Probably true. I would like to say that I’m almost willing to guarantee that the next time we have a pandemic, we will screw it up just as badly - one way or the other. That’s what politicians are for; getting roasted for anything that doesn’t go right. Does that look anything like your planet?
“Numerous conspiracy theories circulated during the 1918 Spanish flu pandemic, primarily driven by wartime secrecy and a lack of scientific understanding of the virus. Many of these theories blamed the enemy, Germany, for deliberately spreading the disease as a form of germ warfare.
Key conspiracy theories included:
German Bioweapons: The most prevalent theory was that German spies or submarines were intentionally spreading the flu. Rumors ranged from agents spraying germs into the air in public places like theaters, to planting the disease on the US East Coast via U-boats.
Poisoned Aspirin: There were widespread suspicions that German-made Bayer aspirin tablets were poisoned with the virus. The American branch of Bayer even published advertisements to reassure the public that its manufacturing was under American control.
Vaccine-Induced Illness: Some people, already suspicious of government vaccination requirements for other diseases (like typhoid and smallpox), believed that doctors and nurses in military camps were intentionally inoculating recruits with the flu itself.
Wartime Byproducts: Other unfounded claims attributed the flu to poison gases used in the war, gases from decomposing bodies, air stagnation in trenches, or other environmental factors like coal dust or dirty dishwater.“
And,
“numerous conspiracy theories and scapegoating narratives emerged during outbreaks of the bubonic plague (specifically the Black Death in the 14th century), primarily because the medieval public and medical professionals lacked a scientific understanding of disease transmission.
Primary Conspiracy Theories
Poisoned Wells: The most pervasive and deadly conspiracy theory was that Jewish people deliberately caused the plague by poisoning wells and water sources to harm Christian communities. This false belief, despite Jews also dying from the disease, led to the massacre and burning of thousands of Jewish people and the wiping out of entire communities across Europe. Some historical analyses suggest that a potential genetic difference in immunity (related to the FMF gene mutation) may have contributed to slightly lower death rates among some Jewish populations, which in turn may have tragically fueled these baseless accusations.
Divine Punishment: A widespread non-conspiracy theory, though related to the lack of scientific explanation, was that the plague was a punishment from God for human sinfulness and immorality. This led to intense religious devotion and the rise of groups like the flagellants, who whipped themselves in public processions to seek forgiveness.
Astrology and "Miasmas": Medical professionals often attributed the disease to astrological alignments or "miasmas" (noxious air), which, while rooted in the limited medical knowledge of the time, contributed to the general confusion and fear that allowed scapegoating to flourish.
Blaming Other Minorities: Witches, Romani people ("Gypsies"), and other marginalized groups were also targeted as potential culprits. In later, smaller outbreaks of plague in France, lepers were accused of spreading the disease.
These historical reactions demonstrate a recurring pattern during pandemics where fear and a lack of information lead people to blame outsiders and minorities for an unseen threat.”
OK, but that's been known to the public since the pandemic, hasn't it? There was a lot of press coverage of doctors, nurses, firefighters, cops, soldiers, etc. resisting vaccine mandates.
Yes, so why are you assuming that they're going to be loyal to your personal cause? You understand that they're not robots that are programmed to follow orders of the sitting president, yes? Especially if the election itself is seen as fraudulent...
Yes, all you need to do to stop fascism is to tear apart all checks and balances and start jailing the opposition party. It's going to work this time, I just know it!
I mean, that was literally how fascism was stopped the first time around. It was just done by political parties external to the countries that had fallen to fascism. When you're far enough gone, having somebody walk you back from the edge at gunpoint really is the preferable option.
All that being said, I don't trust the Democratic Party to do any of that well or responsibly. If they actually did pull a Cincinnatus and restore some semblance of fairness and sanity to U.S. politics before abdicating power and calling fresh elections, I'd do a happy dance. But even if I trusted there benevolence, after the past 10 years I sure as hell don't trust their competence.
I think it’s important that those times it was stopped before involved an end to the current constitutional order and the start of a new one. That’s not what it sounds like you’re talking about if you’re just talking about removing two Supreme Court justices and replacing them while keeping the current cabinet offices.
I would want this to be a situation where there’s a whole new Supreme Court type thing, which would likely have a different name and constitutional role, and even if it involved some people who were on the previous one, they would have to be appointed/elected/selected newly however the new system works, rather than doing it the Trumpy way of just breaking some of the rules of the existing system in retaliation for (perceived) unfairness.
There's an interesting discussion to be had about what the next President should do with his newly expanded powers, immunity, and how to deal with the legacy of this presidency, but you're approaching the questions from the wrong angle.
