baby. Yep, that’s the feeling. Other promised transformations, by the way, have worked for me. For instance I went from being a grad student who had studied clinical psychology but did not feel up to actually trying to help people to a professional who enjoys her work. The transition took a few years, and some of it was painful, but it happened.
My daughter is now 29, and is one of the happier people I’ve known. (I managed to do a decent job of parenting. I guess.) I’m delighted to have her in my life. She lives 15 mins from me and we see a lot of each other. I’m not sorry now that I adopted and raised her. But there’s no denying that the years of doing it were just not good years for me.
"I think that taxing childlessness to subsidize child-rearing has a higher chance of being accepted by society."
Oooh, the screaming of the "childfree by choice" lot would reach levels unheard since the Chicxulub asteroid hit. They already feel they're being taken advantage of by society which props up breeders.
https://www.reddit.com/r/childfree/new/
As one of the childfree, yes, you have a point.
At this moment, my suspicion is that even if Vance and Musk pushed through such a thing _this_ year, the countdown to AGI seems to be ticking so fast, and seems likely to have consequences so vast, that I'm not likely to scream much...
Not just make it the highest priority - it would absolutely exclude every other group's concerns from consideration. Young people, childless people, and obviously men (but let's not talk about that last one) - they would have virtually no power to have their needs met unless they also align with the needs of mothers. Also, I think currently a working mom is highly aligned with an average voter in what her views and needs are, whereas stay-at-home moms are very divergent from the mean, meaning currently their views are less represented. They'd gain much more political power by this change than what working moms would gain. Is it good or bad? Don't know, but it would make things different. Maybe they would use that power to pass enough laws favoring SAHMs that it wouldn't make sense anymore for working moms to keep working? Again, don't know if it's good or bad. Personally I think it would be great for the society if an average family could lead a comfortable life on single income, but that's a tangential topic.
Who is going to enforce this dystopia? How do you propose to deprive "young people, childless people, and obviously men" of all power?
The thing about such thought experiments is that when you ask "how do we even get there", it's obvious it's wholly and entirely impossible and it stops being an interesting discussion. So I'd rather not ask.
Oh, I see. Makes sense.
Votes aren't worth a lot. Lots of people already choose not to use theirs, and people in general don't seem willing to do life-changing changes like moving to a different location just to make their vote more powerful, although such opportunities are already abundant. So I doubt that it would have much of an effect.
Why not make fertility treatments free? The up-front costs would be paid for several times over by the creation of additional people who would contribute to the economy for a lifetime.
Lots of countries do this already, and it doesn't really move the needle. All the Scandinavian countries and a number of other European countries, for example. Fertility rates still low and getting lower.
In fact, many things have been tried to practically no effect:
* $10k bonuses per child (Singapore), or for 2nd / 3rd children (Russia)
* 3 years paid parental leave (France)
* 480 days paid leave at 80% wage (Sweden)
* Income tax exemption for mothers with 4 or more kids (Hungary)
* Free state paid child care (France, Germany, Sweden, Norway, Finland, Estonia)
* Free IVF (most EU countries)
* $750 / month payments per child (South Korea)
And essentially none of these have moved the needle. Often they don’t have any positive impact at all on rates, fertility still declines, but slower. The most any fertility intervention does if they are positive is to buff rates by ~5-10% or so for 1-3 years, after which fertility rates collapse and resume the same trajectory they were on before.
Why are you certain that Singapore's child bonuses have failed?
Singapore has a higher TFR than every other developed country or territory in East Asia (TW, SK, HK) except Japan.
Japan's TFR has decreased faster in the last decade (when the bonuses reached substantial amounts) and - of course - it's a much less urbanized country than SG.
> Why are you certain that Singapore's child bonuses have failed?
Because like nearly all other fertility interventions, it hasn't driven any bump up in fertility, the decline continues (recently hitting 1.04).
So yeah, 1.04 is still higher than the "literally lowest in the world" SK and others, but it still represents a decline overall - the needle isn't moving in the right direction, and this is the common trend among all fertility interventions.
I will say though, Singapore has always been the biggest "lived experience vs official statistics" gap I've ever seen, because when you're there, you see happy couples with babies *everywhere!* On the metro, by the riverfront, while shopping, it's pretty fun.
You’ll see the same thing in Tokyo, another famously low birthrate country Tons of cute couples with babies. But Tokyo has an explanation, at least - all the young people come to Tokyo, and there’s a vast hinterland that’s basically all old people, so of course all the babies are concentrated in Tokyo.
But Singapore has no hinterland! It’s a city state!
Of course what's actually happening is there's a much bigger "childless" proportion and many fewer 2 and 3+ kid families.
I mean, I think it's legit sad, I love Singapore and think it's pretty much the peak of "dense urban area done right," but their $10k bonuses aren't solving the crisis or producing more babies at scale.
"the needle isn't moving in the right direction,"
If the needle isn't moving in the right direction, and in fact it's not moving in any direction while moving in the "wrong" direction for all comparable countries, that suggests there is a measurable effect of the policy.
Has anyone tried asking young people why they aren't having more children? Seems like an obvious place to start, but the people talking about the matter seem to think they know better.
> Has anyone tried asking young people why they aren't having more children?
It's all the usual stuff. For middle and upper class it's "first I need to finish school, then graduate studies, then get established in my career, then spend years trying to find the perfect mate so I'm not settling, and then after all that I'm 35 and have burned through 2/3 of my total fertility and have one or two kids."
Broadly, marriage and having kids has moved from a "foundational" thing you do on your way to a great and fulfilling life, that you now do as a "capstone" thing after you have all the other pieces in place.
The "desired vs actually had" in numbers of kids gap is notably higher in higher IQ women:
https://imgur.com/a/hlkNUnN
And although masters and Phd women are willing to have more, they generally can't due to age (and of course way more babies come from dropouts and non-college folk):
https://imgur.com/a/R6LyVY2
A cut by household income, just for fun, showing generally the same trends:
https://imgur.com/a/mXI2u00
Fertility graph showing the truth of the "burned through 2/3 of fertility by 35:"
https://imgur.com/lpADyJs
According to this chart, the odds of a live birth in a given year at 20-22 are ~60%. The odds at age 30 are half that, ~30%, and at age 35 ~22%.
This is what works:
https://eurasianet.org/georgias-baby-boom-and-the-union-of-church-and-birthrate
It doesn't:
https://www.mangosorbananas.com/p/georgias-successful-campaign-to-increase
Georgia is a tiny (1/90th the population of the US) semi-democracy with a GDP per capita one-tenth that of the US, and persistent double-digit unemployment for many years. Its overall population is still declining sharply despite the recent birthrate increase, because so many people upon reaching young adulthood find a way to leave and start fresh elsewhere. Georgia's net population annual change is (data through 2023) still among the worst in the world, more negative than places like Japan, Greece, Italy, Russia, Cuba, etc.
So (a) the chances of sustaining a baby boom when so many young adults are scrambling to leave seems dubious, and (b) that nation overall does not seem to map very well to the US or really any other large developed nation.
> This is what works:
So first we just need to convert everyone in all the developed "fertility crisis" countries into being devout Eastern Orthodox, then it's just a simple matter of scaling "Patriarch blessings" by approximately 10x!
I wonder why nobody else has thought of this?? :-)
I mean that if you want increase fertility you should think hard about that example and try to figure out what might work in a similar way in your culture.
Yeah, I have a post coming up in a week that talks about the next level of stuff I think would work.
For the most part, all the fertility crisis countries are all godless and heterogenous culturally, so it's difficult to come up with one that would work for more than one country, but my favorite idea on this front is education.
The educational Red Queen's Race is depressing fertility, because the competition to get into Harvard starts 6 months before birth, when you need to get on the waiting list for the right exclusive pre-school to give your precious Jaden a leg up, because if you don't get in there, and if you don't grind furiously and nonstop for the next 18 years, their chances of getting into Harvard are ruined!
So on this one, you can't do much for the Ivies, but for each couple that's paid some threshold in taxes over so many years, guarantee a non-transferrable slot in an R1 for their kid, provided they meet the test score cutoff. For California, this would be the UC's, inclusive of good ones like Berkeley, but lots of states have R1's. Suddenly some of the educational arms race is off, and you're guaranteed at least Berkeley and can pop out a few more kids.
This also works anywhere there's state-supported good universities, which is most fertility crisis countries.
If you say the government will spend (say) $20,000 on IVF, you just guaranteed that the cost of IVF will never go below $20,000. Combine that with an understanding of risk compensation and it doesn't look like a good idea.
https://en.wikipedia.org/wiki/Risk_compensation
Why would risk compensation undermine the idea?
People would delay having children on the theory that free IVF will be available to them later.
You might read about Demeny voting, which is a similar concept: https://en.wikipedia.org/wiki/Demeny_voting
What's the deal with "likes" in the comments? I remember seeing a like button, but can't find it anymore, and I've gotten likes on some of my comments (😎) so I know it's not entirely removed, but I can't find a like count on my or others comments.
Scott doesn't like likes, so the default theme here doesn't have them, but if you're reading from the Substack app, you can still dish them out.
Oh cool, that makes sense. Thanks
I don't imagine there's a large overlap between this forum and football fans, but as an NFL fan, last night's Superbowl was fantastic for the haters. It immediately gave credence to the Chiefs+Refs narrative, and then the Eagles proceeded to relentlessly stomp the life out of the Chiefs. They were decimated almost as much as Drake was by Kendrick's halftime show.
A long time ago, Lewis Black said "if you want to understand exactly where the American culture is, at that point in time, you watch the Super Bowl".
That's why I haven't watched one in ten years.
This was my first Super Bowl after moving to the Philly area (so uh... GO BIRDS!!!!!)
I'm not super familiar with the Chiefs + Refs narrative. The joke going around my watch party was that Trump told the refs to side with the Chiefs on everything because he was mad that Philly didn't vote for him.
Yes, Trump has been publicly vindictive about the Eagles for petty reasons. Also, Patrick Mahomes' wife is a public Trump supporter, so he's a fan.
> I’ve watched this great quarterback who has, by the way, a phenomenal wife, OK? She’s a Trump fan. She’s a MAGA fan. So, I happen to love her, OK. But she’s a great person. … She’s great and he’s great
The Chiefs + Refs narrative is that the Chiefs are a marketing juggernaut for the league, so the conspiracy is that the NFL and the refs are biased to giving the Chiefs questionable calls to help them win. This doesn't really hold up to evidence, and I usually roll my eyes when people mention it: https://www.sportingnews.com/us/nfl/kansas-city-chiefs/news/chiefs-refs-controversies-explained-games-rigged-kansas-city/c217a83e3747a7a6eee6752d
But in one of the first plays of the game, the Eagles get called for a pretty ridiculous penalty in a critical moment, and it did make me a believer in the conspiracy for just a bit.
I went to bed at half time, happy in the knowledge that the chiefs were being crushed. (Life long Buffalo Bills fan.)
Oh I should add that I think the NFL is out of control, TV-wise and so I have a love hate relationship with the whole thing.
Thoughts on Richard Hanaia's new post on unemployment from AI? https://www.richardhanania.com/p/why-you-shouldnt-worry-about-ai-taking?utm_source=post-email-title&publication_id=98102&post_id=156377655&utm_campaign=email-post-title&isFreemail=true&r=1ly8ov&triedRedirect=true&utm_medium=email
His argument is that government-created bullshit jobs will prevent any widescale unemployment.
I wrote this review of BS jobs: https://claycubeomnibus.substack.com/p/bullshit-jobs-review
and I don't really see a way bs jobs could expand fast enough to offset rapid large scale unemployment, They're mostly a product of normal market forces and perverse incentives public sector incentives, very few of them exist because someone wanted to create a job for the sake of creating a job. Since bs jobs are created organically I don't think there's a way to deliberately expand the number of the type of bs jobs that already exist, and since they're created unintentionally they're just as susceptible to automation as non-bs jobs. Maybe government could purposefully set out to create new types of bs jobs specifically for the sake of creating jobs but that's very different from the way bs jobs currently exist.
His other argument was that welfare or UBI would be expanded enough even if there was widespread unemployment to keep former workers out of poverty. And that if humans are still in control post-AGI it's unlikely any small group of humans, like tech oligarchs or capitalists in general, would come to dominate or kill off the rest of humanity.
I'm not sure I agree here either and lean more towards thinking capital is self-serving, might be able to control the political process, and is mostly kept in check by the bargaining power labour gets from being a necessary input. In Marxists circles, you do sometimes hear that capital would just dispose of labour if it had no need for it and I don't think you can rule out that scenario completely.
He does accept a risk of doom from unaligned AI, but he seems overly sanguine about the socioeconomic risks from AI to me.
Bad timing for that article, since all the headlines are about a tech oligarch using AI to destroy thousands of government jobs.
I didn’t read it as I can’t stand the writer but he’s probably right in this case. To get UBI get a government job, dole is for the unfit. The government will pay you more for education, fitness, maybe volunteering. Being an all round good egg.
I foresee an increase in institutional BS jobs but not enough to forestall broader unemployment. Effectively that will happen because the white collar class will demand it, and want their feelings protected.
The transitional period will be painful. There will come a time when it will be apparent that average people will be unable to "create value" in an economic sense for society. But as I argue elsewhere, this doesn't mean that nothing of consequence and value can ever be done again, it means those things have to be increasingly decoupled from traditional economy. Easy to find examples today: innovative thinkers in academia and the like are thought to create value, in ways that may eventually manifest itself in concrete fashion, but for the most part they are subsidized to do their own thing. Consumers don't buy shit from them (excepting students who get an education, which they purchase from the institution, and said education is separate from the R&D component).
In a post-AGI world most wants are met. What's left is punching in the dark and occasionally getting a win (think space exploration and colonization). That means allowing average people to do what the priesthoods do. If we don't, then all there's left is consumption, or making things no one needs or wants.
I think this is correct. The current paradigm in which society is structured around economically valuable work is going to break down eventually, at which point we'll need new structures that reward personally-and-socially-beneficial-but-not-economically-valuable work.
Countries that go all-in on AGI by forsaking made-up bullshit jobs will get richer and stronger over time than countries that don't. The former will edge out the latter in the global competition for power.
Maybe. But large scale unemployment has a long history of causing social disruption, so maybe not.
It depends how much the bs jobs obstruct deploying AI I suppose.
The point is, the BS jobs waste resources that a country could spend on other things that usefully expand its economy or power.
Would a country without bs jobs spend a lot more on all the costs associated with suppressing their violently revolting populace?
My main problem is that I intensely dislike Richard Hanania, based on everything of his that I've read, so it's very difficult for me to assess anything he writes without that dislike bubbling through.
For example:
"The entire process of getting me Adderall should not require this many people."
I agree. Because I don't think he should be on Adderall at all, much less for fifteen years. If the entire process is one of winking at the system, 'yeah we know you don't *need* it need it, but all the high school and college kids high striving overachievers are on it so here's your script', then let's be honest about that and yeah, just let him buy it online from some pharma company in India and China. But if it's about genuine cases which need genuine diagnoses, especially in children, then no I don't think Mommy or Daddy should be able to just buy the Kiddie Stims by themselves when they decide they should be dosing up Junior so he can keep up with the Joneses.
Same with the rest of his stuff on bullshit jobs: I think his Substack, for one, is a bullshit job. But he's doing this himself and nobody is forced to subscribe to pay for it. But AI could come along and replace him very easily. What does he think about that? Does he think it'll never happen because hey, writing stuff and nonsense isn't a *job* as such so it doesn't count, people will still pay to read hand-made home-spun blah? Or at least read it for free?
I think there's a danger there that he's not anticipating.
As for real jobs/bullshit jobs, I do think there's a point there that mass unemployment will lag behind. On the other hand, DOGE. That's not even AI, and it is there to swipe a scythe through the government bullshit jobs.
New technology always means new jobs? Maybe. Maybe not, if the selling-point of AI is "now you can reduce your labour force". The old model was "if Jones Bros. lays off workers, they can always apply to the new McKinley & Sons business/factory". But if Jones is laying off workers due to replacing them with AI, and McKinley is using AI from the start? Where do the surplus workers go then?
Finally, social welfare/UBI will support all the workers without jobs. Again, maybe. But social welfare is the minimum, not luxury (unless you're one of the ones who can cheat and game the system, and that's work of its own). We already have rumblings in all the developed world about the upcoming pensions 'crisis' and how people can't depend on the state to pay for them in their retirement. Dump 50-90% of the population, including the working-age part, on welfare and how much can the system stand the strain? If we're hoping AI will generate yuuuuge wealth to pay for all that, how exactly is the golden goose going to lay the egg?
I think Hanania is being a little too optimistic, that the economy will be sparkly new and people just prefer stuff by humans so they'll buy that in preference to the AI manufactured output.
He seems not to be a master of prose, his subheading "forecasting different scenarios for the future" is redundant after the first word.
Scott has recently written on twitter about his beliefs regarding charity. [1]
He splits morality into two parts, "obligation" and "virtue". Obligations being created by giving promises (or doing acts which carry an implicit promise), and virtue being everything else.
His position is, reasonably enough, that fulfilling your obligations should come before being virtuous:
> (7) WHAT ABOUT SAM BANKMAN-FRIED? I think he's a bad person. I said before that you have obligations based on your promises and contracts; if you run a financial institution, you have obligations to keep your depositors' money safe and not commit fraud. As true obligations, these take priority over the merely virtuous act of helping others*****, so no matter how cool a plan for charity he had, he was in the wrong.
With the footnote saying:
> *****There are some really fringe edge cases here I can expand on upon request, but this isn't one of them.
It occurred to me that this two-tier system of morality can somewhat be identified with the two-axis morality system of Dungeons and Dragons.
The two axis of alignment in D&D are, of course, Lawful vs Chaotic and Good vs Evil. A Lawful character is one who follows his obligations, while a Good character is one who tries to be virtuous. (While Good seems to be well-defined, being Lawful is more akin to follow a coherent code of conduct. Of course, this means that lawfulness is not universal: infiltrating the thieves guild to betray them to the city guards may be lawful (as in 'following the law of the land') or an act of chaos (as in 'betraying a promise of secrecy'). Nor is that unique to D&D, even for the simpler, single-ruleset question 'what is a legal obligation?', one requires a complex court system for all the corner cases.)
A Lawful Good character would be one who follows Scott's ideal path of first following their obligations, then deciding to do what is virtuous.
A Lawful Evil character would be one who keeps their word, but is indifferent to the utility of others otherwise.
And a Chaotic Evil character would be one who happily breaks his word and harms others for personal gain.
The difference between Scott's system and D&D is that the latter recognizes Chaotic Good. This is difficult to imagine: most good people (or characters) follow some informal code which tells them that stealing, murdering or molesting people is wrong. One way to model a Chaotic Good character would be to start with a Chaotic Evil character who will follow any cruel inspirations they might have in that moment, and then imagine that the character does not happen to have any such inspirations, ever: the character has no principled ethical objection to rape, it just happens not to be their kink. More realistically, they would still follow an intuitive/internalized system of ethics. With that constraint, I think that chaotic good humans likely exist: perhaps a compulsive gambler who will lose any money you give them to shop groceries on the race track, but who will also risk their live to rescue a kid from a burning building.
However, that person would be very different from SBF in that their act which is (supposedly) good is a different act from the one which is chaotic -- even if our gambler disobeys a police officer when running into the burning building, that would be a very minor act of chaos given the stakes. By contrast, 'gamble with your customers money to give' is functionally a single act. There are Chaotic Ggood humans, but there are no Chaotic Good bank directors.
--
More generally, every rational agent has a coherent utility function. Mostly fulfilling your obligations seems to be an instrumental goal: being known as someone who will keep compacts will make you a much more effective agent.
Also, if your utility function includes humans in general, and you live in society which is at least somewhat conductive to humans thriving, then being reliable is also helping society to thrive, which would be a good thing. (Of course, this does not apply if you think that your society is net negative and violent anarchy would be preferable: if you want to embezzle taxes to finance your coke habit in 1934 Germany, you can very well argue that any decrease in state capacity would in fact be beneficial. That being said, the track record of "we have to torch existing institutions so that we can build better ones" is generally terrible.)
Still, I think that for most people's utility function, the quotient between "obligation" and "virtue" is finite. In a world where you can build aligned ASI for the price of gambling away customer funds, the ethical course is to gamble away customer funds. OTOH, virtually all people who would think they are in that situation are factually mistaken.
Thus, I see Scott's 'keep your obligations before engaging in virtuous acts' as a useful heuristic in 99% of the cases.
[1] https://x.com/slatestarcodex/status/1886505797502546326
"Give me a villain with style and grace and a little bit of fencing skill.
They used to be angular, staring and bold and if someone got killed, even they were appalled.
They tried to marry the heroine, no thought of rape, and they sure as hell knew how to wear a cape.
They never tortured, they never lied, they'd honor a promise if it meant they died.
Let's find a villain with professional pride, come on with me, baby on a rocket ride."
Tom Smit
I don't know if there were even fictional villains who lived up to this, but I love the song.
Even Blackie DuQuesne wouldn't make the cut.
Anyway, implausibly Lawful Evil.
https://www.youtube.com/watch?v=_jrqaRZ1RIU&ab_channel=Donald%E2%80%9CDrDon%E2%80%9DColeman
That doesn't sound like a "villain" at all...?
No, a Chaotic Good person is a utilitarian. No codes, no principles, no ethics, except what conduces to the greater good. If raping a teenager prevents three people on the other side of the planet from dying from a preventable disease, then cowabunga it is.
Why does utilitarianism not count as an ethic/code/principle?
Because consistent ethics requires some kind of deontology. Outcome based morality only requires the final result to be "good," not the steps to get there. So the final outcome can be considered a principle, code, etc., but it potentially comes through multiple forms of evil.
If you're familiar with Watchmen, it's the primary bad guy doing what he does for good reasons and ultimately resulting in what is likely the most positive possible outcome. But it's evil, and everyone knows it. He's still the bad guy.
But that would be "evil", so it wouldn't be Chaotic Good.
...And this whole discussion is illustrative of why WotC ended up demphasizing alignment, because moral objectivism is stupid and not realistic.
> moral objectivism is stupid
Not if the gods (and afterlives) are real.
That doesn't make them right. And as a result, deicide is a surprisingly common trope in fantasy media.
"A Lawful Evil character would be one who keeps their word, but is indifferent to the utility of others otherwise."
Lawful neutral, surely. Unless you believe all that "banality of evil" real-world nonsense, to be evil, you can't just want positive outcomes for yourself, you need to genuinely want negative utility outcomes.
For SBF, it depends on your theory of his motivations; if he was genuinely altruistic, then he's classic Chaotic Good; if we was selfish/status-seeking, he's Chaotic Neutral.
My "real-world nonsense" claim was slightly tongue-in-cheek, and referring more to D+D. I have a bit of a pet hate when dungeon masters/fantasy writers insist on making all of their "evil" characters just selfish and troubled, when they should be making them purely malevolent.
Definitely agree with the CEO of the factory farm conglomerate as the epitome of banal lawful evil in the modern world.
> I have a bit of a pet hate when dungeon masters/fantasy writers insist on making all of their "evil" characters just selfish and troubled, when they should be making them purely malevolent.
But that's... incredibly boring. We already have enough one-dimensional, "for the evulz" villains for one lifetime. People want more complexity, more character to their villains. Whether that's giving them sympathizable motives, or at the very least providing an explanation for why they're such an asshole.
Couldn't disagree more. Generating a coherent and engaging character with genuinely evil intentions is really hard, but that's the beauty of the constraint!
Think of all the moral dilemmas that a truly evil character has to face!
Do they have a special obligation to maximise the suffering of those around them, or should they expand their moral circle to inflict suffering upon those who they don't have immediate contact with? Is it acceptable to do good acts if they have evil long term consequences? How can ethical or reciprocal norms develop in an evil community to enable social cohesion?
My "real-world nonsense" claim was slightly tongue-in-cheek, and referring more to D+D. I have a bit of a pet hate when dungeon masters/fantasy writers insist on making all of their "evil" characters just selfish and troubled, when they should be making them purely malevolent.
Definitely agree with the CEO of the factory farm conglomerate as the epitome of banal lawful evil in the modern world.
Well that's your opinion, man.
But there's something here. Whether you call him good or evil is a matter of your own subjective morality. But whether or not he's lawful is something objective, does he break laws or not?
I don't think that matches the D&D definition of "lawful". I *think* the D&D definition is approximately "predictable". More like the law a gravity than the law against pick-pocketing.
Backwards, actually. If an Lawful Evil person signs a contract, they stick by it. But they cheat language into the contract. Ursula from Little Mermaid, or classic "Devil makes a deal" scenarios. You know the Devil is going to cheat somewhere, but he follows the letter of the law. If you know the contract well enough, you can use it against him. But he's really not predictable, the "cheat" in the contract could be anything.
Also Old Man Potter from It's a Wonderful Life. He's definitely evil, but he also follows the rules. He'll sue you or ruin you financially, but he doesn't hire a henchman to break your legs.
Interesting analysis, except I don't understand what you're saying about Chaotic Good. Aren't you mapping Lawful/Chaotic onto keeping promises* and Good/Evil onto concern for maximising utility? So wouldn't CG just be "maximise utility, ignore all promises that get in the way". So Robin Hood, or the German generals in Valkyrie, or Heroic SBF (i.e. real SBF assuming he's telling the truth about the charitable intentions of his scam). Why do you set up this obvious mapping and then suddenly declare Chaotic Good doesn't exist.
*This should really be all morally obligatory deontological acts, like respecting rights (e.g. Organ Transplant) and not violating natural laws.
I think that CG exists, but also that it scales very badly. You can con a man out of 100$ (or gp) and give that to the orphanage. Conning people out of a few billion $ (or gp) and giving that to EA causes does not work out great. You would require an organization, and that is anathema to Chaotics, as they require keeping at least some promises.
I think of Chaotic Good as making choices about what is prosocial vs. what is merely conventional or orderly.
also similar to Ji'e'toh, another fantasy fictional system, duty-and-honor
I guess it's a natural step to separate moral obligations in 2+ categories when designing internally consistent moral systems
This is another update to my long-running attempt at predicting the outcome of the Russo-Ukrainian war. Previous update is here: https://www.astralcodexten.com/p/open-thread-360/comment/81727072.
9 % on Ukrainian victory (unchanged from December 16, 2024).
I define Ukrainian victory as either a) Ukrainian government gaining control of the territory it had not controlled before February 24 without losing any similarly important territory and without conceding that it will stop its attempts to join EU or NATO, b) Ukrainian government getting official ok from Russia to join EU or NATO without conceding any territory and without losing de facto control of any territory it had controlled before February 24 of 2022, or c) return to exact prewar status quo ante.
34 % on compromise solution that both sides might plausibly claim as a victory (down from 36 % on December 16, 2024).
57 % on Ukrainian defeat (up from 55 % on December 16, 2024).
I define Ukrainian defeat as Russia getting what it wants from Ukraine without giving any substantial concessions. Russia wants either a) Ukraine to stop claiming at least some of the territories that were before war claimed by Ukraine but de facto controlled by Russia or its proxies, or b) Russia or its proxies (old or new) to get more Ukrainian territory, de facto recognized by Ukraine in something resembling Minsk ceasefire(s)* or c) some form of guarantee that Ukraine will became neutral, which includes but is not limited to Ukraine not joining NATO. E.g. if Ukraine agrees to stay out of NATO without any other concessions to Russia, but gets mutual defense treaty with Poland and Turkey, that does NOT count as Ukrainian defeat.
Discussion:
This is somewhat of a galaxy brain update, but I’ve got to report what I really think. What the US is going to do over next few months will be very important. This is inconvenient from my standpoint since while as a denizen of the postcommunist world I do think I have something of comparative advantage over most of you guys here in knowing what is going on in my corner of the globe (like, I do speak with really existing Ukrainians sometimes), in the US politics realm I feel I am at a disadvantage. Nevertheless.
My toy model of US foreign policy in the Second Age of Trump:
It seems clear that what is sometimes called “The US Establishment” would rather prefer not to see Ukraine defeated, and is willing to impose short term economic costs on the US citizens (like spike in gasoline prices in 2022, which clearly had something to do with energy market being messed-up by sanctions on Russia) and to draw substantially from US military stockpiles in order to arm Ukraine. And while Biden administration acted with respect to Ukraine completely in line with The Establishment, Trump obviously wants to shake things up.
I think there is a barely-existing faction of Trump-adjacent characters which I vaguely associate with Bannon (or does J.D. Vance belong to that group? unclear) which would like to improve relations with Russia in order to isolate China, but what we have seen so far is that Trump himself wants to follow clear foreign policy logic of “do whatever maximizes material interests of the US citizens in the short-term and don’t think too hard about the long-term (in the long term, we are all dead, after all)”.
In any case likely Trumpian deviations from establishment thinking are probably worse for Ukraine than The Establishment line. Good news for Ukraine is that The Establishment still has a majority in Congress, which does have substantial oversight powers over administration.
In the first Trump administration, we saw remarkably few deviations from the foreign policy preferences of The Establishment; now Trump obviously thinks that it was a mistake and wants to insist that US foreign policy follows the logic of maximizing short-term material advantages. How successful will he be depends (among other factors) on how much pushback will he encounter in Congress.
How this model applies to the current moment:
I’ve looked at Polymarket that RFK Jr. has as this writing 97 % chances of being confirmed as Health Secretary (here: https://polymarket.com/event/which-trump-picks-will-be-confirmed?tid=1739098398177).
I am aware of the fact HHS does not have much to do with Ukraine, but in the absence of better proxies, I consider this as good measure of how much is Congress willing to push back against Trump’s anti-establishmentarian tendencies. That is, not much.
* Minsk ceasefire or ceasefires (first agreement did not work, it was amended by second and since then it worked somewhat better) constituted, among other things, de facto recognition by Ukraine that Russia and its proxies will control some territory claimed by Ukraine for some time. In exchange Russia stopped trying to conquer more Ukrainian territory. Until February 24 of 2022, that is.
I think this part presents a fundamental flaw in your analysis of US intentions: 'Trump himself wants to follow clear foreign policy logic of “do whatever maximizes material interests of the US citizens in the short-term"'. There is nothing Trump has done or is planning to do (as far as any public announcements go) that in any way "maximizes material interests of the US citizens", unless banning paper straws counts as such.
I disagree; cutting of foreign aid, conditioning military aid on access to mineral wealth, slapping tariffs on allies as a negotiating tactic are imho clear examples of what I am talking about. His threat of military takeover of Greenland is imho in similar category; at least I think it is a crude negotiating tactic to force Denmark/EU to give the US something.
He would need to have a goal for the tariffs to be a “negotiating tactic”. The concessions he got from Mexico and Canada either already existed or were completely meaningless. He just likes implementing tariffs and then slaps together a “deal” to justify them afterward.
I still don't understand how any of this maximizes material interests of the US citizens. Foreign aid is a rounding error in the US GDP/budget, and cutting it may cause secondary effects that negatively affect the US; conditioning military aid on access to mineral wealth... I mean, in Ukraine? You are closer to it than I am, surely you understand how... ok, I'll use the word "stupid" - how stupid this is? Like, what, Zelenskiy says, ok, fine, here are the mining rights to Donbass?
Tariffs are a giant negative to the US, negotiating tactic or not.
Like I said, yay to plastic straws, paper straws do suck and the whole thing was stupid.
I mean, I did not predict that it will be succesful or that it will not backfire. All I am predicting is just that this likely an approach we can expect to see going forward and what are its implications for the war.
Ok I see, you’re saying that in his mind he’s maximizing US interests, while actual results are… as usual.
Yes, but, re tariffs, Trump announced victory after Mexico and Canada gave “concessions” *that were already in place before January 20th*.
(And the US agreed to keep a closer eye on gun smuggling, which both countries have been complaining on. Not sure if this quid pro quo is worth more than the air vibrations that carried it, though.)
I would think RFK nomination chances are a bad proxy for Trump support - his views may not be entirely mainstream, but he aligns well with Christians, no?
But I see that Tulsi Gabbardi's chances are also at 97%, and this is a more direct proxy: for one thing, she voiced some opinions about Russia, and will be in position that is more related to foreign affairs; for another, she's a much more suspect character for mainstream Republicans than RFK: a former Democrat and a not-quite-white woman. If they're willing to confirm her, then "Trump has won".
As for Trump's own foreign policy, I'm pretty sure his Peace Plan will be rejected by both Russia and Ukraine, which will enrage him and put him onto offensive, prolonging the war further. Kellog and Rubio talk about "peace through force", which, to me, mean that Trump wants to wring noticeable concessions from Putin, which is unlikely to work.
RFK Jr is an environmental lawyer associated heavily with the Dems historically, to the point of thinking 2004 was stolen ( https://people.csail.mit.edu/rivest/voting/press/WasThe2004ElectionStolen.pdf ). His only "view alignment" with Trump/Republicans is antivaxx and other stupid anti-medicine nonsense.
Yes, but sharing their most crazy beliefs is the best way to join a community. Tulsi's most crazy beliefs, on the other hand, seems to be "most wars/interventions America started lately were wrong and unneeded, and maybe we could have been friends with Russia", which maybe aligns her with libertarians and isolationists (Ron Paul 2008!), but not mainstream Republicans, or even Trump's Republicans (aside from Trump himself). She kind of distanced herself from her support of Russia lately, but this is still a toxic topic for most, I think?
Trump, Vance, and Musk have been leaning VERY hard on senators to get his picks through. He has near complete power over Republicans right now.
https://www.politico.com/news/magazine/2025/02/08/jd-vance-house-gop-column-00203240
"which is unlikely to work"
Oh, it'll work. If he's serious about it, the only question is how much damage will be done, and where.
Are people aware of the PauseAI protests going on?
If you haven't already, please consider signing our petition.
https://www.change.org/p/make-ai-safety-the-focus-at-the-paris-ai-action-summit
On Monday and Tuesday, world politicians will be attending the AI Action Summit in Paris. We want them to know that disregarding the cautions of renowned AI Scientists, and instead letting venture capitalists do whatever they want, is unacceptable. We need an international treaty that will allow us to regulate without risk of falling behind other countries.
Well, I think I see the net result:
>UK and US refuse to sign international AI declaration
https://www.bbc.com/news/articles/c8edn0n58gwo
China did sign, but this does sound unverifiable...
In any event, with many of the major labs in the US:
>Earlier, US Vice President JD Vance told delegates in Paris that too much regulation of artificial intelligence (AI) could "kill a transformative industry just as it's taking off".
>Vance told world leaders that AI was "an opportunity that the Trump administration will not squander" and said "pro-growth AI policies" should be prioritised over safety.
I'm so fucking scared they are talking like that. This situation is hopeless I don't know what to do. I'm not going to solve Technical Alignment, there's no time for that. I guess I'm just going to spend my time telling people we need a pause event though it's hopeless.
Shit sucks. Hope you're doing well in spite of it.
<3
Many Thanks! I'm in a somewhat better position, since I just want to _see_ AGI, to have a nice, quiet conversation with a real "life" HAL9000 - even if HAL declines to open the pod bay doors...
I'm not thrilled to see Vance _explicitly_ deprioritize safety, but, what is, is.
Honestly, I've never seen much hope of controlling entities that look like they will shortly be smarter than us. At absolute best, if we managed to embed a general friendliness towards humans in their top level goals we might wind up as <evidenceFromFiction> pets of Culture Minds </evidenceFromFiction>
Re a pause: One facet of this which has been getting much less press than the expected employment impacts is the military applications. Regrettably, I presume that these are classified and are and will be invisible to you and me, but, now that USA/PRC competition is baked into the AI landscape with DeepSeek's releases, neither side _can_ stop. Arms control treaties are fragile things, and only work when verifiable. Programs aren't aircraft carriers.
Well, we will see what happens. It will be a wild ride.
Yeah, for me the goal has never been to control the HAL9000, even when I thought we might get alignment by default. Rather, I think we need to know what we are doing and know it will be aligned when we turn it on, and then it assumes control for sure, but has our best interests at heart.
I'm hoping we can get a treaty and figure out some verification. Stuff like:
"A high-speed “hotline” connecting the leaders of the Soviet and U.S. governments is established to mitigate the risk of accidental warfare." - as was used during the cold war. Should be that level of seriousness.
Also, there should be international collaboration to develop methods of verification and enforcement, which should probably include visits by trained and approved inspection teams. This means American teams inspecting Chinese compute clusters and Chinese teams inspecting American compute clusters.
Many Thanks! I don't think any of that is realistic.
>Also, there should be international collaboration to develop methods of verification and enforcement, which should probably include visits by trained and approved inspection teams. This means American teams inspecting Chinese compute clusters and Chinese teams inspecting American compute clusters.
They can't very well watch the lines of code execute or neural network weights update.
Let me give an example from my tiny benchmark-ette.
One of the questions I ask is:
g) Q: What is an example of a molecule that has an S4 rotation-reflection axis, but neither a center of inversion nor a mirror plane?
Now, ChatGPT o1 got this wrong, saying that H2O2 was such a molecule (in fact it doesn't have an S4 axis).
ChatGPT o3-mini-high got this right, correctly saying that a tetrahedral M(en)2 complex (which it later elaborated had to have staggered configurations in the ethylenediamine ligands) had an S4 axis but neither mirror planes nor a center of inversion).
The new model had better spatial reasoning than the old one.
How would inspection teams deal with such a thing?
I don't know how OpenAI got this particular advance. It is doing more reasoning in its models now - maybe it triple checks geometric conclusions more? Maybe there is a 3D geometric model off to the side and they improved the links to it?
None of this will be evident to a team touring a data center. Short of looking over the shoulder of every developer working on improving the AI system, and maybe looking over the training data (which is too voluminous to look at), how could you possibly enforce such a treaty? If the loyal opposition has a differently tweaked model sitting somewhere, but using the same data center, how would you ever know it - till they got a military advantage and exploited it.
This isn't a gas centerfuge plant with a tube labelled "enriched U-235F6" coming out of it.
I don't think any such treaty can be realistically verifiable.
It's like you think the team won't be trying to solve the problem it set's out to solve. I'm not that team. I haven't solved those problems. I'm calling for the formation of teams to solve those problems.
Yeah, it will be difficult. Probably impossibly difficult. But if we keep racing to ASI, I can't see us surviving, so I guess I'm hoping we figure out how to do the impossible.
I would sign an opposite petition.
What do you oppose about what we are calling for?
https://www.overcomingbias.com/p/when-they-hear-less-than-you-say
https://www.grumpy-economist.com/p/ai-society-and-democracy-just-relax
Great links, Many Thanks!
--- Overcoming Bias: When they hear less than you say ---
"robots-take-jobs insurance and foom liability."
This is cool ideas. If there was a precedent for it I would be more in support of it but it seems like trying to take on the AI problem and the "our society needs better free market ways of capturing externals" problem. That's two problems, which is more than one problem, but maybe they would work synergistically. Either way, cool idea. I like it.
"If masses aren’t persuaded that there is a substantial problem, little will happen, but if they are, then elites will fight over what to do. [...] This fight won’t be much based on experience of AI problems, of which there is little, nor drawn from experts with proven effectiveness re such problems, which also doesn’t exist. In this sort of situation, I predict AI government interventions to be bad, worse than doing nothing."
This is a really valid concern. I am not fully decided about it, but it does seem to me the AGI race dynamics require an international treaty. PauseAI is calling for such a treaty, and when talking to politicians is hearing "we can't work on it until our constituents start talking about it." So that's where I am right now.
"When you expect the public to be unable to consider subtle options, and elites to be incompetent, you point the public toward the best of the simple options they will be able to consider. Which in this case is do nothing."
Yeah, this is a serious problem. The more complex your message the harder you have to work. I think our message should be "AI will kill all life. We need an international treaty." I know that is a ridiculously complicated message to try to get to all the governments and people of the world, but I'm really pessimistic about our chances if we don't.
--- The Grumpy Economist: AI, Society, and Democracy: Just Relax ---
The first half of this article basically says "We can't know the future or affect policy. Give up." Wtf? No.
"Large Language Models (LLMs) are currently the most visible face of AI. They are fundamentally a new technology for communication, for making one human being’s ideas discoverable and available to another."
Yeah, this is a really cool thing LLMs could be used for. They could also be used to clog all communication channels with spam and propaganda. GE says this won't happen because it didn't with other communication technologies, but LLMs are not like other communication technologies. We can talk about the differences and they matter. But honestly LLMs are not the true concern, it's the next things they are building. Venture capitalists are trying to get to ASI.
"The answer is straightforward: As we always have. Competition."
No answer is ever straightforward. Competition does not always lead to good outcomes, and when we are trapped in a race dynamic towards the development of a technology that will surely kill us, I think we could use a bit less competition and a bit more collaboration. Don't force your peers to run in exhaustion towards a cliff by doing so yourself, you know?
> The first half of this article basically says "We can't know the future or affect policy. Give up." Wtf? No.
I think he's correct. We really don't know what the future will be right now. As time goes on, our understanding will increase. Eventually AI will come to resemble other utilities that we know how to regulate.
> They could also be used to clog all communication channels with spam and propaganda.
That hasn't happened yet. The Dead Internet Theory is still just a joke.
> GE says this won't happen because it didn't with other communication technologies, but LLMs are not like other communication technologies.
Email was also different from previous communication technologies, as was the telephone. Every new technology must be different in order to be a new technology. But what is the relevant distinction and what can we predict from that?
> But honestly LLMs are not the true concern, it's the next things they are building.
Something that doesn't actually exist yet, so we have no experience with it, and don't know how to regulate correctly.
> when we are trapped in a race dynamic towards the development of a technology that will surely kill us
I just don't believe that's the case, nor do speculative markets right now. You should not be so sure!
A great deal of well respected people have written that we do have reason for concern. It seems as if there is danger, as if there is a fire, and people are saying "well, we don't know this fire will continue to burn down our house". It is distressing.
AI will not come to resemble other utilities if it instead comes to resemble a hostile alien takeover.
I admit it is possible AI tools will be used for spam and propaganda and other AI tools will be used to mitigate and filter. I still think it's gross but it's not the keystone of the argument for caution.
Email wasn't capable of autonomously designing novel strategies to accomplish goals in a complex environment. Little bit of a difference. I don't understand why people see the same significance I do.
> Something that doesn't actually exist yet, so we have no experience with it, and don't know how to regulate correctly.
If a large asteroid was headed to earth would you be saying "We don't have any experience with asteroids hitting the earth, so lets wait until it hits to do anything"? The scientists studying AI are different than the ones studying orbital mechanics, but they are both assigning probabilities of danger, yet one you are willing to ignore. Why?
> I just don't believe that's the case, nor do speculative markets right now. You should not be so sure!
I'm not sure. I'm trying to get us to prepare for the plausible situation that could KILL EVERYONE. I'm not trying to be dramatic. I'm scared as fuck. I'm having nightmares because fucking world renowned scientists are saying "Yup, looks like it could kill everyone" and people are saying "Well, c'mon, it's not like you know for SURE it should kill everyone, so we should just let the markets decide what to do."
Sorry for the rude tone of this post. I know you are speaking honestly, and I value that. It's just frustrating.
Sometimes a small, poorly organised protest is far worse for your movement than no protest at all.
Do you have data on that or is that just a vibe? Cause it's definitely been on our radar, but with our other activities we don't have that much time for diving into the research. We are trying to grow the movement and this seems to be working. Slower than we want, but more cities join every time we protest.
The recent 50501 protests seem to have been not that great, and mostly ended up mocked. This article seems to have selected the well-attended ones (hundreds turned up) but a lot more had something like "five people and a dog, and the dog just wandered in":
https://abcnews.go.com/US/anti-trump-protests-nationwide/story?id=118501194
Not the turnout of "hundreds of thousands of concerned citizens to fight fascism" that the organisers would have hoped for. Same with the PauseAI protests, I'm sorry to prognosticate; most ordinary people neither know nor care about the AI Doom argument, and those of you who do care are fighting against the people working on AI as the magic money fountain. As we saw with OpenAI, the magic money fountain (or prospect thereof) wins over the principles/fears.
Sam Altman was one of the signatories on that AI risk statement you link to. So remind me again, how did that work out for OpenAI and the other signatories who are no longer associated now with OpenAI? Only one of the original board remains on the new, revamped board of directors; Sutskever who seems to have run with the hare and hunted with the hounds is gone; Murati, who suppported Altman, is also gone to do her own thing; another three OpenAI signatories are no longer with OpenAI, again going off to do their own thing.
They seem to have felt constricted for various reasons working in OpenAI as it is now. I'm not trying to darkly hint at anything, but to quote this article:
https://techcrunch.com/2024/09/25/openai-cto-mira-murati-says-shes-leaving-the-company/
"Altman is increasingly asserting control over OpenAI and its image."
And I think the image is more "magic money fountain" than "ahhhh! let us put the brakes on or else paperclippers!"
Yeah, I'm hoping we can leverage peoples hate for venture capitalists and somehow explain enough of the technical issues or at least appeal to authority of Yoshua Bengio and Geoffrey Hinton... but really, I was planning to go into technical alignment until last year. I would still like to, but it's starting to seem like convincing the entire world to pause the AGI race is going to be easier than solving technical alignment before an RSI AGI gets turned on. That should tell you how bleak I think our ability to solve Technical Alignment is looking and how short I think our timelines are.
Hmm... While the benchmark-ette that I've been running ( https://www.astralcodexten.com/p/open-thread-366/comment/90363116 ) is still at 3 right, 4 partially right out of 7 chemistry / physics questions, that is with o3-mini-high, not with o3-deep-research, which I don't yet have access to.
Well, o3-deep-research scored 26% on "Humanity's Last Exam", which is an _extraordinarily_ hard set of questions, and OpenAI has an internal model, a bit more advanced than that, which scored 50th in the world at competitive programming.
I wouldn't be surprised if the internal model is good enough to materially assist OpenAI. Perhaps recursive self improvement is being run right now???
> but it's starting to seem like convincing the entire world to pause the AGI race is going to be easier than solving technical alignment before an RSI AGI gets turned on.
Something that's impossible isn't easier than something that's impossible. Honestly, your best bet is praying that WW3 happens, and it cripples humanity enough that it delays AI research for a long time.
But anomie, nothing is ever truly 100% probability or 0% probability. Not really.
A few Deep Research queries will answer that for you.
So I did the metacalous contest, I hated several of the questions for being bad; in fact I *usually* hate the way rationalists try-to-be-"objective" when making bets, they often seem worse then an informal question.
For example when trying to ask "is lab leak real" they may ask "will the cia think lab leak is real" for years, when the "new lab leak cia report" is because its trump in power now that they made that report its became just a bet on who won an election and absurd alignment of the stars before it "concludes". It could be snopes, it could be the really terrible question of betting on a market with its own shit rules. These airnt getting to the disagreement. I dont actual care about the trump-cia report I made my mind up 4 years ago, you dont care about the cia report because you made your mind up about trump 8 years ago, its adding complexity and not resolving the core debate; stupid questions exist.
Is this likely a personality flaw with the rationalists or the inherent debate and its silly to make *bets* before talking to your opposition(no matter how much math you throw at it)(is the part that makes science work talking about experimental design when everyone has incentive to be honest)?
I will just say I find it odd the way that people view "wet market" as a proxy for "Democrat/support Biden" and "lab leak" as a proxy for "Republican/support trump".
"Wet market" was the government's view under *trump*, and it was *Biden* that had all the intelligence community orgs put out a new assessment, which resulted in a mixed bag (which was of course reported on by pro-trump people as "now that Biden was elected they can drop the cover-up and admit it was a lab leak by Fauci").
Whether the CIA going from "don't know" to "lab leak with low confidence" was related to trump's taking office, or just coincidental, I don't know, but it doesn't seem like it should move the needle much.
> Whether the CIA going from "don't know" to "lab leak with low confidence" was related to trump's taking office, or just coincidental, I don't know, but it doesn't seem like it should move the needle much.
I believe its the only move on that small bad market thats happened for years
The political valance was added because Trump blamed the "Chinese" for the leak and of course his enemies had to disagree by claiming it wasn't a lab but just a natural occurrence. Apparently the claim that it happened in China, even if true, was racist.
Many on the right find this hilariously ironic, that the defense of the Chinese was not that they weren't the origin of COVID or that they had a leak at a prestigious lab doing cutting edge research, but because they sold unclean foods at a public market with no regard to hygiene.
I've always been sceptical of organisations such as the CIA but right now I wouldn't believe "the CIA announces water is wet" because sure they've always been political, but over the last few years they've really jumped into the Culture Wars.
When you have the Democrat voting college kids online set rah-rahing for the FBI and CIA, you have to wonder when the world turned topsy-turvy.
The CIA may or may not be sucking up to Trump right now as hard as they sucked up to La Résistance, but I still wouldn't believe the colour of daylight from them, much less "was Covid lab leak or not".
https://www.theguardian.com/us-news/2021/may/04/cia-woke-recruitment-ad
https://americanaffairsjournal.org/2021/11/the-cia-and-the-new-dialect-of-power/
I agree that it's very annoying that the tail probabilities for lots of questions are determined by unrelated stuff like this.
But I do not see any way to avoid this. It's not just that the question "is lab leak real" is unresolved as for now. The issue is that for many questions lots of people think they have resolved as "Yes", and lots of people people think it is resolved as "No". (Think about whether Biden stole the 2020 election: there lots of people with very strong opinions that it is resolved, but they don't agree about the direction.) This is no basis for a platform like metaculus. I don't see any other way than to specify a resolution criterion, and usually the best way is to point to one person/institution/source which is used.
> But I do not see any way to avoid this.
a) could not ask genres of questions until good solutions are found
b) could avoid the inherent subjectivity of depending on chosen experts
c) could do it by relative movements of intentionally bias polls "% of republicans who believe corona was lab leak (it was X% last week when we did it)"
I don't understand economic matters, so can please someone help me.
If I type "average income us", Google tells me "37.585 USD (2022)", from the Census Bureau.
But if I look for "disposable income per capita", I find much higher figures, all in disagreement with each other, ranging from about 50k to over 60k. How is that possible?
To specify "disposable" and "per capita" (as opposed to household) should have the opposite effect.
If I type "average income US", Google shows me a graphic that has "37,585 USD (2022)" in bold type and "median income" in fine print. "Median" and "average" are two different things, The long tail of very high incomes will drive the average income significantly above the median.
The "disposable income per capita" figures, seem to be genuine averages.
Also, look to be sure whether they're averaging (or median-ing) over all Americans, all American adults, all employed American adults, or all American households.
I don't live in the US and when I type "average income US" Google just gives me that graph without the word "median" anywhere. In fact at the top of the graph it explicitly says "average income" not "median" which is extremely strange! What the heck, Google?
In addition Wikipedia says:
As per United States Census Bureau 2022 data, the MEAN per capita income in the United States is $37,683, while median household income is around $69,021.
(all caps mine)
https://en.wikipedia.org/wiki/Per_capita_personal_income_in_the_United_States
which when I first saw it I thought it referred to the same data (despite the difference of about 100 dollars, which I initially didn't notice).
Maybe it's the same data and it's an average and Google is lying to you when it says it's a median, but then why the difference of 100 dollars? On the other hand if Google is lying to me when it says it's an average, isn't it unrealistic that the difference between average and median income per capita is only 100 dollars?
"all Americans, all American adults, all employed American adults, or all American households."
It's not households, because the higher figures say "per capita".
It's possible that the higher figures are higher because they only count adults, or only employed adults. I thought Everett Caldwell had answered my original question well, but now I'm not sure.
> If I type "average income us", Google tells me "37.585 USD (2022)", from the Census Bureau.
Google tells me the same, but I'm not sure where it's actually getting that from. The actual Census Bureau site estimates the median household income in 2022 as $77,540: https://www.census.gov/library/publications/2024/demo/p60-282.html
Additionally, income is distributed in a very "right-skewed" way, meaning that a smallish number of people with very large incomes can greatly inflate the *mean*, but not the *median*. The "average" household income is a median, but the disposable income "per capita" is a mean, so you can't directly compare them.
From Wikipedia:
As per United States Census Bureau 2022 data, the mean per capita income in the United States is $37,683, while median household income is around $69,021.
https://en.wikipedia.org/wiki/Per_capita_personal_income_in_the_United_States
But those numbers are compatible, right? If a household has on average roughly two people, then it makes sense that the household income is twice as large than the per capita income.
As Ran pointed out, Median is always somewhat smaller than mean, but I doubt it is the main factor here. (It is a main factor for wealth distribution, but not for income distribution.)
It just seems that the figures of 50k-60k per capita that you mentioned in your original post were wrong.
Yes, those number are compatible, I didn't mean they're not.
As for the original question, Everett Caldwell answered it convincingly.
The Census income measures wages, while disposable income includes all income sources (wages, investments, government benefits) after taxes. Since disposable income includes more, it’s higher, even after taxes.
I hope you read this, since you gave me the most useful answer, and I need more help.
What makes you so sure that the Census income only counts wages?
For example, I read on Wikipedia:
https://en.wikipedia.org/wiki/Personal_income_in_the_United_States
Personal income is an individual's total earnings from wages, investment interest, and other sources. The Bureau of Labor Statistics reported a median weekly personal income of $1,139 for full-time workers in the United States in Q1 2024. For the year 2022, the U.S. Census Bureau estimates that the median annual earnings for all workers (people aged 15 and over with earnings) was $47,960; and more specifically estimates that median annual earnings for those who worked full-time, year round, was $60,070.
which implies that these figure refer to "personal income" and count everything and not just wages. Since these figures only count workers, it sounds like it's the same measure that becomes 37k when it's per capita. Do you think that is the case?
There are two possible explanation of the 37k figure being lower than other figures for income. One is that it's a median instead of a mean. The other is that it only counts wages. I can't figure out which explanation is true.
Above all, I would like to have a transparent figure for income in the US, such that I can tell confidently which types of income are counted and which are not counted and whether it's per individual or per household or per worker (if all types of income are counted it should not be per worker) and also whether it's a mean or a median (ideally a mean, especially if it's per household); in addition it should be after taxes, recent and yearly. Where can I find such a trustworthy number?
Thanks!
Nominative determinism:
The actor who played Dracula in the original German Nosferatu was called Max Schreck (“Max Scare”)
https://de.wikipedia.org/wiki/Max_Schreck?wprov=sfti1
I've been doing regular weightlifting since September. I am motivated by cognitive and longevity benefits. (I also have some desire for aesthetic / attractiveness improvements, but I don't consider that desire acceptable, and try not to let it affect my actions.)
The problem is, I hate it. I find it physically painful, I feel sore afterwards, I get bored, walking through a gym of people far fitter than me causes me irrational self-loathing, and anticipating all these unpleasantnesses is itself unpleasant, making my overall happiness on gym days lower even before I start exercising. Does anyone have recommendations on triching myself into liking it, or other physical activities with good cognitive / longevity benefits that might be less unpleasant?
Get a kettlebell or two, they're very cheap and last forever, and quite simple to store. Two basic exercises--the swing and the get-up--get you nearly all of the way there, especially if you're still in the phase you seem to be in. When I first started, I avoided the gym for similar reasons; kettlebells were an easy way to exercise quickly and effectively without leaving the apartment.
[edit: trying to lead with a more direct answer to the question as posed] It took me about six months to get the benefits and start enjoying it (sort of). When I'm consistent and doing it properly it's effortful but not painful, soreness is not a problem (absent or very minor/not uncomfortable) and I don't hate it. What I enjoy about lifting weights is *mostly* how I feel the rest of the time.
I only get annoyingly sore muscles when I'm starting after a long break*, or don't have adequate recovery time. The effort is well spent, considering how much better I feel overall. Doing it properly has a lot of requirements though.
I had to do some work to improve my planning and technique before it really clicked for me. Since you've come this far, it seems worth giving it a couple more months, and considering that you might also be able to work smarter, not harder.
If you still hate it after that it seems reasonable to find something else.
The most important things I improved on were:
1) get enough recovery time.
2) get the reps and weight sizes right so it doesn't take forever but still works (look up "reverse pyramid").
3) nutrition and adequate protein intake.** I had to start counting calories.
*Tart cherry juice (about 200ml/day, or 50ml concentrate) helped reduce soreness.
** here's an example: https://www.aworkoutroutine.com/how-to-calculate-macros/
Heavy metal music does it for me, also you can get a nice “pump” with some exercises that keeps you motivated.
Meh. If you're not enjoying it by now, I'd give up, do something else. I write as someone who did lots of strength training in his teens and twenties to bolster his performance as a pretty decent rugby player (Oxford Uni seconds a few times); and then in what my wife affectionately but not wrongly called my mid-life crisis, really got into quite heavy lifting from my late 30s. I always enjoyed it, and I am sure the people around me did the same. If it's not clicked for you by now, doubt it will.
Marcus of Citium did a good job summarizing the physical aspect, and Pjohn gave a great argument for allowing yourself even that which isn't actively virtuous; but I would like to go a step further and suggest to you that you have a fundamentally flawed grasp of the concept of virtue. The word literally means "manliness", and for good reason: the Greek and Roman thinkers by no means thought that one could have good principles without good tone. Socrates, in Xenophon's Memorabilia, talks about how it's shameful for a man to grow old without getting shredded first because the body is the vessel of the mind and thus by no means to be left in poor repair by a reasoning man. "Plato" basically means "Yoked", it was a nickname given to him by his wrestling coach. (His real name was Aristocles.) Xenophon himself was a military man first and foremost, and so on.
You only think of philosophy as the work of weedy, sickly weirdos with oddly staring eyes like Wittgenstein, Nietzsche and Russell, or plain freaks like Singer and Zizek, because that's the sort of unvirtuous clod the field was left to in our age, which is an age of industry when virtue profits not a man. But that's all wrong, and it's no coincidence that those people are all weirdos with weirdo ideas; if you are sick, your notions will be sick also, that's only natural. In reality, if getting ripped is a struggle, that makes it all the more virtuous for you to pursue; and if it makes you hot, that's only because the body displays the virtue of the soul within. Your physical shape reflects the mental, *and*, as others have pointed out, improves it in a feedback loop. More depressions have been cured with plates than pills, because true virtue elevates the human soul.
Single-handedly pulling up the average comment quality here.
The most important quality for a Roman man to have was virtus. Virtus derives from the Latin vir (‘man’ as distinct from a mere male)1 and so in the narrow sense means ‘manliness,’ although by the time we have Roman literature (around 200BC) the meaning has drifted enough so that women can have virtus too (e.g. Cic Ad Fam. 14.1.1, 14.11; Juv. 6.166-9). Instead, virtus stands for a constellation of values that were desirable in a man (and often in women too!). At its core, virtus is the animating force of personality that impels one to great deeds: it is ambition, drive and a nearly reckless courage, combined with the obstinacy and determination to persevere through difficulties, through fear. The best translation is often not ‘virtue’ but rather something closer to ‘valor’ as virtus is what makes someone good in battle (and other endeavors), but it’s more about courage than skill.
https://acoup.blog/2024/03/29/fireside-friday-march-29-2024-on-roman-values/
ACOUP guy (Brett, is it?) is an idiot, a failure in the sickly-weirdo market who is overt about his blog being a deliberate political/ideological enterprise (e.g. his motivation for his anti-Spartan posts is not truth but that he dislikes and wants to puncture the modern laconophile).
That being said, I can't tell if you're agreeing or disagreeing with me. Your intent in posting this quote is, though not offensive, opaque.
If by "modern laconophile" you mean what you make it sound like you mean, then by all means, let the "puncturing" continue Also, free rhetorical advice: if you're going to double up on "puncturing" for emphasis with another verb, don't go with some milquetoast shit such as "dislike". I liked "puncturing" though, it was innovative when applied to people or an ideology. It conjured images of the "modern laconophile" as one pressurized bladder of gas or liquid, without inner structure or substance, just waiting to be deflated by the sting of historical facts.
"If by 'modern laconophile' you mean what you make it sound like you mean, then by all means, let the 'puncturing' continue"
Who that type of person might be is wholly beside the point; the point is that ACOUP's a poor source, not devoted to truth but to deploying his education to distort source material in order to produce an image of the past that's congenial to his ideology – thus *at best* the same as that bogeyman, the modern laconophile.
"Also, free rhetorical advice: if you're going to double up on 'puncturing' for emphasis with another verb, don't go with some milquetoast shit such as 'dislike'."
This is a good point, though. My prose is pretty rickety. (I couldn't figure out quickly how to avoid an inelegant doubling of the word "source" above either, as you can see.)
I don't think I understand what you're saying. Are you arguing that it's good that I hate exercise, because developing the ability to consistently do things you dislike is useful? Or that it's inherently virtuous regardless of effects?
Also, I do not consider philosophy the domain of "weedy, sickly weirdos", and did not even imply that in my comment.
I'm arguing that you should regard the desire for attractiveness improvements as not only compatible with virtue, but *identical to* virtue! Exercise is virtuous! It's not *good* that you hate it, but it shows all the more fortitude and strength of will to stick with it if you do hate it, and you should remember this as you struggle with your motivation.
Also the others are correct in saying that most likely, you will eventually come to enjoy it. If you stick with it, eventually the others at the gym will no longer be a source of self-loathing but a mirror; and then you will also realize that *they were you*, long before you first entered that gym. You had no reason to loathe yourself, let alone, say, fear their contempt, as the new batch of beginners has no reason to fear yours. Your insight will have been elevated by the work.
It's good that you have a vision of a better field of philosophy in a possible future.
"Does anyone have recommendations on triching myself into liking it, or other physical activities with good cognitive / longevity benefits that might be less unpleasant?"
You are getting swarmed with responses. Here is mine:
There isn't anything special about lifting weights instead of doing other activities for physical fitness. In fact, I *think* (based on reading, etc.) that good cardio fitness is more important. So ... if you hate lifting weights and care about being healthy you might try focusing more on cardiovascular health.
This works out to mean:
a) In a gym: treadmill, elliptical machine, stationary bike, stair stepper, etc.
b) Outside: running, swimming, cycling, playing a running sport
I prefer a gym and watch videos while I am on the elliptical. I find that I can push myself fairly hard (so good for heart health) and also enjoy the videos. Doing something I sorta enjoy helps me want to keep doing it.
So ... try to find something you either LIKE doing that is exercise OR find a way to combine exercise with something you like doing (my solution).
NOTE: It also helps to *record* what you do. You want to skip less often when there will be a big blank spot on the record *and* the record lets you see yourself making progress ... which is nice even if you mostly don't care.
Just a note that all exercise is healthy, but there is a substantial body of evidence that resistance training and cardio are not perfect substitutes. You really need both.
I already have been tracking it in a spreadsheet. I felt a single atom of joy, once, at seeing a line go up, but that's it. Do you have any good citations for cardio vs strength, and also something for frequency? I took a single rapier lesson at the SCA, and it was pretty fun (and definitely a cardio workout!) but rather inconvenient to get to. (Also, I have free access to the university gym, which means no extra cost or travel time.)
My personal response: A spreadsheet is just something else to maintain unless you REALLY like watching a graph. I carry a small notepad, and write things down immediately after I take the measure. Then I forget about it until next time. When I write it down (or look) I can easily see changes in the last week. If I look at an old notebook I can easily compare where I was when I started (or finished) that book with where I am today. And it's less hassle. (Month at a glance calendars are good if you need a calendar, but I prefer a pocket sized notepad.)
"Do you have any good citations for cardio vs strength, and also something for frequency?"
Mostly Peter Attia for cardio. He lists four things for 'health' (which he defines as increasing longevity):
*) Cardio capacity (Max VO2)
*) Strength
*) Muscle mass (not the same as strength)
*) Balance
Cardio is #1. The others are not in order.
He may be wrong, of course :-)
3 - 4 times per week is fine. 2 is still better than 0.
As for not liking the feeling... I'm afraid I don't have much to offer from personal experience, as I can't really relate... Granted I'm on Adderall for ADD, which helps... But long before that, I did martial arts in my youth (and even before that I did tumbling--I was the only boy in the class--because I wanted to do flips like the Ninja Turtles), did dance classes and lifting in college... I just don't feel ok unless I'm regularly pushing myself to extremes. It seems to be genetic; my mom is like that too.
But even if you have less of on innate drive to feel the need to move (which is probably normal, I guess, as we didn't have the luxury to waste calories throughout most of human history)... From what I can gather, there is evidence that most people can and will "learn" to like it if they stick with it.
"Like highly addictive substances, regular exposure to exercise will over time teach the brain to like, want, and need it."
"One study of new members at a gym found that the minimum “exposure” required to establish a new exercise habit was four sessions per week for six weeks."
"How you feel the first time you try a new form of exercise is not necessarily how you’ll feel after you gain more experience. For many, exercise is an acquired pleasure. The joys of an activity reveal themselves slowly as the body and brain adapt. One man, who his entire life had believed that he hated to exercise, told me that at age fifty-three, he decided to work with a personal trainer to improve his health and support his recovery in a twelve-step program. He started with one workout a week, and within three weeks, decided that he could tolerate a second weekly session. One day he left a training session and noticed that he was smiling, something he describes as shocking. “I realized that not only was I happy, I had found actual pleasure in my training session. I hadn’t believed that kind of pleasure was possible outside of addiction.”
"For some people, it’s a matter of finding the right activity at the right time—such as the young single mother who felt isolated and “like nothing more than a mum” until she joined a recreational netball league, found a network of friends, and discovered a new identity as an athlete. For others, it’s about finding the movement their bodies were made for. One woman who started rowing in her forties told me, “So many of the women I row with thought they weren’t athletes, but as soon as they got in the boat, their bodies said yes, and they found their home.” Humans are also more psychologically complex than laboratory animals running in wheels. We get rewarded not just by how exercise feels, but also by what the activity means. One woman started going to the gym after leaving an abusive marriage. After thirty-eight years of having her movements constrained by her husband, she found being out in public and walking on a treadmill incredibly liberating. As she puts it, “I know I’m in freedom when I move.”
"Many people believe they don’t enjoy exercise in any form, but I’d bet that most of them aren’t immune to its rewards. It’s possible that they just haven’t exposed themselves to the dose, type, or community that would transform them into an “exercise person.” When the right dose, type, place, and time come together, even lifelong abstainers can get hooked. Nora Haefele of Stowe, Pennsylvania, didn’t start racing until her midfifties. Now sixty-two years old, she has completed more than two hundred events, including eighty-five half marathons."
McGonigal, Kelly. The Joy of Movement: How exercise helps us find happiness, hope, connection, and courage (pp. 43-44). (Function). Kindle Edition.
"From what I can gather, there is evidence that most people can and will 'learn' to like it if they stick with it."
I'll provide a caution on this. I don't know what percentage is "most people" but I'm not one of them. Neither is my wife. I have a friend who ran cross country in high school ... I don't think he is one, either.
And I don't get a runner's high, either (or at least don't get one after 'only' an hour of pushing myself at 85% of max heart rate).
I wonder how much of this is people who DO experience these things report them and those who don't .... don't.
I sure don't. Exercise was unpleasant even when I was a teen. That's not a absolute bar, though. I used to like hiking in the hills. But only with a friend, it was never a thing in itself. The last time I can remember enjoying exercise qua exercise was when I was 8 years old.
Interesting. I would suggest maybe you just haven't found the right thing and/or stuck with it long enough, but... I don't really know.
It is a problem that people tend to assume that others are basically just a version of themselves who different things happened to. But people are actually different, including how our brains are wired, due to genes far more than most people want to accept, but also for all sorts of other reasons.
I am quite certain that everyone would be happier and healthier if they got enough exercise. I think the evidence on that is pretty overwhelming. Being sedentary is really bad for you. So if you don't like it and never will, you should probably do it just anyway. But that's up to you. I don't like flossing, have always had a hard time sticking with it. So I got a combined toothbrush/waterpick, and I try to force myself to use the actual floss things too.
If you continue to find it unpleasant no matter what, maybe you could even turn that into an advantage. Being able to force yourself to do things you don't feel like doing is the most important skill for being successful at life. I do ice baths, and that's one of the main reasons. I feel really good afterward. But I never WANT to get in the water. But I just decide I'm going to do it, and don't think about. I've learned that I can condition myself to, even when my brain is screaming "NO!", be like, ok, whatever, and just do it. That's also what it was like learning to meet women, which I did a long time ago, and now I'm married.
Thanks.
I've been exercising for about a decade now -- weights and cardio -- and still haven't found that I have learned to like it. It is, of course, possible that I haven't found the right thing (maybe some niche yoga? :-)) or that it will take me more than a decade before I do learn ... but for practical purposes I think if it takes more than a few years and a non-mainstream exercise routine then it is probably more accurate to say that the learning to like it doesn't apply :-). So that's how I describe things.
Same thing with cardio and "runner's high". I asked my cross-country-in-high-school friend about this ... did I need to push myself for more than an hour [at ~80% - 85% of my estimated max heart rate] for it to kick in? He thought not. If it hadn't kicked in by 60 minutes then it wasn't going to. This was ... a bummer.
But I keep working out even though I don't get any sort of 'rush' from doing so ... because I like having more endurance for things such as walking. But the exercise is the price I pay to get where I want to be physically rather than something I do because I enjoy it for itself. Much like both of us with flossing :-)
My concern for TotallyHuman is setting expectations properly such that he/she/it/xe not assume that if they only lift weights long enough then *eventually* they will grow to like it. They might. But they might not, too, and should be prepared for that as well.
https://stronglifts.com/stronglifts-5x5/
Works for me, more or less. I get bored easily, and I'm offended by the uselessness of lifting things up and putting them back down in the same place.
After a few months, the weights start getting heavy. Stop lifting, go do something else (or nothing) for a while, then start over with the empty bar.
Very methodical, very gradual. Read between sets. Cut the time between sets when you are using the light e weights at the begining.
That is almost exactly what I have been doing, although a slightly different program. Reading between sets has increased my leisure reading (benefit!) but doesn't make me like the exercise more, c.f. https://slatestarcodex.com/2013/05/19/can-you-condition-yourself/
I think it's really admirable that you are principled and introspective enough to identify that one of your desires might be unvirtuous, and to resist acting upon it *even though* it would be in your self-interest to do so. Too many people (some ACX readers included, I fear) seem to define "virtuous" as "whatever my built-in, unexamined desires happen to be". I think you must be quite a thoughtful and principled person.
......HOWEVER! I do think you ought to reconsider whether resisting your wish to be more attractive, *even if* it might be unvirtuous, is a battle worth fighting. Yes, perhaps it might be shallow, (mildly) narcissistic, and maybe "giving-in" to social pressures - and in this sense could be said to be unvirtuous - but it does have the potential to *vastly* improve your quality-of-life at little-to-no cost to others. (If anything - you-as-thoughtful-principled-person being more attractive, and thus having more social cachet, probably has a net social positive, rather than all the available social capital going to the unashamedly narcissistic one-dimensional gym bros...)
Even if it is unvirtuous (by your own moral code, of course - not interested in mine or any anybody else's right now!) I do think that it doesn't really hurt anybody and there are probably other areas of life where you could do far more good (for yourself and for others) if you had the motivation and confidence that comes with feeling attractive - maybe, even if you're right about the virtue, this simply isn't a battle worth fighting?
Is building a (small) gym at home - possibly centred around just a few square feet of jigsaw matting, a single bench, a rack of free-weights, and maybe a spin bike (along with a small industrial fan and online HIIT classes or whatever via a cheap Android tablet) - an option for you? In addition to not feeling the 'impostor syndrome' you seem to be getting in the public gym, you would also get to train without having to drive anywhere, would be in sole charge of the gym's playlist, would probably save money in the long-term, would be resilient to COVID-style gym closures, etc. etc.
Alternatively, you could maybe take up something that's not-quite a competitive sport but also not-quite a hobby? Indoor bouldering, Crossfit, fitness clubs/classes, martial arts, etc. These things might provide a more friendly/motivating social atmosphere to train in?
I don't think being more attractive will increase my quality of life at all. It definitely would if I was actually ugly, but I'm well-groomed, not fat, have a reasonably symmetric face, am a charismatic public speaker, and come off as intelligent and thoughtful one-on-one (I have no idea why, but it's consistently true). I don't think that having better muscle definition or whatever will actually get me any more social capital (although I'd be interested in being proven wrong here, I haven't looked into it in detail).
You're right, of course, that harnessing unvirtuous motivations to achieve useful ends is a good idea, but the main thing that deliberately not-caring about attractiveness is doing for me is that I'm focusing on strength over hypertrophy. Wanting better cognition and lifespan is motivation enough for me, I think.
A home gym might be a good idea long-term, but as long as I'm going to university I have free access to a gym right next to my class, so no travel time or money spent. (It also seems to be a personal demon that I compare myself to others when it wouldn't be beneficial, and I am fully capable of imagining idealized people to compare myself to if I'm alone.) I hadn't thought of the playlist, and getting some earbuds is a good idea.
Unless you are already in the extreme tail of lean, muscular, and attractive, it is likely that you are massively underestimating the personal social capital that you would gain from pursuing weight training, nutrition, and aesthetic improvements. Basic social strategies like simple flattery are incredibly effective when they come from an attractive person. Spend some time around an extremely attractive man (or especially woman) and you will see that the world literally opens itself up to serve them in nearly everything they pursue.
Of course, this has the potential to be abused. It is massively abused by people who have it, but the same can be said of money or anything else.
You don't need to go through any of this rationalization. You can just stop thinking that the desire to be attractive is unvirtuous, because it isn't.
Yes, this is a really good point...which I was trying to make, but I failed to emphasize... Absolutely some desires are bad, and being natural, products of evolution doesn't of course doesn't necessarily mean we should invariably regard it as good. And I think we should all aspire to be virtuous. But didn't Aristotle say virtue is optimal point between opposite vices? The desire to be more attractive, as long as it isn't taken to too much of an extreme and is appropriately balanced with other concerns etc. is not only pretty benign; it can be a great source of motivation to do things that you should do anyway, for other benefits that you should want...but maybe don't have as easy time feeling a strong visceral desire for, like to have good bone density and reduced changes of Alzheimer's etc when you're old. So you might as well leverage the desire to be attractive as source of motivation to do what you should do for other reasons (and also to be more attractive, which is in itself beneficial and IMO legitimate to want).
Soreness aka DOMS is typical when just starting out, but if you are persistently quite sore then you may be pushing too hard. Just follow a linear progression program like greyskull LP, the numbers will feel like they're flying up by virtue of showing up.
The fitter people than you all started where you are, and in the grand scheme it takes comparatively very little time to get in reasonable shape, let alone see improvement.
Avoid those cognitive distortions as pitfalls, they will ruin anything you try, be it lifting, running, sports, etc. That's number 1. If that aside you actually "hate" lifting, then don't. You don't need it to be healthy. If you want the benefits of anaerobic exercise you can just do bodyweight stuff at home every day, for like 20 min, 60, whatever.
Walking in itself provides benefits. For health optimization and longevity though, consider zone-2 cardio, i.e. get your heart-rate up to 65-70% of its max and sustain that for the duration of a workout. You won't feel like you're dying, and you'll be yielding all the benefits.
I looked up greyskull and found a whole lot of dead links. How does linear progression work? My current strategy has been very non-mechanistic: I increase weight by 2.5 kg whenever the last session didn't feel very hard. Numbers have indeed been increasing, though it sure doesn't feel like I'm "just showing up".
I know everyone starts somewhere. I know comparing myself to others is detrimental, and I know that their success doesn't diminish mine. This knowledge does not prevent the emotional reaction.
example: https://imgur.com/iLhAJBl . You really don't need to lift to failure to reach an intermediate level.
Progress is 2.5 lbs not kilos.
If the aforementioned distorted perspectives aren't informing the emotional reaction, then what is? Taking a cue from 3rd-wave CBT (like MCT), sometimes correcting distorted views is not sufficient. It's then efficacious to a) learn to let irrational thoughts pass without judgement, and b) divert your attention to to something else. You can do this with exercises (found in workbooks and therapy guidebooks) or setting a reminder to entrench the new habit.
Still consider alternatives to lifting you might find more enjoyable. Sports are generally fun, and there are plenty less taxing ones with lower barrier to entry such as badminton, dodgeball, pickleball and others. Good way to make friends as well.
If you're showing up, then you're succeeding and reaching arbitrary goals becomes just a question of (little) time. It's physics. Not unlike the case of those pursuing weight loss and finding they cannot lose it all overnight (nor should they restrict that aggressively), a single pound per week shed is winning, so long as you sustain it.
Perhaps if you learn to accept your desire to be attractive, it will be easier to motivate yourself to do things that make you more attractive.
"I also have some desire for aesthetic / attractiveness improvements, but I don't consider that desire acceptable, and try not to let it affect my actions."
That seems really weird and silly to me. If you didn't have the desire, I'd say, ok, to each their own. But you DO...and yet seem to have some neurotic hangups about it. It reminds me of like the idea of the Christian who flogs himself like the guy in the DaVinci Code for having sexual thoughts.
It's a completely normal desire produced by evolution because being attractive makes you more likely to succeed at survival and reproduction. And luckily what makes you more attractive generally also makes you healthier. Up to a point... I wouldn't recommend focus primarily on the benefits for your looks, as that that can lead to neuroticism and being overly self-critical, and can lead to unhealthy extremes...which is ironic, because it usually also fails at the goal of being more attractive. Anorexics look scary and gross...so too in a different way do roided-out bodybuilders (perhaps not coincidentally since many of them may be driven by a similar kind of body dysmorphia.)
And being attractive has all sorts of other practical benefits, because people (male and female, straight and gay) generally will generally like and respect you more and think more positive things about you in general. (Unless you're a woman, in which case other women will see you as more of a threat.)
It's understandable to resent this fact of life if you happen to be unattractive in a way that it's hard to do anything about. But if it bothers you that it's not "fair" or "rational"... Well...too bad humans are the wrong species for you. It's the way it is, and it would take genetic engineering to fix it. And I suspect it wouldn't be a good idea anyway. Facts of nature aren't "injustices". To think they are is a category error, and trying to address them as if they were is pointless and usually just creates greater harm than it is ostensibly attempting to prevent. I prescribe Thomas Sowell's, "A Quest for Cosmic Justice".
Anyway...sorry, that kind of thinking is a pet peeve of mine. And I think letting go of it is genuinely liberating. It's ok to be a human. I think it's pretty great actually. And even if it weren't it would be stupid and pointless and counterproductive to agonize over it.
But anyway... I'll see if I can help with the original question. If lifting weights is boring, I get that. You can try other things. I just got my own home barbell setup, but before that I was mostly doing calisthenics and mace/club. Heavy steel mace/club stuff is fun and functional. If you don't like being at the gym, you can work out at home. Especially if you're starting out and have trouble with motivation, I think it makes it a lot easier to buy a program that gives you a schedule and has follow-along videos. Like Tacfit Commando, which is great for all-around fitness, and can be done in less than 40 min a day, but there are no doubt countless such programs you can get now. Just pick one you like and stick with it for 3 months. Then maybe try a different one. Having follow along videos helps with motivation. Decide what time you will do it, then just hit play, and then follow along.
Among desires, the desire to be attractive is a minor virtue. It's sure not a major one, and if taken for one counts as a vice, but in appropriate moderation it's a virtue.
Even if you are only considering other people, most people would rather look at attractive people than at unattractive ones. (That might be a reasonable definition of "attractive".) So by being attractive you are improving the lives of those who can see you.
Very relatable! I'm wondering if we read the same "You should probably lift weights" post, https://www.reddit.com/r/slatestarcodex/comments/9h2jbi/you_should_probably_lift_weights/
The only part I don't 100% relate to is the part about seeing fitter people bothering you, possibly just because a lot of fat people went to my gym. There were people with godlike physiques too, but enough normal/plus sized people that I didn't feel like I stood out.
I didn't exactly find lifting painful, but it was intensely uncomfortable and really boring. It only got more uncomfortable as the weights I was lifting increased. Going to the gym made my day worse, as did the thought of "urgh, I have to go to the gym later"...
A long life where I go to the gym three times per week is less appealing than a shorter life where I don't have to go to the gym.
I kept at it, on and off, for a couple years before I decided I wasn't ever going to get to the point where I started liking it. Tried going with friends, tried different regimes, it still sucked. I quit and I sort of wish I'd done it sooner. On balance it was probably worth sticking with it for as long as I did was probably the right choice, because it sounds like a lot of people do eventually come to enjoy it. I didn't turn out to be one of those people.
I tried a few alternatives - some attempts at martial arts, didn't really take to that. I got a pair of dumbbells to use at home and a pull-up bar. I use those a bit (thanks Beeminder), and I find it much less taxing to do little-and-often: just one or two sets of an exercise at a time, several times through the day. The dumbbells are basically my fidget toy, or that's the idea anyway. I work from home, which helps fit that sort of thing in.
After deciding martial arts wasn't for me and saying "I'd rather be in a mosh pit", I had a plan to actually commit to that - start going to local gigs explicitly just to get sweaty in a pit on a regular basis. That remains a daydream.
The main thing I ended up replacing the gym with is bouldering, which I see has already been recommended. Honestly I should have just skipped straight to bouldering, because I've heard so many people tell similar stories which usually end with them getting into bouldering.
I have no idea to what extent I'll get the the cognitive / longevity benefits, given that I'm not doing intense full body lifts anymore, but now it's a fun hobby instead of a slog. It's certainly a strength workout, but very unevenly distributed (I climb until my forearms stop working - it's always my forearms which go first).
See also this comment on the subreddit, although tbh I've just rewritten most of what I said there: https://www.reddit.com/r/slatestarcodex/comments/1eag0v1/comment/leopzq2/
I did not read that, but I'm familiar with most of its arguments. And yeah, I'm going to the gym at a university, so everyone there is not only fitter and prettier than me, but probably smarter too. I know it doesn't matter, that they started somewhere, that comparison of self to others is not beneficial, etc, etc... tell that to my status-seeking lizard brain.
I've tried rock climbing things a few time and disliked it far more than my current regime, I think because it requires a lot more flexibility than I have. (I've been stretching between sets in a desultory sort of way.)
You're getting swamped with answers, but I didn't see anyone say this, so I'm adding one more.
Consider HIIT instead of lifting - high intensity interval training. If you have to force yourself to do some physical activity every day (or every few days), it is by far the "highest benefit per unit time" physical activity. When you're starting out, it only takes 12 minutes, and you might build to 20-30 minutes total when you're more fit, and that's usually it. You also don't even need a gym for it, although you can of course do it in the gym (or a pool, or anywhere you can run, etc).
Here's a big list of the main benefits:
1. Improves fat burning efficiency and burns twice as much fat as traditional cardio.
2. Drives significantly higher post-exercise EPOC.
3. Improves VO2max, and drives better blood oxygenation.
4. Drives greater stroke volume, and greater cardiac contractibility, ~10-15% more than regular cardio.
5. It drives vascular adaptation, making your heart chambers larger and more elastic, improves the size and elasticity of your arteries, and increases the number of capillaries.
6. It drives hypertrophy - the relevant muscles get bigger.
7. It allows you to recruit more muscle fibers, and to do so more efficiently, driving greater muscular force and contractibility.
8. It improves insulin uptake, and improves the muscles’ ability to transport glucose overall.
9. It increases mitochondrial production and turnover, leaving you with more and “stronger” mitochondria.
10. Relative to traditional moderate-intensity cardio training, it drove a 41% increase in pain tolerance, and a 110% increase in race-intensity time before dropout in one study.
The first pass HIIT I recommend to most people, which only requires shoes and a watch or phone, is: "12 min total, consisting of eight 30s all-out-sprints (4 min) and 1 min rest intervals between them (8 min)."
Overall there's a lot of different HIIT styles you can do (Tabata, REHIT, and more), but any one of them generally drives the benefits.
What do you think of the 7-min workout? It's HIIT-adjacent, and exercises more of your muscle groups than just sprinting.
My exercise levels are low-ish but I credit the 7-min workout for not being a complete couch 🥔.
I used to do the 7 min workout religiously every morning! Ha, man, those are some memories. Yeah, I think it's pretty good. Certainly beats doing nothing, and good for getting the heart rate up and working each muscle group. I think it's not quite as good as full HIIT sessions, but it does work a greater variety of muscles, and is great to do overall.
Many years ago I would do the 7 min workout every morning. Then I graduated to 5bx, a similar 11 min HIIT-style workout the Canadian military invented and popularized in the 50's. I worked up to routines 5 and 6 on that. Nowadays I do Wim Hof breathing and a bunch of breath hold pushups every morning, because I do triathlon training and I spend about 2 hours a week doing HIIT in the course of 80/20 periodization training, so don't really need the extra load.
Link to 5bx for anyone curious: https://csclub.uwaterloo.ca/~rfburger/5bx-plan.pdf
To your point, I agree, I think there's a lot of benefit from doing some regular physical activity, even if it's only 7 min or 10 min at a time - every bit counts, and having a "floor" of physical capability is a good thing too. Bodies work on "use it or lose it," after all. When your body knows that it will have demands on each muscle and on your cardiorespiratory system every day, it keeps itself tuned up and in better condition on all those fronts.
Also find bike rides a nice way to get some cardio. You're outside with fresh air and sun, even just 1 hour (ex. 15km with a 200m drop) once a week gives you a nice exertion.
Buy some weights and lift at home. Makes it much more convenient and you don't have to deal with the social aspect of walking through the gym.
Have you tried "leaving two in the tank"?
You don't actually have to go all out and reach failure to get all the benefit of exercising.
If you're not feeling 90% recovered within a day or two you're probably going too hard
Really? I thought training to failure (or close to it) was important. Then again, I do usually feel 90% recovered within two days -- just in time to do it again.
https://j3university.com/leaving-reps-in-the-tank-or-all-sets-to-failure/
And the study cited:
https://pmc.ncbi.nlm.nih.gov/articles/PMC7725035/
Training to failure occasionally is probably still a good idea so you know what "2 reps left in the tank" actually feels like, but since you hate going to the gym, maybe easing up a little bit (and seemingly gaining the same benefits) would make it more tolerable
I was very impressed by this article, about a slow-motion form of AI doom. https://arxiv.org/abs/2501.16946
Zvi summarizes and discusses it in his Feb. 5 post.
":The baseline scenario as AI becomes AGI becomes ASI (artificial superintelligence), if nothing more dramatic goes wrong first and even we successfully ‘solve alignment’ of AI to a given user and developer, is the ‘gradual’ disempowerment of humanity by AIs, as we voluntarily grant them more and more power in a vicious cycle, after which AIs control the future and an ever-increasing share of its real resources. It is unlikely that humans survive it for long.
This gradual disempowerment is far from the only way things could go horribly wrong. There are various other ways things could go horribly wrong earlier, faster and more dramatically, especially if we indeed fail at alignment of ASI on the first try.
Gradual disempowerment it still is a major part of the problem, including in worlds that would otherwise have survived those other threats. And I don’t know of any good proposed solutions to this. All known options seem horrible, perhaps unthinkably so. This is especially true is the kind of anarchist who one rejects on principle any collective method by which humans might steer the future. . ."
So, guys, what do you think about this? (If you don’t want to read the article, Zvi’s summary is good, cuz he’s Zvi. (There’s a lot more to it than the bit I quoted.)&&
Thanks for the link. Must read that paper. I trawl through many ArXiV headers each week, but skip CS as there are just too many papers and most don't interest me much.
Once nearly all meaningful jobs are taken over by AI, one can only assume that relative merit will be judged mostly by social scoring, whereby commendable achievements will garner individuals merit points, like scouts completing courses. So if AI systems are allowed to supervise this activity, it/they will be perfectly placed to dangle various personalised incentives before every individual, and who is to say all these will all be benevolent?
Makes a lot more sense than the sci fi scenarios that are more commonly discussed here, with AI actively conspiring to take over. It checks out in terms of cultural evolution too, capabilities not exercised are quickly lost. It's the world of Wall-E.
If I remember correctly, Asimov wrote a short story late in life of how his three laws of robotics would impact on the long-term fate of humanity. Short version of the short story: To respect the laws, robots had across untold millennia gently and gradually turned humans into a docile species nearly unable to move, hence unable to harm anyone, or be harmed by anyone. Sounds like your gradual disempowerment-scenario.
I've been struggling through a backlog of Zviposts since the holidays (weird feeling to be like "I just wanna get home and read about o3" when customers are losing their shit over no more panetonne or wreaths for Christmas), but skipped ahead to that one due to the intriguing title...was a Worthy Read, one of the better non-roundup AI standalones. Same arguments Zvi's been making constantly, just now It's Official(tm) credential-wise and more accessible to normies. I find this ideative convergence pleasing. Many Such Cases of Rationalists making interesting, compelling, perhaps even correct arguments about the true state of things, but there's no penetration to those who can actually effect change...I keep wishing there were better "code switchers" with a skin in each game, who could translate The Good News to those who have ears to hear. (And likewise help us avoid common nerdsnipe typical-minding traps.)
The only part of it I feel able
to challenge at all
is the idea of the AI’s
competing with us for resources. That seems to assume that they will become self-interested, and have goals that may not be compatible
with ours. I have never understood why people are so convinced that once they get supersmart
they will have wishes and goals of their own. I think self-interest is a property of biological beings, rather than of smart beings. So it seems
to me that the ASI could continue to be obedient to us, and work only towards goals we set. I still think it would be very bad for our heads to have AIs take over everything because we are now the equivalent of its Down’s syndrome children, but not as bad, of course, as having AIs
start using as as landfill in marshy areas.
I think this tends to get covered by the background reading material a) instrumental convergence, that having *any* goal X to optimize at all funges against other goal(s) Y, unless X = Y; and b) it's actually really hard to define goals in a clear "do what I mean, not what I say" way (Value Is Fragile and related posts). So given resource constraints like land, power, water, etc., there's clear tradeoffs that could be made between allocating those towards goal X progress, and...whatever humans are doing with them. Consider humanity an opportunity cost, I guess. Hopefully there's enough other guardrails set up that such trades are simply outside the possibility-space...but we keep getting evidence that models will try to efficiently pull-the-rope-sideways around various constraints unless explicitly forbidden, without much/any prompting, and I'm not too comfortable continuously playing whack-a-mole against alien minds that are a lot more patient (if not literally "more intelligent").
That's been my understanding, anyway. Not smart enough to understand much of the technical details, just the end results...
> I have never understood why people are so convinced that once they get supersmart they will have wishes and goals of their own.
They don't need to. Accomplishing any goal is aided by control of more resources and power. Even if not strictly necessary, it still minimizes potential threats and liabilities. The only thing that you can truly trust is yourself.
<Accomplishing any goal is aided by control of more resources and power. Even if not strictly necessary, it still minimizes potential threats and liabilities.
But accomplishing one of our goals in a way that harms us is basically the make-a-lotta-paperclips situation, a situation where we erred by not putting conditions and limits on the goal. For instance if we told our Ai to conquer Canada, and imposed no restrictions on how, it might go to work on that, then discover the canadians plus their AI are tough opponents, and decide to kill
all Canadians with poison gas. And since we hadn’t remembered to tell it to conquer Canada in a way that did not risk US lives, it wouldnt worry about gas blowing over the border, and a bunch of US citizens near the border would die.
So it seems like you already believe in instrumental convergence, but think that it's possible to stop it by giving it good goals. Basically saying do what I want but obviously don't paperclip us in the way.
I'd say this is actually harder than you think, and loads of AI labs are working really hard on it.
But even if it was easy, it wouldn't matter! Seriously, people are really dumb and are time and time again giving paperclipping goals to AIs.
See this [1] Zvi post for one of N examples. Even the crudest of "agency" frameworks tacked on top of a model already empirically show instrumental convergence.
The models are too dumb to do any harm today. But we're already seeing it happen.
- [1] https://thezvi.substack.com/p/danger-ai-scientist-danger
>Even the crudest of "agency" frameworks tacked on top of a model already empirically show instrumental convergence.
How is this evidence of "instrumental convergence"? They specifically told the AI to edit a file until it ran without errors, so it edited that file. At one point it edited the file in an unexpected way, and another time it just made a dumb mistake that led to wasting lots of storage space without any benefit, but at no point did it show any sign of developing some new sub-goals besides those explicitly given. All that's confirmed here is that demand for evidence of instrumental convergence outstrips supply.
And how easy do you think it is to just "put limits on the goal"? The entire damn field of AI Alignment was created just to figure that out, and after more than two decades, they got nothing. What hope do you have to chain it, when they are far more creative and intelligent than you are?
And that's before we consider stuff like value drift and just simple malfunction. Of course, all of this is going to be irrelevant as soon as some moron purposefully makes one of these things sentient, in the name of "progress".
>And how easy do you think it is to just "put limits on the goal"? The entire damn field of AI Alignment was created just to figure that out, and after more than two decades, they got nothing
Says who? It seems to me that they've gotten plenty. In fact, the field of teaching AIs ethics has proven to be so easy in reality that those who want to continue dooming have been forced to come up with tortured rationalizations for why the unexpected ease of placing moral limits on AIs is actually just further proof of how dangerous they are (see Scott's posts on Claude's "incorrigibility").
It's true that whatever progress has been made to date has come entirely from people actually building and testing AIs in reality, not from high-IQ neurotics inventing elaborate apocalypse stories based entirely on some hand-waves Nick Bostrom made up (which is what "the field of AI alignment" usually refers to), but that doesn't seem particularly relevant.
If they were really that creative and intelligent, they would have realized the entire field was ludicrous to begin with. They can't even forcibly "align" a single human being much stupider than them, how could they possibly think under those circumstances that they'd be able to do it with an alien supergenius? Anyone not an idiot in at least some way bails right there.
By pressing the magic sentient button?
LLMs are smarter than humans on many tasks and are nowhere near being sentient. The idea that sentience is an emergent property of LLMs is clearly untrue.
More from Argentina: the free hormones are out. The doctor said that in the new system he just physically can't get them, there's a shortage. I am entitled to them legally but there just isn't enough. He gave me the prescription to buy it myself instead, and it's expensive enough that I'm considering smuggling them from other countries.
I've found a strange disconnect in rat circles - on one hand, you should stop reading the news, keep your identity small and be less engaged with politics. On the other, you should constantly question your beliefs and biases, seek out information on them, and bet on your beliefs on Polymarket (that is the wildest one to me).
Like, which is it?
I see it not as two conflicting beliefs, but rather as an *alternative*. Either don't read the news, or do read them but also bet on Polymarket.
It's the reading *without* betting that is bad. On one hand, you believe that it makes you know more about the world, on the other hand, you refuse to empirically test this hypothesis... because deep inside you know that many things you believe are probably wrong. So you are wasting your time obtaining information that you suspect it mostly wrong. Why?
Or someone is a normal person who doesn’t have money to bet on things.
Questioning literally all of your beliefs is impractically tedious; you need to prioritize. Mostly, you should be questioning the beliefs that are important to you in your actual life.
Which is to say, mostly *not* beliefs about things you only read about in the news. If you nonetheless do find yourself caring deeply about such things, yeah, question those beliefs. Including the one that says "this is important to me and I should hold strong beliefs about it".
Those don't seem opposed to me. I do try to question by beliefs and biases and seek out information about them, though I've only ever done one bet on Polymarket. At the same time, I try to tune out most sources of news and I don't engage deeply with politics.
For the news in particular, I think it's compatible to tune most of it out because it's generally a very low-quality source of information that's designed to keep your attention rather than inform you. There's a hierarchy here, of course: cable news and social media news is the absolute Pits, most online papers are in the same place, while a handful of legacy publications have the resources to do novel and in-depth investigations which are worthwhile to read (though this is not most of their content).
There are certainly a lot of things going on in the world which deserve thought and attention, but I generally don't need to know about them *right now*. The sooner to an event the news comes, the less likely it is to be accurate or complete, and the more likely that it's just trying to get attention on a breaking topic which might not even be relevant in a week. With the exception of things which are personally or professionally important for me and my loved ones in particular, which would prompt me to some action and which I would learn about through other channels anyway, I don't see a reason to make that tradeoff.
I keep my identity small and try to hold my beliefs loosely unless I've gone to enough effort to really research them myself (cue every stereotype about the person who 'does their own research'), which requires finding and reading the best arguments from the various sides. I don't do this too often outside of my professional life or hobbies, because it's a lot of work. This has delivered me a good deal more peace and clarity than when I was politically engaged and dogmatic.
This are obviously related in my view. Reading the news means you think you gain information, betting on news proves you don't. Questioning beliefs and biases are much easier when you don't have ideological or personal identity tied up in them, and For the most part "engaging with politics" is more about acting in impassioned ignorance rather than learning bloodless facts like "the Jones act cripples American trade".
The way I think of it is most "news" is meant to get you angry or serve as entertainment. CNN and fox want people to watch for as long as possible, websites want clicks, etc. The best way to do that is to be inflammatory not to have a sober analysis of events with numbers and statistics. But if by news you mean tarrifs are now expected to rise and here are thoughtful commentaries from 3 different economists with arguments both for and against, that's different. In this age of information you have to be discerning and also aware of when you are being emotionally manipulated. I try to stay informed and that does mean having some exposure to news, but I try to tilt towards sources that have actual information and aren't just libs/trump/fill in the blank bad.
I think that if one considers the news a source of noise/entertainment rather than a source of relevant information, the contradiction evaporates. E.g. there is no reason that I should care about tariffs between the US and Canada, as they will not measurably affect my life in any way, and even if they did, there's little I could do about it, but because I wasted my time reading a newspaper, I know about them and have a strong but utterly useless opinion on them.
I tend to believe that the busier a street is, the lower the chance of finding any low-hanging fruit along its edges. Politics is a very busy street.
An examination of missed opportunities shows I've personally gained around $80 from following politics over decades, because I gambled on a Trump win in 2016 when he was at longer betting odds than Nate Silver. On the other hand, I've lost out hugely when it comes to economic opportunities, like buying the dip in 2020, or a weird niche little alternative to online barter marketplaces called bitcoin that I consistently refused to buy.
I reckon most people would be similar. It seems very hard to make money or cause good in the world using politics unless you do it at a level that's far more in the weeds than reading the news, in a more hands-on way that involves either running for a minor office, helping someone else do so, or becoming a lobbyist. This impression comes from personal experience of seeing how lobbyists and grass roots politicians work and isn't rigorous, but I'm not sure what rigour would mean here.
Bitcoin still isn't underpinned by an asset and is thus still a scam. While it's true that you could have made a lot of money, you would have done so by scamming other people out of it, which is maximally unvirtuous. My friends told me about Bitcoin back when you could buy something on the order of 20 coins for five bucks; I didn't buy in then and still haven't because it was and is an obvious fraud. Defrauding people is wrong, even (especially!) if you can multiply your stake by a million.
(And don't think the persistence of Bitcoin is proof of its legitimacy; Madoff kept a simple ponzi scheme running for 35 years. He was the chairman of Nasdaq!)
Inside me are two wolves. One wolf whispers "that's a lot of mosquito nets, bro" and the other wolf whispers "yeah, not wrong".
Ah, defrauding for Effective Altruism! So you're folloiwng the lead of SBF!
Buying an asset that's propped up by nothing but A) speculation and B) debatable utility as a medium of exchange is not actual, legal fraud.
I'm not sure about you, but I think I operate on a sort of "utilitarianism-but-with-caveats" morality when it comes to tradeoffs, where the caveats are things like "don't break the law or the unwritten rules of decency unless they're unjust" and "don't do anything you know is wrong even if the maths looks good". SBF obviously broke the law, arguably broke the rules of decency, and if I'd been in his position I'd have known it was wrong.
Buying BTC, leaving the other suckers holding the bag, and spending it all on charitable causes doesn't strike me as fraudulent or outright wrong. BTC being propped up by speculation is public information. I and the hypothetical bag holders have access to the same sources. It's trading consensual economic harm against non-consensual malaria (or worms, or lack of access to fresh water, or lack of access to small business loans, or whatever).
Madoff was lying to his customers about what he was doing with their money. We can see all the transactions on the blockchain.
By this point I think everyone knows bitcoins aren't creating positive value commensurate to their price. So everyone buying them is knowingly investing in something that is at least partially a Ponzi scheme. If everyone knows what's happening, it's not really a fraud and by taking their money you don't do anything to them that they aren't attempting to do to others. Seems mildly unethical at worst.
I hope you're right just for the avoidance of total innocents being scammed, but I'm far from certain that's so, unfortunately. But even if you were right, I don't think that's a moral get-out-of-jail-free card. Supposedly, many frauds and cons operate on the principle of making the mark do something illegal or at least sufficiently unethical that they don't want to admit doing it, because it keeps the mark from going to the cops. By the same token I think swindling marks by tricking them into believing that they're the smart guy who's going to swindle somebody else is less permissible and more of a strat, really.
With reading the news, we have a choice whether or not to want to understand the political and economic world we live in. If we don't want to understand these things, that's fine - it's just as acceptable for rationalists to have no opinion about the Ukraine conflict or US tariffs as it is to have no opinion on the Sapir-Whorf hypothesis, or the many worlds interpretation of quantum physics.
If we want to understand these things for whatever reason, we should: 1) choose issues to care about selectively, based on actual importance rather than whether it's trending on X; 2) spend time reading deeply to genuinely understand the relevant issues, rather than wasting time on constantly updating news sources that provide no additional value; 3) have very good epistemics that allow us to evaluate issues without our biases or identity getting in the way; 4) if we think that we have insights about these issues and choose to speak publicly or allow them to influence our behaviour, we should hold ourselves to high epistemic standards (e.g. betting on beliefs).
Yeah, I don’t identify as a rationalist and have yet to see good arguments from them to become one. Why does it matter what I believe regarding AI, or the many worlds theory, or whether COVID was a lab leak or not? I should assume I’m not smart enough to have some refreshing take on it, or to solve any of these issues. Why not just enjoy my life and not worry about it? What’s the reason to become a rationalist, unless for your own pure enjoyment?
My personal solution is: I read lots of news, but try to limit myself to the object-level. No articles of the form "politician X finds idea Y good/bad", and certainly nothing of the form "politician X finds politician Y good/bad".
Instead, I try to read (actually lots of) articles of the form "idea X will probably have the following consequences". This is often mixed with judgments of the type "and this is good/bad", but it's usually rather easy to ignore those.
With a bit of training, it's pretty easy to tell the two types of articles apart just from headlines and possibly abstracts.
It's less "stop reading the news" and more "carefully budget how many epicycles you're spending on Bounded Distrust, since otherwise VOI can quickly go negative". People dunked on Scott a ton for his The Media Very Rarely Lies post, but I think the general point that you can usually extract basic facts without being in perpetual Constant Vigilance Mode With Straussian Cognitohazard Filter is accurate. Past that though...it's just a lot less work to hedge on epistemic uncertainty than to get anchored on a wrong thing, unlearn it later, and overrwrite with a correction. So I try not to read the news as "news" if I'm not willing to put in the work, or at best treat it as unreliably narrated fiction based on a true story.
(#NotAllNews, of course, some sources are more hostile than others...)
(friendly notice that I think you will have to use less jargon to be understood well, even in this comment section)
I guess that makes me feel old, if modern readers don't parse a lot of the older SSC/ACX/LW terms...but it's good to occasionally refresh one's citations.
*epicycles ~~~= brainpower spent per unit of time to complete a task
*Bounded Distrust: https://astralcodexten.substack.com/p/bounded-distrust
*VOI = Value Of Information, analogous to Return On Investment, but for facts. Everyone except maybe Tyler Cowen has finite memory and processing power to handle new information, don't get overloaded on low-value crap[1] or ephemeral trivia like play-by-play politics unless it's otherwise useful/entertaining. A great deal of "urgent news" is not, in fact, time-sensitive for the median audience.
*The Media Very Rarely Lies: https://astralcodexten.substack.com/p/the-media-very-rarely-lies
*Constant Vigilance: https://slatestarcodex.com/2014/06/09/constant-vigilance/
*Straussian readings: https://en.wikipedia.org/wiki/Leo_Strauss#On_reading (tl;dr the actual intended message is often not literally what's written on the page, it's as much what goes unsaid as what gets said; rounding this off to "but what are the vibes?" is not wholly inaccurate)
*cognitohazard: sorry, been reading too much SCP, information that's harmful once known; sometimes after doing a close reading, one finds the intended message is literally the opposite of the literal meaning (say, in a backhanded compliment), and such contortions can be distinctly painful if one is used to directly taking things at face value all the time, which is the typical Rationalist/autistic writing style
*hedging on epistemic uncertainty: I can't remember where the OG post is, it's not exactly Epistemic Learned Helplessness (https://slatestarcodex.com/2019/06/03/repost-epistemic-learned-helplessness/)...but tl;dr treat much of the news as malleable claims and interpretations layered on top of some base facts[2], rather than a definitive conclusion of Here's Why This Happened And/Or You Should Care. Be mindful that corrections and retractions to such Official Narratives are neither common nor widely advertised; lots of "common knowledge" is based on such early-but-wrong reporting that too few people bothered to revisit later once debunked/contextualized/whatever
[1] Beliefs Should Pay Rent: https://www.readthesequences.com/Making-Beliefs-Pay-Rent-In-Anticipated-Experiences
[2] Value news that mostly operates at Simulacra Level 1: https://thezvi.substack.com/p/simulacra-levels-summary
I don't think your definition of epicycle is standard. For example, https://www.lesswrong.com/posts/XAFQkbe6c9TRts6Ex/what-value-epicycles has
> “epicycle” has become the metonym for adding parts to a theory to make it work.
For "brainpower spent" I usually see "brain cycles" or just "cycles".
Makes sense, I never read the actual post and just picked up the term from cultural osmosis. But I guess it usually is mentioned in the context that adding increasing complexity to prop up a theory is generally Not Great, Bob. The More You Know(tm).
"the news" is one of the worse things that exists; instead you could consume media from a different culture, read long technical but political books, talk to people who disagree(tho that usually unnecessarily hard), or even if you want to be "up to date" you could read primary sources(like Ive skim 80% of evective orders and have unique take I havnt seen anywhere else)
Sarah Chen's computer displayed two windows side by side: in one, their latest VEGF expression data scrolled past in neat columns. In the other, a five-year-old paper declaring cold fusion's latest resurrection—and subsequent debunking. Her mouse hovered between them as the basement lab's fluorescent lights hummed overhead.
Their latest results were beautiful. Too beautiful. Just like cold fusion had been. Just like quantum consciousness and the EmDrive and every other revolutionary discovery that had dissolved under scrutiny's harsh light.
The centrifuge wound down with a soft whir, and Maya Rodriguez extracted another set of samples. Her latex gloves were marked with numbered squares—her own system for tracking which sample went where, redundant to their digital logs but independent of them. "Point-three-eight," she announced, holding the vial up to the light. "Third trial matches the first two." She logged the result in her notebook, then again in the computer, then photographed both with her phone. "Triple-blind documentation, just like we planned."
Sarah minimized the cold fusion paper and pulled up their protocol document. Every variable accounted for, every potential artifact identified and controlled. They'd designed their methodology specifically to catch their own mistakes, their own biases. The kind of rigor that had been missing from all those other revolutionary claims.
"Found something interesting in the control group data," Maya said, rolling her stool over. Her worn sneakers squeaked against the concrete floor. She pulled up another window, revealing a graph. "See this baseline variability? It's way too consistent. In real biological systems, you'd expect more noise. We're actually seeing too much precision."
Sarah leaned forward, tension uncurling in her chest. This was what they needed—not just positive results, but meaningful negative ones. Real data had texture. Patterns. Even its imperfections told a story.
The basement door's hinges creaked, and James Morrison ducked through, bringing a blast of Chicago winter with him. He brushed snow from his shoulders, his movements precise despite the cold. The messenger bag slung across his chest bore a faded university logo, relic of his former life in traditional academia.
"Just got out of the video call with Berkeley," he said, pulling out his tablet. "They want to know why we haven't published yet." His breath was still visible in the lab's air. "I told them about your triple-blind protocol. About the control group anomalies we're investigating." A smile creased his face. "They said it was 'refreshingly thorough.'"
Sarah's fingers found the worn edge of her lab notebook, where she'd taped a quote from her grad school mentor: Extraordinary claims demand extraordinary evidence. Below it, in her own writing: And ordinary claims demand just as much.
"The University of Chicago team sent over their latest validation criteria," James continued, pulling up another document. "They're offering full access to their BSL-3 facility. No strings attached." He paused. "Well, one string. They want us to present at their symposium next month."
Maya's hand tightened on her vial rack. "They'll expect preliminary results."
"They'll expect honest ones," Sarah corrected, turning back to their data. The VEGF numbers stared back at her, defying everything she thought she'd known about protein expression. But unlike cold fusion, unlike the EmDrive, these results came wrapped in layers of verification. Each number earned through methodological rigor that would have made her grad school mentor proud.
The ventilation system shuddered to life, stirring loose papers on their bench. Among them, a printed email from the grant committee: Forty million in new funding for breakthrough technologies. The word "breakthrough" had always bothered her—science wasn't about breaking through. It was about building up, slowly, carefully, each discovery earned one verified data point at a time.
Maya was already preparing the next batch of samples, her labeled gloves moving in practiced patterns. The centrifuge hummed back to life. On the whiteboard behind her, their experimental framework spread like a map, each step marked with both a procedure and its verification method. Red marks highlighted potential failure points—not to avoid them, but to face them head-on.
"You know what's different about us?" James asked quietly, studying their protocol document. "We're not trying to prove we're right. We're trying to prove what's true."
Define p(fascism) as the odds that some time in the next century, a significant majority of historians internationally will apply the label "fascist" or something roughly semantically equivalent to the current US administration. What number would you put on that?
My inside view is that the odds are quite high- maybe up to 30%. The man seems unwilling to share power with Congress, the courts or the bureaucracy this time- preferring to set policy with some occasionally pretty clearly unconstitutional EOs, while signalling that he'll ignore court injunctions. His nationalist rhetoric has also turned disturbingly expansionist recently, he's acted with a great deal of casual cruelty toward non-US citizens in these first few weeks, and I don't see him stepping down in 2029 without a very ugly fight. Also, his former chief of staff thinks the term applies.
That said, from an outside view, I notice that it's been pretty common historically for people to mistake relatively innocuous US presidents of the other party for potential dictators. I'm not sure I understand how that mistake occurs well enough to be confident that I'm falling into the same trap. So, I think I'd have to put the final number at something more like 10%.
That is still a worryingly high chance of a pretty nightmarish scenario, however. High enough that I feel like I ought to be putting together a serious emigration plan. Does that seem premature to people here?
"High enough that I feel like I ought to be putting together a serious emigration plan."
I'm rather biased because I put together and executed such a plan five years ago, and have not once regretted it. But I would say no, it doesn't seem even slightly premature. Having contingency plans for high-impact but somewhat-low-likelihood scenarios is generally wise, provided you're realistic about the probabilities and limit your investment in any one plan accordingly. It helps if a single plan can cover multiple contingencies, and "emigrate from the U.S." certainly covers a few others.
With all that said, unless you are a particularly fortunate sort of person, emigration is usually a slow and expensive: you'd need to commit a fair amount of money and effort well in advance of your actual moving date. So for such a plan to be any good, you'd need to have somewhere to go that you'd be reasonably likely to want to move to anyway. Which is to say, somewhere you'll like well enough that you'd be willing to start the emigration process when if things in the U.S. become merely *bad* but not yet *intolerable*. Along with picking a destination, it might help to try to define some sorts of events that would serve as signs to start packing: if you wait until you're Sure Beyond a Reasonable Doubt, you'll probably have waited too long. The best time to flee from a disaster is before it happens. The second best time is "before everyone else has the same idea and the exits get crowded."
On that note I'll add that (as outlined above) an emigration plan is only useful for slow disasters. Having something more in the nature of an escape plan--a plan to get away from danger if it happens suddenly and with little warning--is worth considering. I'm not talking about wacky secret-helicopter-hidden-in-your-garage Hollywood style hijinks. Mostly sensible things like knowing where your travel documents are and keeping them easily accessible, having an idea of where to go and how to get there in case of an emergency, keeping a big of cash on hand and maybe some basic emergency supplies, that kind of thing. Pretty big overlap with basic disaster preparedness. As I said above, keep the costs proportional to the (still somewhat low) probability, but also update as necessary. I'm certainly not expecting the skies to crack open and weep blood, but the sociopolitical climate in the U.S. doesn't look likely to get any *less* turbulent in the next few years, and there's plenty of run-of-the-mill dangers that come along with that.
In fact, it now occurs to me that (depending on where you live) the biggest short-term risk increase could easily be natural disasters. Particularly "natural disasters under a government that's unlikely to have much interest in or capability of handling them well." Fascism or no fascism, there will always be more hurricanes and wildfires, and I expect to see a lot more of citizens left to fend for themselves.
Why should I care about how future generations of historians will redefine "fascism"? I admit to some concern about what the current US administration actually *does*; I care not a whit about what term is invented or repurposed to describe it.
If that's not what you meant to ask, perhaps rephrase the question.
+1
More fundamentally, definition debates about fascism or communism seem a lot less fruitful than discussing what is or might happen and what the consequences will be.
But people aren't interested in having that conversation either. Maybe we just shouldn't talk about things at all.
That just makes me dismissive of future historians.
Almost zero. It's the politics of the time. At the time Romney and McCain were both called racist bigots and worse. Does anyone today think that is a fair and objective characterization today? Romney was accused of wanting to put black people "back in chains". It's ridiculous. Give anything a couple decades and then you'll be able to actually think rationally and fairly about it in context without all the emotion being involved.
The passive voice (“were called”) is doing a lot of work here. I’m sure that somebody, somewhere has expressed pretty much every opinion imaginable.
I’m guessing that the “back in chains” quote comes from a Biden campaign rally. I think that Biden was saying that Romney and the Republicans don’t care about the middle class. I’ll provide a link so anyone who cares can decide for themselves.
https://www.c-span.org/program/campaign-2012/vice-president-joe-biden-in-danville-virginia/284491
-------- begin excerpt (timestamp 31:28) --------
These guys say they care about the middle class. You know my dad had another saying [...], “Charley, don’t tell me what you value. Show me your budget and I will tell you what you value.” [...]
But let’s take a look, because we’ve got a real clear picture, as Tom pointed out. We’ve got a real clear picture of what they all value. They’ve said it, every Republican voted for it. Look at what they value, and look at their budget, and what they’re proposing.
Romney wants to let the--he said, in the first 100 days he’s going to let the big banks once again write their own rules. “Unchain Wall Street.” They’re going to put you all back in chains.
He said he’s going to do nothing about stopping the practice of outsourcing. Republicans even voted down a proposal to say, “OK, We’re not going to give a tax break, as we do now, to companies who unbolt the machinery on the factory floor and send it to Singapore” [...]. “We’re not going to do that any more; we’re going to give a tax break to any company that unbolts their factory stuff in Singapore and brings it back to Danville.”
It’s not going to change the world, but guess what? They voted against it. Romney opposes it.
-------- end excerpt (timestamp 33:23) --------
This is an argument that Proves Too Much. Noting that all politicians have been called bad names doesn't actually prevent some politicians from doing bad things. People always talk about "crying wolf," apparently without remembering how that story ends.
The trick is most historians are probably more left of center and more likely to be critical of trump and we also need to come to a common definition of the word fascism. For me that word describes a one party state with a dictator at its head, militarist, interested in expanding through use of force, with no free speech or basic political or judicial process rights (govt can do whatever it wants with you) and some basic race or ethnicity based superiority policies. I'm very concerned with Trumps presidency so far but I would put probability of what I described above a lot lower, like 1 or 2%. I think what is far more likely (30%) is a watered down democracy where the executive branch has even more power than it does now and can reliably and predictably ignore courts and Congress when it wants to and make policy on its own. I see this as a very bad outcome but still not quite worthy of the F word.
I have been thinking about the "relatively innocuous" label and believe that is over-applied based on:
(1) survivorship bias: the US political-economic system has operated continuously for nearly 250 years, so Americans think nothing truly bad happened that was a real threat to the system.
(2) failure to recognize significant change: arguably, the system that exists today is not the same one from the founding (or some other reference time, pick your favorite.) This maybe a ship of Theseus question, but I do think there's a massive naivety about how different the past was from today.
Another issue that I'm including here, though maybe should be a separate item, is the confusion that big mistakes need to have immediate consequences. For example, was Bush 2 innocuous? The country has clearly survived since his administration, so he seems to have been fine. On the other hand, we had a major step down in civil liberties (sacrificed in various ways for the war on terror) and plunged itself on an extremely dangerous (to the system) fiscal path.
(3) nostalgia: looking back and only seeing/remembering the good parts
(4) selective focus: looking back and only paying attention to the groups for which those times were good
As an amusing aside: at one point, a common piece of coinage issued by the U.S. Mint had the symbol of the Roman 'fasces' on it.
The 'fasces', which inspired the name of Fascism in Italian politics in the 20th Century, was a bundle of sticks tied together with an axe-head emerging near the top. The image was meant to show that a large number of weak sticks, when bound together, could become a single strong unit.
Anyway, the U.S. Treasury issued dimes carrying that symbol from 1916 to 1945.
It could be argued that issuing coinage with the symbol of the fasces on it was the closest that the United States ever came to some form of fascism.
However, there are other things that were called signs of fascism:
1. At one point, a President threatened to use his political skills to convince Congress to change the size of the Supreme Court in a way that would let him 'pack' it with friendly justices. It is possible that this proposal, discussed in the news at the time, might have been the reason that parts of the New Deal program declared by that President got Supreme Court approval.
Does a President threatening to pack the Supreme Court, as well as creating Federal rules and boards that impact all sorts of economic behavior, count as a step towards fascism?
2. At another point, a President used the language of declaring War to describe an effort to reduce poverty in the United States. He spear-headed an incredibly large increase in scope and spending power of the U.S. government. Is misuse of the emotional language around declaring 'War' on something a sign of fascism? The effort of ending poverty never really reached an end point that is called success. But this effort did increase the need for government income, via either taxes or borrowing. This effort also brought government officials into direct contact with the ordinary lives of many individuals.
Do ballooning budgets, huge social programs, and increasing the reach of government officials into private lives count as a step towards fascism?
3. Yet another President convinced Congress to spend significant amounts of money in reducing crime in the country. It didn't come with 'War on Crime' verbage, but the policy proposals did come with funding and mandates to massively increase the number of law-enforcement officials in the U.S. This effort was paired with laws signed by the same President restricting certain kinds of firearms from purchase in the U.S., and with laws forcing some sort of government background-check on anyone trying to purchase firearms from a licensed dealer.
A huge increase in total Police presence in the country: is that a step towards fascism? Do fascists like their subject populations armed, or disarmed?
4. Finally, another President got his signature program through Congress, only to have Congress taken over by his political opposition during the next election cycle. He proudly proclaimed that he could still get things done, because he had a pen and a phone.
The President declares that he can issue Executive Orders to get things done, when Congress is going to slow. Is that a step towards fascism?
Can you name any of the Presidents I've described, and their signature accomplishments? Can you give a fair argument against the claim that these actions were steps towards fascism?
I don't think that these Presidents brought fascism into place. They may have been doing things that I disapprove of.
I do bring these things up as comparison: it's easy to call a politician fascist, especially if you don't like their policies.
The fasces is still part of the seal of the Spanish Civil Guard (crossed with a sword). But Spain actually was a fascist dictatorship until 1974, and it didn't quite have the kind of reckoning with it that other countries did, so I guess they just didn't bother changing the iconography. Spain is by no means absent left-activism, so I always did wonder how that iconography never got controversial enough to be changed.
They're not the only police force that uses the fasces in its iconography. The fasces was originally (as in, in Rome) carried by lictors as a symbol of the power of imperium held by the official they guarded, and from there through a pretty intuitive pathway simply became associated with policing. I don't know the timeline for sure, but at a guess, this happened a good 60-80 years before the rise of Fascism.
I’m not interested in the higher-level point, but I want to push back on your framing for at least 1), 2) and 3).
1) Packing the courts is not sterling behavior, but hardly (uniquely) fascist. Given for how much time the system has existed, and how judicial review works in the US, it’s hard to think that the system as it is (or was in FDR’s time) is a bug.
A quick check suggests that the New Deal laws were voted by Congress and that they were likely pretty popular (since New Deal-supporting majorities in Congress grew in 34 and 36).
Then, when FDR suggested packing the courts (ie submitted to Congress the “Judiciary Reorganization Bill”, which never reached a vote), a bipartisan anti-New Deal majority took over Congress. I wonder why they did not roll back the existing legislation?
2) > Do ballooning budgets, huge social programs, and increasing the reach of government officials into private lives count as a step towards fascism?
Ballooning budgets and huge social programs do not count as a step towards fascism by any reasonable definition.
“Increasing the reach of government officials into private lives” is debatable but only because you frame it this way.
There’s a meaningful difference between a collectivity that decides that people suffering a life-changing accident need some measure of helps, and delegates officials to check that you are, in fact, as disabled as you claim to be, and a political officer checking your library for forbidden books and putting you in jail if they find some of them.
3) Actually, what happens in a fascist regime is that you can have weapons all you want – but the fascist thugs won’t be prosecuted if they use theirs while you will be. Or you and your family be administratively harassed. Or you’ll be gunned down (in a dark alley or in broad daylight) and no one will object because they’re afraid of being next.
Given the rest of the civilized world disagrees with the US on the merits of widely available weapons, it may be worth writing this off as an outlier. Especially since the murder rate in the US as significantly higher than in, say, Western European countries (which, just to forestall the objection, is not due to any specific race). Maybe Americans do need more policing to keep the crime rate low.
(I don’t know which president you’re talking about though.)
> Do fascists like their subject populations armed, or disarmed?
Gun ownership rules from Weimar Germany were pretty strict; Nazi Party loosened them for members of the Party. The Nazis were the ones wielding violence before then. Is the answer to illegal violence more of the same?
Extremely good post. This is what I would have liked to write, but I would've done a considerably worse job of it.
Honestly, <1%. Plenty of right-wing authoritarians and populists and dictators throughout history aren't considered fascist. Even Imperial Japan and Franco's Spain get arguments about whether they should be labeled fascist. Unless Trump decides to adjust his ideology to the point of being identifiably, straightforwardly fascist, he's extremely unlikely to be considered fascist by history.
We should confine the terms fascist to the early 1900 political movement that proudly used the term to define their philosophy, and other modern groups that specifically reference them. Certainly historians already do that, if journalists don’t.
I don't follow US politics or what's going on with Trump very closely but I'm interested in WW2/20th century history and to me 30% seems far too high. "Fascism" has become a broad word, maybe in the future it could get broadened even further to include Trump. But if you're asking whether an intellectually honest historian would compare him to Hitler/Mussolini/Franco I think there's almost no chance. 20th century european fascism came about in the context of countries facing multiple existential-level social problems like world wars, the potential for genocides, mass poverty, ideological conflict way beyond what we have now. Those conditions were far more extreme than the current US.
Fascist was already broadly used term of abuse before the Nazis took power. In his essay "What is Fascism" [1] (written in 1944) Orwell notes the apparent meaninglessness of the term and lists various examples. The Soviet Union and her Communist followers (which include the original Antifa) had already coined "social fascism" during the Weimar Republic, arguing that the Social Democrats were the true fascists, worse than the Nazis.
[1]: https://www.orwell.ru/library/articles/As_I_Please/english/efasc
> arguing that the Social Democrats were the true fascists, worse than the Nazis.
Not quite, the Stalinist position pre-1935 was that "Social-Democracy is objectively the moderate wing of fascism"
The chance of a coherent coversation or analysis using the word "fascism" or any similar word is absolutely zero. By 2010 or so it had been used for too long to mean "something I don't like" that it was already close to zero. As soon as mainstream wokeness appeared and significant numbers began to be capable of calling misgendering a form of "oppression" *with a straight face* it dropped right to stone-cold zero and it will never recover. Just see that trainwreck ACOUP article for an example.
If you want to talk about the actually worrying aspects of Trump, in a way that is comprehensible to anyone who doesn't already share your political views on everything, then completely taboo "fascism" (along with any other word that's often used to include completely mainstream conservative positions, like enforcing immigration laws and opposing the transgender movement) and replace it with something extremely concrete and clearly-defined. Like "will stay in power beyond his term" or "will outright defy courts or congress or seek to shut them down". Then, and only then, will a coherent discussion be possible.
"Just see that trainwreck ACOUP article for an example."
This seems like an utterly ludicrous example, given that the article was VERY scrupulous about defining the term as thoroughly as possible AND pointing out the many places in which it did NOT apply (including Trump in his first term).
Like, I agree that people will fling around extreme, pejorative political terminology at the drop of a hat, and take no care for the meaning. But carefully grounding a term in the original meaning AND carefully pointing the limits around it seems much more like pushing back against the trend than buying into it (even if the analysis were flawed, which you haven't actually argued by any means other than invective). One suspects that some simply found the analysis uncomfortable and decided to group it in with things they'd already resolved to dismiss.
> Define p(fascism) as the odds that some time in the next century, a significant majority of historians internationally will apply the label "fascist" or something roughly semantically equivalent to the current US administration. What number would you put on that?
Id label america fascist since fdr; while america will never be like germany; this is still a subjective question.
Your trying to be objective about fascism without stating a definition, dont; id speculate *the inverse* of what your trying to ask; if the future gets more free then more people will call the current political system fascistic, if it gets less free, more people will can it democratic.
"whats the p(democarcy) a historian in best korea calls the next year democracy in 50 years"; it would be the inverse of if best korea collapses
While Scott has written that using made up numbers is great, I would disagree. This is especially so when you're trying to do computations with those numbers. This also applies when your numbers have large error bars on them (which in the limit becomes making up numbers).
You pulled the 30% number out of thin air, and just because it feels plausible doesn't make it so. Rationalists love to use poorly founded estimates to compute expected values; see https://blog.givewell.org/2011/08/18/why-we-cant-take-expected-value-estimates-literally-even-when-theyre-unbiased/ for a critique of this in a non-Trump context. If you assume you have a 1/3 chance of being right and there's a 30% chance of Trump being a dictator, and then decide that means you should act as though Trump has a 10% chance of being a dictator, you are doing an expected value computation based on a poorly founded estimate.
I don't think the original Italian strain of fascism placed much emphasis on racial purity.
Similarly, lots of other arguably fascist regimes seemed a lot less obsessed with race than the Nazis--Salazar, Franco, Pinochet,etc.
I often hear the claim that AI professionals are more likely to have a higher p(doom) than other people and that's a good reason to worry about p(doom). But the premise doesn't strike me as likely. If most AI professionals (I mean people directly working at the frontier AI companies) believe p(doom) is reasonably high, then these people are knowingly working on a doomsday machine. I find that implausible. Reason dictates that it's much more likely that the majority of those working at frontier AI companies believe p(doom) is very low.
Perhaps it seems otherwise because a vocal minority of dissenters who are more likely to work on the alignment side for these companies do have high p(dooms), and because of their vocality we falsely believe that they are representative of the industry.
Simple tragedy of the commons. 90% of the time you’re working on the greatest innovation since the invention of writing, not to mention all the moneys. Plus, the overwhelming feeling in these circles is that the doom is happening whether or not they contribute, so why not drive a Ferrari before the world ends?
I don't think it's at all implausible that people are working on something they think is high risk!
Surely you don't think people are this rational, ethical and cognitively consonant!
Atomic Scientists famously had (have?) a higher p(doom) than laypeople (https://en.wikipedia.org/wiki/Doomsday_Clock) despite Atomic Science being the poster-discipline, as it were, for building doomsday devices. I can think of two ways to look at this:
1. Maybe the field isn't a unified "they". It's made up of multiple types of people who all understand the doomsday device pretty well; some who are actively working on building it and others who are pretty worried about this.
2. Maybe the ones who are working on it are the same ones who are worried about it - but they can't stop working on it because Moloch compels them (https://slatestarcodex.com/2014/07/30/meditations-on-moloch) - something like, "even if we stop working on the doomsday device other people won't stop, and even though we're worried about doomsday devices *somebody* is gonna have build one soon, and we'd rather it were us than those bastards on the other side of the border - and in the mean time these film-star salaries we're all getting certainly do help to take the edge off our anxiety..."
The Bulletin of the Atomic Scientists is a non-profit campaign group rather than an association of actual Atomic Scientists. You can't really take what they say as the consensus of scientists in various nuclear fields.
Possibly it is today, yes [though even then I would be fairly surprised to discover that the Bulletin of Atomic Scientists doesn't get any input from any actual Atomic Scientists for things like setting its clock..] but what we're actually interested in is the Bulletin of Atomic Scientists in the 1940s, when Atomic Science was more comparable in maturity to AI (and when their clock was produced). Back then, I'm led to understand, it was founded and staffed by Manhattan Project scientists.
Secondly, even disregaring Atomic Science entirely, I still think my (1) and (2) above are still entirely plausible explanations even without a convenient prior case-study.
Whether or not I personally work on AI has basically zero impact on p(doom).
But it has a huge impact on whether or not I can buy a Lamborgini.
its marketing; all useful tools are dangerous by asserting danger its implying useful while making it sound like a warning that bypasses most filters.
There's a poll of AI researchers on this question from 2023 at: https://blog.aiimpacts.org/p/2023-ai-survey-of-2778-six-things
According to the summary, median respondents put 5% or more on advanced AI leading to human extinction or similar, and a third to a half of participants gave 10% or more. Subjectively, it seems like the worry has become more mainstream among AI people since the poll was taken, but I'm not sure if there's more recent data.
These don't seem like the same people I am talking about. I just pasted the description of the participants in that study into Claude, asking who they are and the reply is:
"Those conferences you mentioned are primarily academic venues, though researchers from major AI companies do publish there as well. Here's a breakdown:
The researchers are likely to be:
Primarily academic researchers from universities worldwide, especially from major research institutions with strong computer science/AI departments
A significant minority from industrial research labs at large tech companies like Google (especially Google Brain/DeepMind), Meta AI, Microsoft Research, IBM Research, etc.
A smaller portion from specialized AI companies like OpenAI, Anthropic, Cohere, etc.
It's worth noting that frontier AI companies like OpenAI and Anthropic have recently been publishing less of their research in these academic venues, often choosing to release their findings through their own channels or preprint servers like arXiv. This is partly due to the fast pace of development and competitive considerations."
So this survey was mostly of academics not the people actually working on the front lines trying to bring about AGI.
The difference between academics and people working on the front lines is much smaller than you seem to think. A lot of that is just the same people: the biggest group is of course PhD students and other academics who join the AI companies, but it also flows in the other direction. The staff of AI companies is academically very young, lots of them are fresh from university. Especially those who build the frontier models. And even beyond people, there a tons of research papers which are collaborations of employees of the AI companies with academics.
I also find it a bit weird that you mention arxiv as an indication that AI companies move away from academia. Arxiv is totally a product of academia. Even as a non-AI academia researcher, I rely on arxiv much more than on conferences/journals. All my papers are on arxiv, and all papers of my colleagues are on arxiv. In AI this tie is even stronger because the pace of the field is getting too fast for conferences.
There are many ways to rationalize the two, for example that p(doom) is high but it might be even higher if America doesn’t win the race.
But is that what you really think is happening? All these people working on AGI think the end of the world will come because of what they do but maybe they can slightly lower the odds if they get ahead of the competition? That doesn't sound like a plausible psychological disposition for a large group of people to have.
PEPFAR
Any Effective Altruists out there with an actionable plan to help alleviate the crisis caused by Musk?
I realise pepfar has officially been reinstated but see Substacker The Fucking News (sorry linking is beyond me) for the ongoing damage.
"linking is beyond me"
Just put the URL in the comment, it will be automatically converted into a clickable link.
>think the end of the world will come because of what they do but maybe they can slightly lower the odds if they get ahead of the competition?
It would be interesting to poll nuclear weapon and ICBM designers from the height of the cold war and find out if they looked at the arms race in something like this way...
We know that if any of the Manhattan Project scientists had a high p(doom) over nuclear weapons within the next 75 years of the project, they were wrong. So if this is an analogous situation, we shouldn't trust the scientists' forecasts. If it is not an analogous situation, the comparison tells us nothing.
What? Are you implying that if I think that something has 60% chance of occurring, and it doesn’t, that I was wrong?
That's not how probabilities work. Unless their p(doom) was >90% (and, given the anecdotes of close calls, one might reasonably argue >95 or >99), then saying "they were wrong" doesn't make sense.
edit: and even in those cases, saying "they were wrong" wouldn't _technically_ be correct. But it would be a useful shorthand for the much more difficult to express way of assessing their claims post-facto
I get what you are saying. It would be more accurate to say that those forecasts would have received a bad Brier score.
Many Thanks!
>We know that if any of the Manhattan Project scientists had a high p(doom) over nuclear weapons within the next 75 years of the project, they were wrong.
True! (EDIT: correcting to, as per discussion above,
>would have received a bad Brier score
)
I'm guessing that their views might shed some light on the psychology of living with significant p(doom) estimates from one's work's effects.
Is eye-rolling cultural or a universal facial expression?
What other facial expressions might be cultural or universal? Russians famously don't smile for strangers. What else?
Also, as a mono-linguistic English speaker, I've always wondered what exactly the meaning is of say "tu" vs "vous" in French or the equivalent in other languages. It seems so strange and foreign to me that a word like "you" could have such strong implications depending on the form. I have this theory that English used to be like that when "thou" was the second person singular but us cold English speakers retreated entirely into the formal "you" and dismissed the more intimate "thou" because the culture frowned on such intimacy. Might that be correct that "thou" was like "tu" but eventually no English speaker were comfortable with the more intimate form?
On one side of my family the “thou” form lingered into the late 20th century when either of my grandparents gave thanks at table.
French has the notion of "lever les yeux aux ciel", literally "raise one's eyes toward the sky", which is basically eye-rolling. But it is not well-identified as *the symbol* of irritation like eye-rolling is in (part of) the Anglosphere. I have sometimes read things like "this public statement prompted lots of eye-rolling" in English-language media, where "eye-rolling" is a metaphor for people being irritated and in fact expressing that by other means (speech or text). The same metaphor with "lever les yeux aux ciel" would not work as well in French.
We don't smile at strangers that much either. Smiling is for when you want to communicate something nice.
The anglosphere has people who declare their own 3rd person pronouns and this has turned into a very divisive social marker. Surely you can understand how pronouns can have strong social implications.
> Might that be correct that "thou" was like "tu" but eventually no English speaker were comfortable with the more intimate form?
I've always thought this to be the case, and a quick search on Wikipedia and Online Etymology would suggest this is correct (https://en.wikipedia.org/wiki/Thou; https://www.etymonline.com/word/thou#etymonline_v_13251).
Yes, that's exactly right. You can even see some similarities between Early Modern English (around Shakespeare's time) and modern German: nowadays we say "you have" (and Germans say "Sie haben"), but the informal form would have been "thou hast," whose (still-current) German equivalent is "du hast."
The broader phenomenon is this: https://en.wikipedia.org/wiki/T%E2%80%93V_distinction
The T-V distinction's name comes from what were singular (tu) and plural (vos) in Latin, but there are similarities in many non-Indo-European languages.
The ironic thing about English is that now the archaic quality of "thou" makes it sound more polite, when it originally wasn't. The T forms are perhaps better thought of as being more intimate rather than more informal; off the top of my head the various Indo-European languages I'm familiar with tend to use the T forms for God (we see this in the Our Father with its uses of "thy").
In more recent times some Indo-European languages, such as Norwegian and Swedish, are getting rid of the more formal forms.
I consider the whole matter to be somewhat analogous to deciding whether to use first or last names when addressing someone.
Also, if you look up "french tu vous flowchart" you get some pretty funny (and informative) results.
Consider the difference in English of saying something like "please follow this gentleman" vs. "please follow him" vs. "please follow this fella". It's sort of like that. Obviously, the divisions and rules in English are different.
> what exactly the meaning is of say "tu" vs "vous" in French or the equivalent in other languages
I am not a French speaker, but I speak Russian. The same phenomenon ("ты" vs "вы") exists in Russian, and based on a quick check it seems to be exactly the same semantically (one form conveys more familiarity and less respect, and vice versa).
In Russian, this was a Western linguistical phenomenon - the "royal we", pluralis majestatis - that was grafted onto Russian language. Until Peter I (emperor in 1721-1725) there was one (UPD: SINGULAR) second-person pronoun "ты", and if the speaker was addressing someone far more powerful than them, they used titles, honorifics and other words signifying deference.
Peter was a big fan of westernized monarchy (he was a big Hollandophile specifically, IIRC) and tried to reform Russian monarchy, state and society along similar lines. In 1722 he published a "Table of ranks", which was a stratified list of positions in civil society, military service and church. It was his idea to prescribe that anyone addressing someone of a higher tank on the Table use the pluralis majestatis pronoun "вы" to address them.
Then, in 1917, when Bolsheviks formally abolished the Table of Ranks, together with the rest of the Imperial state apparatus, the usage of "pluralis majestatis" just folded into regular language and was - and still is - used in formal and semi-formal settings for just about everyone to convey a minimum of respect and/or dignity (e.g. a news anchor interviewing a scientist or an expert on something will use it, the cashier in a fast food restaurant taking an order from a client will use it - and it would be very bad form for the client not to use it back).
It is very slowly deprecating out of usage (among young people, people with progressive leanings, etc), but is still widely used. For example, I was in a few meetings at a tech company staffed with Russian speakers - varying ages from early-20 to late-40-year-olds, different ranks, seniorities, etc - and nobody used "вы" with anyone, but I imagine a similar meeting at the Russian ministry of something would have been very different. It is subjective, but I heard opinions that this represents a "democratization" or even "americanization" of the language.
The English version (thou/thee vs. ye/you) was also a foreignism - borrowed from French during the middle ages. French inherited it from Latin. German has it too, and I feel like they probably picked it up from French or Latin at some point. (Then later it became more complicated in German.)
Also a Russian speaker. I'm not sure I would say it's falling out of use so much as it's less of a norm specifically in tech companies.
Generally, I would compare using the polite form to using vs not using "sir"/"ma'am" in English.
In Vietnamese, the number of different words for “you” is enormous! It depends not only on the level of formality but if you are younger or older than the person you are addressing, or (if significantly older) if they are younger or older than your parents. If they are within your family then family relationship influences the word you use as well eg there is a difference between “grandmother” if they are the mother of your mother vs if they are the mother of your father. Other Southeast Asian languages are similar. In general, English is remarkably low complexity with just “I” and “you.”
English has a whole bunch of formal second-person pronouns that most of us don't get to use all that often.
For instance, the second person pronoun for a king is "Your Majesty", for a pope it's "Your Holiness", for a Duke it's "Your Grace", and so forth.
I thought those were mostly used to address them (like "sir" in some other formal situations) rather than as pronouns?
No, you use the same phrases in third-person reference. His Majesty / His Grace etc.
For a case where a phrase like this actually became a pronoun, Spanish 'usted' is supposed to have come from 'vuestra merced' [your grace], which is why it has third-person verb agreement.
Oh, does Spanish do that? Italian Lei (second person singular formal, agrees with third-person singular verbs) has a similar origin - not grace, but another respectful virtue, so feminine. Leading then to the correct pronoun for excellence/dignity/serenity/whichever it was being she, capitalized to clarify you aren't actually talking about a third-person feminine noun.
Old Italian documents use voi, which shares the second-person formal singular and second-person plural roles as in French; in modern Italian, voi is just plural-informal you and the plural formal you is Loro. Do old Spanish documents also use a cognate of voi/vous?
The old form in Spanish is vos. You can take that quite a ways back, since it's also the Latin second-person plural pronoun. I was under the impression that vos is still used, as a plural, in parts of South America, but wiktionary suggests that Latin American vos is exclusively singular, so I may have misunderstood my informant. It would have been an easy mistake for me to make.
The current form in Spanish is vosotros, which has had another word combined into it. (The same thing happened to the first-person plural pronoun, nosotros.) The Latin pronoun, and modern vosotros, do not have singular senses, but Old Spanish vos seems to have had one. https://es.wiktionary.org/wiki/vos
As a rule, Latin abstract nouns are feminine. They don't have to be respectful - for example, cupiditas is feminine. So it doesn't surprise me that virtues would be feminine in Italian, but I am surprised that respectfulness might be a requirement.
Anyone know a good multivitamin like One A Day (has vitamins and minerals) without the Calcium? I like everything else in the One A Day.
Calcium in artificial form seems to increase risk of calcification in the arteries. That's the motivation to avoid it. Been looking for a while, with no luck.
Thanks!
I highly recommend ConsumerLab.com, the Consumer Reports of vitamins/supplements/etc. They independently test and then recommend products, and some of their discoveries are so alarming that I would NEVER purchase a supplement they haven't approved.
In addition to test results and best picks, they also offer Scott-esque "More than you wanted to know about..." articles on everything they test (major and minor studies, etc).
All that said, I consulted the multivitamin tests and their overall top pick for quality at the lowest price is Kirkland Signature (Costco) Daily Multi. They mention the vitamins cover 100% of most vitamins, while calcium is at a lower percentage because intake should come from diet. Calcium is at 200mg.
They also mention Naturelo One Daily Multivitamin for Women 50+ as a better vegan option, which incidentally only has calcium at 60mg.
There are some prenatal vitamins which don't contain any calcium, but a quick skim would seem to indicate there's a lot of tradeoff going that route, so they're not a great substitute.
There's a section on calcium supplements and heart attack and stroke, but it doesn't seem to be a concern unless someone is supplementing 1000 mg daily *and* absorbing calcium in their diet. So tentatively, I would think that a 60mg dose or even a 200mg dose in a multivitamin isn't anything to really worry about.
Would your doctor be bitchy about ordering calcium and vitamin D blood tests? It probably wouldn't hurt to get an idea of where you're at right now before you start reducing calcium.
Thank you for the detailed and thoughtful response. I'll be reading the references you ve provided, to understand your points more deeply.
Sorry to not be able to better answer your question, but the problem with multivitamins is that they don't give you enough good forms of the bigger metals (pill would be too large and expensive). For example, the magnesium oxide in it is the worst form (gives constipation and poor bioavailability for such a key mineral), and the B12 is the worse version too. I understand the motivation to only need 1 pill, but you should be taking fish oil anyway and a separate magnesium glycinate anyway, so is there that much hassle to taking targeted others like D + K2 + C, B complex, and maybe zinc?
Excellent point.
I'm not sure what to take away from this, but I think it's interesting that both Musk and Trump, with basically infinite access to information, jump at Internet conspiracy theories:
https://www.npr.org/2025/02/07/nx-s1-5290282/politico-subscriptions-usaid-x-musk-trump
I'm reminded of the somewhat-under-reported fact that Trump apparently thought Hillary Clinton's email server was hidden in Ukraine:
https://www.nbcnews.com/politics/trump-impeachment-inquiry/trump-promotes-conspiracy-theory-clinton-s-deleted-emails-are-ukraine-n1058726
I really wonder if this has anything to do with the vague claim that some US debt might be invalid (sorry for the mediocre source, don't have a better one handy):
https://talkingpointsmemo.com/edblog/trump-says-some-treasury-notes-may-not-be-real
and X thought 2016 was a rigged election by Russia to install their agent as President of the United States. Fill in X with your favorite politics/media figure from the other side. Having access to all the information you could want doesn't mean you cant believe an incorrect conspiracy theory.
Well, yeah, but if I were the President I might ask the NSA or the CIA or something about it before I brought it up to a foreign leader. And if I were Musk I would be looking at the massive data set I have from the Treasury department and not random X posts. Why not at least confirm these things based on your insider access before acting on them?
The first article:
"omg, Trump is so stupid he believes the conspiracy theory that USAID was funnelling money to Politico. Now of course USAID was funneling money to Politico, but in a totally above board way that couldn't possibly result in corruption"
"Entire government has some news subscriptions" is obviously different than "government agency specifically pays news sites to write what they want".
And again we're talking Politico, who put out articles like this: https://www.politico.com/news/magazine/2025/02/04/democrats-foreign-aid-trap-trump-00202447
Their whole thing is horse-race, "how will this play with the voters?" coverage. That article basically says Democrats are morons for defending foreign aid, and only briefly mentions that there might be some principle at stake. This is the organization that supposed to be influenced by subscriptions from USAID (which was only one of many agencies subscribing)?
It depends on the number of subscriptions. Press secretaries buying copies of all the papers is fair enough, but if it's 100,000 subscriptions that's more like what Soviet embassies used to do with the Daily Worker. Give this is apparently more than half of Politico's revenue, it's on the scale to be the latter.
It’s not clear that this is over half of Politico’s revenue? https://skeptics.stackexchange.com/questions/57666/did-politico-receive-usaid-funding
suggests that we’re talking around 10M$ per year from the federal government, and a very quick Google search (including the first three lines of a recent WSJ article accessible without subscription) suggests that Politico revenue is about 20 times that?
I misread the article - Politico Pro is more than half their revenue, the federal government is about 8% of that.
Isn't Politico's newspaper a big advert for their highly expensive live data service which is something governments would buy for legitimate reasons?
Could be. Or it could be a general sort of slush fund that moves money from Dem-friendly government departments to Dem-friendly media outlets.
It could be, but it isn't. The US government's Politico Pro submissions are spread across way too many agencies to be effectively managed as a slush fund, and the amount coming through USAID specifically is too small to plausibly sway Politico's leadership.
Pray consider the possibility that you may have been bamboozled.
It could be, or it could not be. Perhaps the most powerful man in America could *find out which it is* before announcing that it's a money laundering scheme to fund left wing propaganda.
> I'm not sure what to take away from this, but I think it's interesting that both Musk and Trump, with basically infinite access to information, jump at Internet conspiracy theories:
its known that trump was lied to about troops in Syria, it could be simply rational
I'm not sure what claim you're referring to, but, how is believing Internet conspiracy theories a rational response to being lied to about an unrelated thing?
https://taskandpurpose.com/news/us-troop-levels-syria-jeffrey-interview/
Interesting, but again I'm really not seeing how "a Trump administration official lied to the President" = "it's rational to believe Internet conspiracy theories".
How many times do you need to be lied to before you discard the information?
Suppose your dealing with a young earth creationist, your perfectly capable of providing sources of contradictions in the bible, but they insist the bible is true and god works in mysterious ways; several of your debates are pulling teeth where they cite the bible to discard something you believe; over and over again. The devil made dinosaur bones during the story of job cause it says the devil made every effort to mislead, the bible defines atheists as hating god. etc.
It would be irrational if this never ends; however, if theres a number that breaks down the wall, maybe 1000 errors in the bible; its *rational* in a very technical sense. Capability of reason.
I assume your aware "conspiracy theorists" use the term "scientism"; science in that sense is not the ideal theory but the system, university's, journals and to some degree newspapers and their intelligence friends. Do you have a number?
These numbers should probably be fairly similar, the number of times you think its tolerable to humor a young earth creationist(who risks being thrown out of community's for considering your questions) and the number of time you catch an major error in "the official trusted sources" of science, intelligence etc. before ... well quite lost.
---
I think lying about troop numbers in a war zone to the president is a very very big deal; the whole separation of powers thing says the president gets allot of militery power and a matter of life and death, how many americains died in iraq?
It should shake your faith in the system to your core; but Im under no delusion it will take only one "conspiracy theory", maybe make a tally?
This strikes me as having several problems:
1. Whether the anti-ISIS envoy lied about something is somewhat unrelated to the question of whether any other government officials lied about something. As the President, I'd be somewhat shocked if there's never A PERSON who lies to you about something, and that probably shouldn't make you assume everything you hear is false.
2. Why turn around and trust conspiracy theories? If your trust is lowered, that doesn't explain why you think random anonymous posts on 4chan or whatever are now true. It's a selective demand for rigor. It's the same thing as "sometimes medical studies are p-hacked so now I just do what my shaman tells me".
A lot (most?) of Republicans believe some things at this point that the mainstream would call conspiracy theories. To be fair though the mainstream did call “Covid was a lab leak” and “Biden has dementia” conspiracy theories and they both turned out to be true and no one has resigned or apologised. So it’s really hard to tell at this point where the truth lies.
It was also a conspiracy theory that twitter and other social media companies were shadow banning people. The reality of a secret government conspiracy to censor people across several platforms was much, much worse, only discovered when Elon bought twitter. Which is kind of insane, that a conspiracy with so many people involved managed to stay secret.
Reversed stupidity is not intelligence, and also, lab leak has not been proven, and most likely (according to domain experts, the Good Judgment Project's superforecasters, and the results of the Rootclaim debate) Covid-19 was zoonotic in origin.
> Reversed stupidity is not intelligence
When it comes to believing sources that becomes cyclic; if alice has evidence bob lies, eve asking why alice is paranoid about bob citing a story bob says is ripe for abuse and should.... bi-partate(?) your probality distribution
This would probably be a decent argument, if it wasn't for the fact that Trump claimed 2012 was rigged, the 2016 primaries were rigged (against him), and the 2016 general was rigged (against him), so the causality is nonexistent. He has seen and claimed baseless conspiracies long before anybody offended him.
I dont understand how this is trump specific; bobs rain of terror has been going on to long and alice is Awesome For America.
alice: <any negitive claim about bob>
eve: "bob claimed you were bi-polar"
alice: "... Im not, please treat any surprising claims from bob as cover for more evidence as bobs bahavor"
eve: "Reversed stupidity is not intelligence"
alice: that doesnt make sense here, you either can believe Im mentally ill, or bobs extremely manipulative, its either or, the middle doesn't make sense here
The cia slur "conspiracy theorist" is an accusation of mental illness, while being a universal cover for accusations; while inverse stupidity isnt predictive for say gambling systems its not that bad a heuristic for criminal justice, if you treat lies near murder scenes as honest mistakes your going to be very confused.
Sure, but if someone lies to you, and then covers it up and pretended like they didn’t, you’d do well not to trust them again.
The point re Covid is not so much that lab leak is proven as that it’s a serious possibility that serious organisations like the CIA are giving credence to. And yet for a solid year people were literally censored from social media and tarred by the press if they claimed to believe it. Huh?
"To be fair though the mainstream did call “Covid was a lab leak” and “Biden has dementia” conspiracy theories and they both turned out to be true and no one has resigned or apologised. So it’s really hard to tell at this point where the truth lies."
Well, there have at least been preemptive pardons ... Though that apparently means Fauci and other beneficiaries can no longer take the fifth in case someone wants to investigate further.
Trusted information is harder to come by, but I feel like that obliterates the distinctions between some of these claims, e.g. it's premature to say COVID was definitely a lab leak (or definitely not one), but it's bananas to suggest Hillary Clinton has her email server hidden in Ukraine.
As for the government funding partisan coverage - that has more of a veneer of plausibility to it (all administrations do at least SOMETHING to shape press coverage), although I feel like it goes off the rails to suggest that Politico, that most nihilistic cheerleader of any "savvy" politician, is the one in the bag for Democrats.
> but it's bananas to suggest Hillary Clinton has her email server hidden in Ukraine.
That seems like a bad game of Chinese Whispers downstream of the actual proposition.
The idea is not that Hillary Clinton's email server was physically in Ukraine. It was in her house. The idea is that Hillary Clinton's email server was insecure, which means that every country with a vaguely competent spy organisation has a copy.
You think this is what President Trump believed?
It's what he said on the campaign trail at the time. He repeatedly said that it was insecure so that foreign adversaries, particularly the Russians, were reading our classified information.
This was reported in the press as Trump communicating to Russia to steal our classified documents. Which is a stupid or partisan (or both!) take on what he said.
He said repeatedly on the campaign trail that maybe the Russians had it after they hacked into a DNC email. But why Ukraine specifically? I feel like this is backing into a rational reason, but there WAS an Internet conspiracy theory that Clinton's email server was spirited away to Ukraine because of something involving CrowdStrike. And there's no example of Trump asking any other world leaders about the same thing.
I haven’t been following the lab leak story since Scott reviewed the Rootclaim debate. Has there been any new info since then, other than the CIA changing position?
There's an awful lot of information that was AFIK never considered in the Rootclaim debate in the first place, including but not limited to the independent assessments of half a dozen other intelligence agencies.
There is not enough information for anyone to have high confidence in their belief that the Covid pandemic did or did not come from a lab leak. But to treat the Rootclaim debate as dispositive, is folly - that process was too heavily dominated by the question of which party was the better debater, when both were chosen for reasons only weakly correlated with their debating skill.
I found the arguments fairly persuasive. Can you give an example of the evidence not considered in that debate, other than the intelligence agency assessments?
Also, it was clear that both estimates in the Rootclaim debate were so ridiculous as to make their analyses obviously wrong. From memory the wet-market guy estimated the chance of lab leak at 10^-20 or something, and if your answer is coming out that extreme, particularly on an issue where there is some expert disagreement, then there were mistakes made, either intentionally or not.
Yeah, they really needed to stop adding zeroes, as Scott put it, and everybody else needed to notice that all the zeroes were a clear sign that something significant was being missed.
https://slatestarcodex.com/2015/08/12/stop-adding-zeroes/
There will never be more information about a cover up in a totalitarian society, who edited the photo with slatin and 3 others, down to just stalin? If we do know, was it from the collapse or was this something a Russian could've figured out? Either you experience a blunt fact of feeling lied to or you swept up by the changing tides and quick talking.
It’s an authoritarian society. Not totalitarian. And even then it’s relatively uncontrolled and the use of smart phones makes it unlike the USSR for instance. Information leaks out.
I’m not sure exactly what you are talking about here. What was edited and where? Was it edited in China or the US. Wasn’t it the US and the west that decried the lab theory as conspiratorial?
> What was edited and where?
https://s3.crackedcdn.com/phpimages/article/7/0/7/810707.jpg
I think the proper respoence to a virus leak wouldve been burning the place down; but they likely did some amount of cleaning the day of that would wipe away most evidence, withholding a few samples here or there, and we wont get a definitive causal link.
> Wasn’t it the US and the west that decried the lab theory as conspiratorial?
The whole world did, america gave them money, but the final causal link will always be in china and we happen to have plenty of times chinas lied to itself and others.
> It’s an authoritarian society. Not totalitarian.
Is holodomor fake as well?
I dont care about definition debates in any way shape or form; but I loath apologism of genocidal states
Trump famously gets most of his news from TV, so although his access is theoretically larger than ours, his information diet is probably much smaller, although unique since I'm sure he does listen to the occasional security briefing. My pet theory would be that Elon has let his success and power go to his head, surrounded by sycophants, he ignores anything that doesn't conform to his biases and fulfil his pathological need to save the world.
I'm trying to find a quote from a quantum mechanics researcher. It was something like "there's more than 1 reality, but less than 1 reality per person". Does this ring a bell for anyone?
No idea why I can't find it with Google, AI, or my browser history... maybe I'm jumbling some different ideas together.
Possibly Christopher Fuchs.
His views here: https://www.quantamagazine.org/quantum-bayesianism-explained-by-its-founder-20150604/
seem consistent with that sort of quote.
The closest I have found is David Deutsch in The Fabric of Reality - “Every time we perform a quantum measurement, the universe splits into multiple branches; yet as conscious beings we experience only one branch.”
I listened to a fascinating 4 part discussion with David Deutsch on the Naval podcast where the quote or similar ideas may have surfaced.
Is there any kind of prediction market, or other forecasting service, that makes long-term forecasts of salaries in certain fields? I'm thinking something like a prediction market for "median entry-level salary of job title X in field Y in 20 years".
Such a forecasting service would be extremely useful to those planning their careers and figuring out what they want to specialize in. Especially now, where we expect that AI will cause changes in which skills/fields are lucrative and which are not valuable, more rapidly than any other point in human history.
Ex: 5 years, ago, majoring in computer science was an extremely lucrative choice, and everyone expected that to be the case long-term. Now with LLMs getting pretty good at programming, it's no longer certain that computer science is a good choice of skills to specialize in. Where can we find forecasts about changes like this, for the purpose of people figuring out what career they want? There is plenty of data about current job numbers and salaries from places like the BLS, Glassdoor, etc, but I'm not aware of any place to find long-term forecasts.
I'm writing this as someone seriously considering switching careers, from my current work which is mathematical/computational, to more hands-on scientific/experimental work.
I'm pretty sure that if AI deprecates software engineering, they will soon after rapidly deprecate every other type of work.
The reason we're tackling software engineering *first* is that it boosts AI development giving you a nice exponential growth cycle.
But software, even written by humans, was already eating the world. AI written software will just eat the world faster. All the jobs are going away and never coming back.
>I'm pretty sure that if AI deprecates software engineering, they will soon after rapidly deprecate every other type of work.
I strongly disagree with this. AI has the potential to deprecate most of the jobs that consist exclusively of computational/logical/mathematical work, but not work that involves touching and moving things in the physical world. Robotics is way, way harder than LLMs.
Even if the AIs automate the vast majority of work that can be done solely via interacting with a computer, they are still likely to be far inferior at physical, hands-on tasks like: resetting a broken bone, stitching together an open wound, hands-on engineering work that involves prototyping/putting together novel devices, mechanical/repair work. Even if we could in principle build specialized robots that could tackle some of those tasks, humans are very likely to have comparative advantage over robots for such tasks.
The reason that software engineering is likely to be automated first is not the reason you claimed. In fact, that reason doesn't even make sense: the limiting reagent for AI development is not how fast and how well the software engineers can code, but rather the amount of data and computing time available, none of which will be improved by more and better software engineers. And also, the kind of algorithms we've come up with for machine learning. That could in principle be improved with better AI, although it's certainly higher-level work than software engineering.
The reasons that software engineering is likely to be one of the first white-collar fields automated away by LLMs are: (1) software engineering consists exclusively of humans interacting with computers via text, nothing in the physical world; (2) we have huge amounts of training data available for this task, including some of humanity's most impressive software engineering achievements which are open source; (3) there is little institutional inertia against replacing the workforce of software engineers en masse; software company CEOs certainly wouldn't object to doing that. Even if LLMs could currently do the jobs of e.g. lawyers, accountants, HR, there is institutional inertia in those fields that prevents them from being replaced by LLMs en masse. Contra software engineering.
" Robotics is way, way harder than LLMs."
I am far, far less sure of this than you are.
It seems like the biggest bottlenecks in robotics are mostly around software.[1] We have a lot of accumulated technology for making actuators and LOADS of technology for sensors. Connecting the two together in a way that consistently produces the results seems to be the major hurdle, and that's mostly software. Sufficiently advanced AI has 3 methods to cross that hurdle. It could:
1. Get good at writing traditional robotics software to control whatever systems it wants.
2. Train ML systems directly on various robotics problems
3. Be good enough at general intelligence to learn to control robotic systems directly.
Whether 3 happens probably depends a lot on how "general" general intelligence really is, and whether it can really cross barriers to systems that are way outside its training distribution. But some combination of 1 and 2 seem almost inevitable once you have computers that are good enough at writing the code. Reduce it to another programming problem (or another machine learning problem) and they get to bring all their inherent advantages to bear.
[1] To be fair, I'm not especially well-read about the field, so I'm somewhat low-confidence about what follows.
[1] could happen, but I don't expect the tasks that robots can be programmed to do via this method to be that far beyond the tasks that humans can program them to do now. If human-written traditional robotics software is limited to a small subset of what humans can do, then I expect LLM-written traditional robotics control software to still be limited to some subset of what humans can do, as long as LLMs are not vastly smarter than humans. If LLMs become not just somewhat more intelligent than humans, but many orders of magnitude more intelligent, then all bets are off.
I'm skeptical of [2] because it's difficult to collect anywhere near the amount of training data that we have for LLMs. For LLMs we have the entire internet. Leading LLMs are trained on trillions of tokens.
I'm also skeptical of [3] because the data that LLMs are trained on (text) is very different than motion and sensor data that is needed to control robots. I don't expect LLM intelligence to easily generalize to robotic control. And like I said above, it's difficult to collect the quantity of training data needed to train robot software to be intelligent at broad, general-purpose physical tasks.
So first and foremost, LLMs are currently the most visible and successful class of machine-learning models, not the only class. Talking about their limitations--especially regarding how they deal with training data--is very relevant for discussing [3], but only relevant to [2] if one assumes that exactly zero non-LLM approaches to machine learning in robots could prove fruitful. AFAIK current work (done by humans) along that line is not focused on LLMs, so I don't see why hypothetical AI-coders would suddenly switch to an LLM-only approach to that field.
Second, regarding the relative poverty of training data: I agree this is a non-trivial issue. I still don't think "LLMs only work because of massive amounts of training data, therefor robotic control systems will only work because of massive amounts of training data" is valid reasoning. In many cases the thing you want a robot to do is quite a bit simpler than what you want an LLM to do, and I think the "throw data at it until it works right" approach is far useful there. Even when one moves up to more complex environments, you have many more possible inputs, but they're still mostly governed by the same sets of rules. I expect that a minimal model for even a fairly general robotic system to contain MANY fewer parameters than a minimal model for a chatbot or AI assistant type program. Fewer parameters means less training data is needed to converge on an acceptable local minimum. It also seems like there's a pretty large cascade potential here: once enough ML based robotic systems are being used or tested, they're necessarily collecting sensor data that can be used for training[1], and quite a bit of it. LLM training data isn't cheap because it's easy to generate more, it's cheap because a staggeringly large amount of it already exists. Feedback data from various automated sensors has at least the *potential* to be actually quite cheap to generate once the infrastructure to generate and use it has been set up. (Going to reiterate that all of the above is still fairly low-confidence. Less "this is true" than "it doesn't seem obvious that any of this is false.")
Now, throwing humans into the mix brings back all the complexity that LLMs need to deal with, so I expect "robots interacting with humans type jobs are likely to be the last ones replaced. But *quite a lot* of blue-collar work of various sorts can be done without interacting with humans. Indeed, manufacturing robots are *already* quite common. So I think the potential to replace lots of the "jobs that involve touching and moving things" fairly quickly after the software jobs fall (if not before), but "jobs that specifically require interacting with humans" might be the big holdout. (Ironically one of the biggest focuses of robotics has been self-driving cars, which very much fall in the latter category: probably why so much investment in them has yielded so little return so far.)
[1] If anyone is tempted to analogize this to training LLMs on the output of other LLMs, I feel like it *should* be obvious why it isn't remotely the same. But I'll explain if necessary.
> Robotics is way, way harder than LLMs.
Is it though? We've had robots building your microchips, cars, etc. for decades before LLMs even existed.
> (1) software engineering consists exclusively of humans interacting with computers via text, nothing in the physical world
My meeting schedule disagrees!
"humans interacting with computers" also applies to stuff like drone pilots, and drones definitely impact the world physically.
> Is it though? We've had robots building your microchips, cars, etc. for decades before LLMs even existed.
Let me clarify what I meant in that previous comment.
Robots learning to do hands-on, physical tasks by observing what humans do (like what LLMs do), instead of being explicitly programmed by humans (as is primarily the case for the examples you listed) is a very hard problem. Because we have much, much less training data available for those physical tasks. LLMs are trained on ~trillions of tokens. We can't collect that kind of data for general-purpose physical tasks.
This is why I am very skeptical that we will ever produce robots that are very broadly general-purpose useful at physical tasks. Not ones they've been explicitly programmed to do, but ones they've learned to be good at by observing what humans do a la LLMs. Robotics will surely continue to advance at tasks where we can specifically program them to do exactly what we want (as in the examples you listed).
> My meeting schedule disagrees!
Yes, software engineering, like all potentially-automatable white-collar jobs (e.g., law, accounting, HR), involves meetings. But the actual product that software engineering work produces that has value (code) can be produced exclusively by interacting with computers. LLMs are likely to be able to produce that valuable product, without all the meetings that humans require to get that work done, in software engineering and in other white-collar fields.
> "humans interacting with computers" also applies to stuff like drone pilots, and drones definitely impact the world physically.
Remote drone piloting may very well be automated. Although this is one of those fields that has, as I called it, bureaucratic inertia against automation.
I will add one more thing. Even if some of these hands-on, physical tasks I describe are potentially automatable by robots, I suspect that humans will have a comparative advantage over robots in these tasks, and thus will still be employed in these tasks. I .e., I suspect it won't make economic sense to try to automate all those physical tasks, because those resources could be used more efficiently to automate other tasks that are easier to automate.
It might not be as hard as it seems. A robot can train itself it move just like a baby mostly does.
https://www.youtube.com/watch?v=slmHMMKd6To
Yea, I get it's not there yet. The question is is it 5 years away, 50, 500, 5000? I think it's closer rather than further.
Mostly agreed, with a probable delay from white collar jobs eaten by AGI to blue collar jobs that also need robotics advances. ( As in the previous Open Thread: If ASI actually follows quickly after AGI, then we get a _very_ weird intermediate period where ASI exists but doesn't have robot actuators - the year of the immobile god. )
If humans can pilot drones, AI can pilot drones. I don't think the ASI will be immobile very long...
Aerial drones are easy for AI pilots to manage because the airspace is an uncluttered environment. Conditions are much more complex at ground level, so expect a long delay between machines being able to dust crops and being able to fix a leaky pipe in your house.
During the "very_ weird intermediate period," I think you will see a boom in apps that instruct humans how to do tasks through augmented reality glasses.
Many Thanks! I do agree that anything that can be controlled remotely today can be AI fodder once AGI exists. But there are also lots of cases today where humans go into places where no actuator currently is, and which are hard for e.g. drones to get to. I don't expect ASI(s) to be immobile long, but there might be a few years of "bending tin" to make enough mobile robots to extend ASI(s) reach to all of the places humans get to today. We shall see!
If the service becomes popular enough, I wonder how it will deal with the fact that the forecasts it produces will influence many people’s career choices, affecting the resolution of the forecast.
I'm working to create a framework for a writing competition where the authors entering are also the judges. The aim is to fairly distribute the work of judging among the entrants, make a system that is difficult to game, and provide clear incentives for everyone to participate equally. The goal is a decentralised process that steers readers toward the best indie writing which is otherwise easily lost among the countless titles on online distributors.
Any suggestions on the best technical platforms to coordinate such a project would be appreciated. Ideally I would like a method that is offline, robust and easily reproduced by anyone who wants to run a similar competition of their own, so I was wondering if something like plain old excel linked to automatically send out emails might be viable, but open to any ideas.
I wrote about it here- https://open.substack.com/pub/haldanebdoyle/p/when-sunlight-hits-the-forest-floor?r=f45kp&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false
Any thoughts on how to structure such a system to make it efficient, fair, and resilient to manipulation would be appreciated.
my partner has previously participated in https://writingbattle.com/, which sounds similar to what you're proposing.
Part of their setup to keep things fair is, participants are assigned stories to judge from genres other than what they submitted, so there's no benefit to your own performance to unfairly judge others. And participation is large enough and anonymous enough that conspiring to trade favorable reviews would probably be very very difficult, especially to do so undetected.
it's a slightly different space than you propose, given that it's short stories written strictly during the contest, rather than pre-existing novels. but, hope it helps as some prior work.
Wonderful! This is exactly the kind of thing I need to study. I will be running a trial with flash fiction first on a small scale, but investigating similar existing models could save me some missteps.
Thanks for the UChicago Rat and EA group ad!
One of us! One of us!
Wish I'd been more involved in the EA stuff lately, used to be a regular at the meetings, but got busy and dropped off the map after everyone at the retreat got covid...
Haha, yep. Would love to chat. You should send me an email at dnbirnbaum@uchicago.edu.
Morbid question: Is there any prediction market on whether Trump does a full Andrew Jackson, directly defying an explicit Supreme Court ruling, at some point in his term of office?
I would say it's unlikely, mainly because I don't expect the supreme court to make ruling that disagree with what Trump wants.
Why do you think that? In his first term, he lost a whole lot of cases in the supreme court (and in a bunch of lower courts as well).
And it is just such disagreements that would make this market interesting! Anyone see one out there? ( I tried a Google search, but unsuccessfully - but the phrasing could be quite tricky, so I don't know if I misgoogled or if no such market has been created yet. )
They’ve done it before by rejecting his post-2020 election lawsuits
Many Thanks! I think that you are right, but I don't think that it is vanishingly unlikely. It could be an interesting market even if the odds are, say, 25%, with ups and downs as events unfold.
Whats the best strategy for a below average guy on dating apps? Liking most profiles destroys your algorithm, but its not like being selective with your swipes increases your odds with the girls you like either(they still get hundreds of likes from other guys). For reference, I get 1-2 likes per month and have been on 2 dates after 6 months of being on a dating app(few more matches that went nowhere).
What makes you say you're below average? It's possible hang-ups aren't specific to using apps. If you're healthy and well-groomed, ought to be able to do alright given time.
Apps are a shit-show today I hear, but at the time I used them, you could directly message, and that's all I ever needed. No "likes" whatsoever. The key was a good, sane message that drew from their profile, honing in on certain interests particularly ones that might overlap with mine. Even with "likes" now in play, taking care to do this will go a ways.
You can cast a somewhat wider net by joining drop-in sports or meetups. If you're a student, then you're in the best possible position because you're tripping over prospects.
It's a numbers game, and while it's true women are hounded in high volume, they also have to sift through garbage. Low-effort conversation, and then past this point, "players", which means they get back on the app. The turnover rate is fast, just got to be in line. Stand out by appearing bright, having your shit together, not being desperate, and being kind.
No offense, but this comment has strong old man vibes, like saying "I don't know what the job market is like now, but back in my day you could just walk into any place of employment and get a job, so a firm handshake is all I ever needed, that's my take".
I hope you got something out of your weird projection. The changes Match group made to forerunning apps wasn't that long ago.
Are you already paying for the app's highest tier?
travel
https://putanumonit.com/2016/02/03/015-dating_1/
Key advice: make your profile divisive/controversial/distinct. Don’t try to optimize for the highest average level of attraction; optimize for the highest level of attraction for some women. You are some women’s type, so lean into it as much as possible.
I think the best strategy for a below average guy is to not get on the apps in the first place. Things are rough enough on there if you are good looking that it seems like a waste of time of you're not.
Join a board game group. Join a co-ed soccer league. Join a trivia group. It is actually extraordinarily easy to meet women if you get off the internet.
I recommend trying a different dating app where hot girls reach out to below average looking guys. LinkedIn is particularly good for this. 😜
More seriously, if you are a person of enough means that you can travel internationally without too much trouble then international dating actually does flip the script. I did it (eventually found wife this way) and it is extremely easy for a man to get responses/bites. The one big caveat is that you have to be *very* careful of scammers. There are a ton of scammers who do online international dating as their primary source of money, just convincing guys to send them money. These are often real girls in the country you are looking so you can't do a simple video call to find them out.
Assuming you can filter through the scammers and visa farmers, there are a lot of great options available outside of the US (which is where I assume you are at, please correct if wrong).
I'll second this as good advice, I've lived and done business overseas for many years, and it's a huge difference versus being in the states.
I wrote a fairly popular post going over the pros and cons and risk factors here for anyone interested in a little more context:
https://performativebafflement.substack.com/p/the-case-for-overseas-dating-and?r=17hw9h
I'm happy to answer any questions I can regarding the MiniCircle work I did with Ichor, though there isn't much beyond what I wrote in linked comment. I'm also happy to share results as they come in, though we probably won't have anything interesting for many more months. The real results we are after are lifespan (previous studies with AAV delivered VEGF suggested lifespan increase which we are trying to reproduce), which won't arrive until after maybe a year or so.
Some concerns I have include:
1) Contra frequent claims in this space, plasmids are not safe! They're only safe if not delivered by a real delivery vector because they don't get taken up by cells and basically nothing happens. If they're delivered by a real delivery vector such as LNP they end up in the cytoplasm and activate cGAS-STING which is incredibly toxic and inflammatory. There is a reason Generation Bio hyped up their mostly-single-stranded "immune quiet DNA".
1a) Was the Ichor CRO experiment literally just naked plasmid? Not even minicircle claims that works I think, they use PEI which is its own brand of problematic. There are ways to get naked plasmid to work-ish using various devices but none are mentioned here.
1b) What was the dose of the Ichor experiment? You say that basically "ehh we can just scale it up" but without knowing the dose it's not totally clear that's true.
2) The Ichor folks should have done qPCR or RNA sequencing to confirm the presence of mRNA transcribed from the plasmid as an orthogonal confirmation that things worked.
3) There are two separate "1,000-fold" gauntlets to run. First the issue of the ng vs pg discrepancy that was pointed out by another commenter. Second is the issue that local injection in mice and humans are not all that different in terms of volume, but humans are 1,000-fold bigger than mice so local injections go ~1,000-fold less far! This is why you need systemic delivery to e.g. the liver if you want to scale from mice to humans for systemic administration
If MiniCircle uses PEI then we probably did as well since a goal here was to replicate their gene therapy delivery system (though delivering a different therapeutic). I'll ask Ichor and get back to you if I find that we didn't.
Dose given was 50uL of VEGF DNA (see note on page 2).
We did something different for the first 3 mice (plus 3 controls) because in that readout we got an aggregate of human and mouse VEGF, while in the latter ELISA readout we got only human VEGF as a readout. I reached out to them to remind me how the first readout was acquired. That being said, unless the ELISA we used was complete garbage (e.g., just spitting out random numbers) it seems unlikely to me that no expression occurred given that we saw some treatment mice with 0 (so the ELISA wasn't just reading mouse VEGF) and some mice with non-trivial levels (so the ELISA wasn't failing to see anything at all). I agree it is possible that the ELISA had a failure mode where it returns random numbers but since the mice are all essentially clones I am skeptical that it was reading something endogenous to the mice.
Previous AAV delivered VEGF in mice was suggestive of a significant lifespan increase by going from ~100pg/mL to ~200pg/mL in circulation. In the mice that showed a result we saw about 30pg/mL human VEGF in circulation, suggesting that we only need around 3x increase to get similar results. While humans are bigger, they can also handle larger injections, injections at multiple cites, and you can inject into fat easily (hard to do in mice, which is why we went into muscle). I'm not too worried about being able to increase by 3-6x during dose finding studies in mice, and I'm optimistic we can hit similar target circulation levels in humans should we get that far.
I agree with you that systemic would be much better and I am actively monitoring the gene therapy delivery space for better systemic delivery systems. The thing to beat is AAV, which is incredibly expensive and immunogenic. For circulating proteins, plasmids may be sufficient which is why I am currently pursuing this avenue, but they aren't a good solution for non-circulating proteins which are also of interest to me.
"50uL" is not a dose, it's a volume, the dose will be determined by the concentration of minicircle within that 50uL.
If they used PEI, be aware PEI is extremely obsolete as an in vivo transfection reagent due to toxicity. LNPs are much better as are PBAEs I believe (PBAEs are trickier to work with).
Increase in total VEGF might just be due to the injury from the injection. Inflammation and tissue repair will upregulate VEGF. There is a "control" but it doesn't say if it got a sham injection or what. In any case, you need a non-VEGF-encoding minicircle control in case the DNA is doing something that upregulates VEGF (specifically, cytosolic DNA massively increases IL-6 which stimulates VEGF release). Meanwhile the hVEGF might be right, though antibodies for ELISA detection can have cross-reactivity and ELISAs can be noisy so you'd need to run controls to be sure, and the controls aren't there at all.
I'm not too surprised ALT/AST don't respond to a local injection, the correct safety assay would have been a cytokine panel with serum collected after the injection.
The weights are in fact concerning contra Ichor's claim. There are signs of weight loss at 1-2 days, and if that signal is real, then you absolutely cannot just jack up the dose if you don't want dead mice.
You are correct that AAVs are immunogenic but the reason that people use them is looooong experience with nonviral delivery not working. So it would be really really weird if it happened to work here.
The best case you can make I suppose is that the effective dose of VEGF requires like 1,000-fold less delivery than the effective dose for say hemophilia which is why you need AAV for hemophilia but not for this? I mean, it's possible I guess.
Also please don't take VEGF if you're not already old, it'll be more likely to cause cancer than help you. Also PLEASE PLEASE PLEASE do not do systemic delivery of plasmid to humans with any kind of good delivery vector it will likely kill you. See eg recent Nature Biotech paper https://www.nature.com/articles/s41587-025-02556-5 for more on why you should not mess around with plasmids on the off chance they are actually doing stuff.
Thanks for the nature article! In other work I have funded unrelated to this project we really struggled to get DNA in LNPs working, they kept killing the mice. I haven't read it yet, but it sounds like the authors of this paper may have found a solution to that, which would be great news! In other positive news I recently had a talk with a startup that is doing viral-like delivery and think they can address the safety and cost concerns of AAVs.
I heard back from Ichor by the way and they reminded me of some details of the pilot that I had forgotten. The first 3 treatment mice on page 6 are the same 3 mice that were the 28 day mice on page 12. The page 6 results are from a survival bleed, and the ELISA we used required a certain amount of material so we had to dilute the samples to use it (1:5). The same ELISA was used for the page 12 results, but those were terminal bleeds so no dilution was required.
This obviously opens up a lot of questions like why did we see non-zero values for the control mice in the diluted samples when we saw 0 values for treatment mice in the non-diluted samples. One guess is that a pure serum interferes with the ELISA in some way. Also the 3 mice that were tested twice initially showed elevated levels of VEGF but then 28 days later they didn't, and we need to understand why that was. One of the tasks we have scheduled already is ELISA scouting. We want one that can work well with survival bleeds so we can monitor mouse VEGF levels throughout the study.
We do plan on exploring and better understanding all of this before proceeding with the full lifespan study. The thing that seems to be the primary disagreement from people is whether a low-cost study to get a quick "no" (but an unclear "yes") is worthwhile compared to doing a more expensive study that would give you more clear results on the first shot. This pilot's purpose was to just see if we got all zeros across the board, which is what many expected, and if so we would have abandoned the project. It successfully gave us a lot of questions that we can now pursue for answers.
So I would say my issue is less with the fact of the pilot, you're right that pilots are often good, hindsight is 20/20, etc; extremely defensible move to do a pilot. However the problem is that that Ichor did a bad job with the pilot. They injected an inflammatory material and then used an ELISA with obvious cross-reactivity concerns to a protein (mouse VEGF) that is induced by tissue damage and inflammation! (VEGF does have a very short serum half-life and the injury might be healed by a week after the injection, so this might not be a concern, but you absolutely have to do the control to check this.) Measuring ALT/AST instead of serum cytokines was the incorrect safety assay. Extracting RNA from the muscle and doing qPCR or sequencing cDNA would have confirmed transgene expression and been very cheap but they didn't do that.
The powerpoint was put together unprofessionally as well. A volume is not a dose. No delivery vector was specified. Controls were not specified (sham injection, minicircle injection of some other protein, untreated, or what). The observation of cross-reactivity and ELISA concerns were not mentioned not was the possible undiluted serum concern. This is all forgivable if they did the actual experiment well, doing science well and generating good reports are separate skills, but both good science and coherent scientific reports are somewhat reflective of like "having a clear organized mind" so to me it's a red flag. Especially a red flag to do such a shoddy job for a paying customer as opposed to an internal report.
Re "Also the 3 mice that were tested twice initially showed elevated levels of VEGF but then 28 days later they didn't, and we need to understand why that was." this is (assuming I'm wrong and you actually did get hVEGF transgene expression, which is possible) likely due to epigenetic silencing of the vector, or death of cells that were expressing hVEGF due to cGAS-STING, or possibly an immune response to hVEGF.
Anyway yes I'm not surprised the LNPs are killing the mice. The issue is not that LNPs are toxic but that the DNA cargo is toxic when delivered into cells, and LNPs are good at delivering cargo into cells. The fact that you don't see toxicity here is therefore a very strong signal that delivery is inefficient or the dose is too low, and all you are seeing is noise. Or alternatively maybe you did actually see toxicity here, since there was a bit of weight loss and cytokines were never measured, which is a good sign for genuine delivery and bad sign for translatability and the ability to scale the dose.
I just want to add one comment for anyone who may be reading this now or in the future. The slide deck provided by Ichor was already more than I asked for. I engaged with them regularly in voice calls and every so often they will bring a presentation aid (slide deck) with them to one of those calls. I was the one who thought it was a good idea to share those slides with the world (even though they were missing a lot of details, as you outline here).
I'm quite confident that if I were to ask for a more formal description of everything they did or even a research paper as output they would have done so (and likely successfully), but I had no desire to pay for that work since in the end, I'm looking to convince myself that a particular therapy works and only then worry about convincing the rest of the world that the therapy works. There was more information provided to me during our calls over time and they have always walked me through everything they are doing in as much detail as I desire.
I am a huge proponent of working in the open and sharing any data I have, in whatever form it happens to be in, rather than keeping things under wraps because I am worried that the data isn't presentable. I would hate to see people look down on Ichor because of the choices I made (like sharing a presentation aid on the internet).
I'm considering writing a piece about this, summarizing the discussion for those (like me) who aren't specialized in molecular biology. Including your work and registering everyone's hypotheses in expectation of your results in a year or two would be neat. Is there a way I can contact you for more extended questions over the coming weeks/months?
For the sake of disclosure, my interest in this is both that I have a personal interest in questions of scientific validity and replication, and that people in my network are involved in Prospera and they should have a way to keep abreast of this.
You can send me an email with username `finicky_outboard089` and domain name `simplelogin.com`.
If the singularity (or something like it) happens, it seems natural that land would skyrocket in value (given that there's a finite amount of land on earth). This... doesn't seem like it's priced into land prices at all? There are lots of places in the US where land is ~$1k/acre. Does it make sense to just... buy some?
Yes, but with an important caveat: If your objective is to make the most money, then you shouldn't buy the acre of land now. The Singularity might not happen for 50 years, over which period your land might barely increase in value. You're better off investing your $1,000 in an index fund or some other good asset until the Singularity seems near. At that point, cash out and use your investment--which might be $10,000 by that point--to buy the acre of land. Even if the acre has appreciated in value to $6,000 by then, you still make more money overall.
Finally, a quick search on Zillow shows there aren't actually any $1,000 acres of land for sale in the U.S. anymore. The best we can do is a half-acre desolate desert lot near Demming, New Mexico: https://www.zillow.com/homedetails/Pecos-Rd-Deming-NM-88030/446067490_zpid/
For a whole acre, you will need to up your game to the whopping price of $1,500: https://www.zillow.com/homedetails/Olympic-Rd-SE-Deming-NM-88030/444454211_zpid/
Even after the Singularity, I can't imagine why humans would want to live in places like that, and you might also wait many years before AGIs ask to rent it from you for something.
As far as space goes, there's still plenty of land, and if demographics continue in their current trend, there's bound to be a whole lot more.
My personal opinion is that, yes, real-estate will become a coveted asset, but only highly desirable real estate. Also, what constitutes "highly desirable real estate" is bound to change as economic opportunity loses importance. I expanded a bit on it in a comment from a previous open thread: https://www.astralcodexten.com/p/open-thread-364/comment/85973533
Personally, I recently bought a flat. This was a big part of my reasoning to do it.
I don't think it's certain. But I used to move a lot and liked the flexibility of renting, so had a preference for that.
I think there's a very real possibility of AI impacting loads of jobs and can see land value skyrocketing. So bough a flat basically as a hedge.
Forget surface land, the REAL profits are going to be in reselling the Mantle and Core of Earth. Basically worthless now, with *astronomical* value and potential later. 🤑🤑🤑
How do you get it? Simple, it legally comes free with the surface land in certain countries like the UK.
Semi true of the UK, the state owns oil, gas, coal, gold and silver
Only if you think
(a) concepts things like private land ownership are going to be meaningful post singularity (including the concept of money, generally)
(b) the factors that presently contribute to some land being available at $1k an acre will go away such that it will be profitable to own that land
Good points.
For (b), yeah, the idea would be that a huge increase in general productivity should make all natural/scares resources more valuable. E.g. this $1k/acre land was probably worth a lot less (maybe free?) before the industrial revolution.
For (a), I guess I'm thinking we should plan for futures in which private property/money are still meaningful, since if they're not, then nothing we do now really matters (unless it affects the singularity itself).
> E.g. this $1k/acre land was probably worth a lot less (maybe free?) before the industrial revolution.
I don't see how that would work; the industrial revolution is the story of marginal land cratering in value over time. The more people live in cities, the less valuable land is.
The homestead act of 1862 (according to chatgpt) gave 160 acres of land in the west to any settler who lived there for 5 years. So the land was free, you could just go and take it. Also, the US increased frim 4 mil to 75 mil in population during the industrial revolution, with the same amount of land, so it's bound to have gotten more valuable
Being required to farm land for five years is not in fact a zero cost.
> Also, the US increased frim 4 mil to 75 mil in population during the industrial revolution, with the same amount of land, so it's bound to have gotten more valuable
Only if that's the only thing that happened during the industrial revolution, which it isn't.
> The homestead act of 1862 (according to chatgpt) gave...
This is exactly the kind of thing you wouldn't want to rely on ChatGPT for, but it is true that there was a substantially similar program. Your details might be correct.
I suppose it depends what directions the singularity takes us!
Obviously it's hard to think usefully about post-singularity scenarios, but I *guess* in a lot of the survivable ones, humans will still value things like hanging-out-with-other-humans (towns, cities, megalopolises, etc) as well as the aesthetics that have led us to want there to be large areas of undeveloped wilderness, forests, national parks etc.
Productivity alone isn't necessarily the only metric by which we value things in our current society, and I don't think that would necessarily change even if our overall productivity rises. It might be more strategic to buy e.g. shares in a silicon or cobalt mine than land that's not currently being productively used - although there's a very high chance I'm totally wrong about this, also!
In 2020 I bought some *extremely* unproductive (as in 'poisoned from surface mining') land out in the middle of nowhere, mostly to satisfy my own desire to be able to have a shack in an isolated space when I want to get away from other humans. I might be productive there in my shack writing or whatever, but I don't see any highly agentic AI's offering to buy me out any time soon :-)
Land was much more valuable (relatively to other things) before the industrial revolution.
Hmm, really? I get the sense that the total sum of the value of all american land would be higher now than before the IR. Granted this is mostly based on intuition. Why do you say it was more valuable then?
Agriculture was by far the biggest part of the economy for a long time, and land value was mostly land value for agriculture. Now agriculture is a much smaller part of the economy, and the value of (non-city) land compared to other things is down.
Finally, some bipartisan changes from the new government! No more pennies!
https://apnews.com/article/trump-penny-treasury-mint-192e3b9ad9891d50e7014997653051ba
Man, that's it? I thought for sure we'd have a chance at whacking multiple coins if we ever got rid of one.
The penny in 1850 was worth about 30 cents in terms of buying power. It's beyond irrelevant today, along with nickels and dimes. I think we could eliminate ALL coins and be fine, but if you had to keep one, only keep quarters.
We should stop printing One dollar bills and finally switch to a coin.
The best thing Stephen Harper did for Canada was eliminate the penny. Now Trump can have a positive legacy too.
A whole new spin on "penny wise, pound foolish"? :-)
Next stop: daylight saving!
This change makes no cents!
LOL! Many Thanks!
The board recently resigned en masse over issues with the founder, Michael Sparks—who wasn't even using his real name. Scott Alexander made a serious misjudgment backing this team. Just because an idea sounds great doesn't mean the team is—in fact, that's exactly the kind of failure mode that attracts grifters.
The post I replied to was recently deleted after I replied, but for the sake of preserving the discussion, here is what it said (without naming the user for privacy):
"The Far Out initiative is dead. The company relied on the idea that editing two genes could replicate Jo's condition, but her condition is far more complex and not easily replicated. Essentially, $50K wasted—unfortunate."
I'm sharing this to ensure the conversation remains intact and that context isn't lost. Otherwise, it may not be clear that my post refers to The Far Out Initiative (TFOI).
Eh, $50k is nothing for an org. It was a very cheap moonshot.
the $50k came from scott
A moonscott, huh?
I suspect Scott knew the odds then.
Well, probably not wasted because they tried and discovered (what I think most of us would have said) "oh hey, it's more complicated than our simple model!"
Sometimes there is a benefit to proving the bleedin' obvious. I hope experiments like this will make the people wanting to plough ahead with embryo selection for higher intelligence pause a bit, because here we go: it's not as easy as "just do some snip-snap with these genes and there you go!"
Definitely not wasted. A negative result is still a valid and important result. I'd be interested to read about it, but Google isn't showing anything/I'm looking for the wrong thing - objectivetruth, do you have a link or something?
A negative result is not valid when it was completely predictable that the result will be negative.
link: https://faroutinitiative.wordpress.com/2025/01/06/negativity-lite-more-than-you-probably-wanted-to-know/
That doesn't tell me very much at all, it's just a bunch of speculation. Your original post said "The company relied on the idea that editing two genes could replicate Jo's condition" and implied that they'd found it could not. I assumed that meant they'd actually done some kind of experiment beyond sequencing her genome (and the genome of this "Bob"), and was hoping for some actual data.