Whoops sorry I meant it’s an edited version with a diff ending - Nathan rewrote it to make it better, but that obvs also removes a bit of the ”this was pure AI” magic
The edited version diverges pretty early. Both versions make very little sense economically. This scenario - a very fast growth of money supply - is not that surreal, this has been tried many times, e.g. in Weimar Germany in 1923 and Zimbabwe in 2007 and the results are pretty consistent and well understood. The only fantastic assumption in this story is that the source of growth is a private bank account, and not a state treasury account. This assumption does not change the macro effects on the economy. The predictable macro effects are 1) a rapid hyperinflation where the price of all real assets is increasing in line with the money supply 2) leading to a loss of faith in the currency and replacement of the inflating currency with a stable one 3) with considerable frictional costs and winners/losers in the switch over. So, for example, the original version's claim that S&P500 price becomes one day undefined is not realistic - S&P500 will rise roughly in line with money supply, but a bit slower to reflect the frictional economy-wide inflation costs and the turbulence, then switch index denomination to euros/new dollars/BTC.
I think you're assuming the entity with the excessive amount of money is actually spending it, which didn't seem to be what was happening in the story.
Yeah, they say "S&P500 will rise roughly in line with money supply" but it's something like the money supply * the average velocity of money that's important (holding other things equal). So someone who gains a large amount of money and never spends it will have no effect
"Over the past several years, we have invested in building out our supply chain and continue to increase our vertical integration, enabling us to move at speeds usually unimaginable in the semiconductor industry. More recently, we completed our first in-house production-quality 300 mm wafer lithography tool, which operates at the extreme G-forces required to meet the throughput of a leading-edge fab."
"extreme G-forces required to meet the throughput of a leading-edge fab" ?????
Note also that getting x-ray lithography to work isn't the problem. Getting x-ray lithography to work at scale (100+ wafers/hour) and at cost is the problem.
> A lithography system (scanner) must work 24/7 with sub-nanometer precision, while accelerating mechatronic modules at incredible speeds. For example, the reticle stage accelerates at close to 16g and the wafer stage to 7g. That’s more acceleration than a jet fighter.
> It’s not possible to mechanically construct a machine capable of this level of alignment and precision, accelerating at those speeds, and with the level of reliability and repeatability required to make today’s computer chips without the help of in-scanner metrology.
I agree that ~10G of acceleration is impressive for those elements of a production lithography system, but it still feels hyperbolic to call it "extreme" for people used to high accelerations in other contexts (e.g. ejection seats, car crashes, piston engines, turbomachinery, centrifuges...).
Substrate almost halfway to their goal. All product creation can be broken down into two phases as described in this excellent TED-like presentation of a similar product: https://www.youtube.com/watch?v=DkGMY63FF3Q
The two phases are "ideation" and "development." Substrate is far along on the ideation phase and just has to hire some good engineers, which are plentiful. /s
If you're not already suspicious, it's very easy to get taken in by this kind of thing, because why would you think a startup was actually a complete fraud (like Theranos) instead of the normal kind of hyped-up gamble that might or might not get anywhere?
That's what "due diligence" is for, and it doesn't mean reading the startup's shiny prospectus and taking their word for it. If it's just you investing your own money, sure, be as foolish as you want. But it is exceedingly foolish to invest $150 million in a startup without e.g. hiring an independent expert to accompany you on a tour of their factory, signing an NDA and reviewing the technical details of their proposal.
And if not all of that $150 million is your own money, then it's your fiduciary responsibility to do due diligence.
Since this happened, and Theranos happened and FTX happened, it's pretty clear that Silicon Valley has some rich fools and that selling silicon snake oil to rich fools can be a viable business plan. But everyone who falls for this sort of thing, should be embarrassed.
Ben Thompson is an analyst, not a reporter. In his interviews with startups he explicitly states that he is trying to get the founder's perspective/narrative, and usually won't be able to get hard evidence to verify the claims they make https://stratechery.com/2022/an-interview-with-rippling-founder-parker-conrad/
"One of the recurring questions over the years that I have written Stratechery is why I don’t write more about startups (a related question on Twitter is where is the Stratechery of startups). The answer is pretty straightforward: I am an analyst, not a reporter, and there is a lot more information publicly available about public companies; information about startups is much harder to come by, and often unverifiable when you get it.
At the same time, I do regularly come across companies that are very intriguing to me, particularly in the context of the themes that I write about, and I want to learn more — and share those learnings with you. That is the context for a new type of interview I am launching on Stratechery: startup founders.
The idea is this: I sit down with a founder with the goal of learning more about their business and the opportunity that they are pursuing. Instead of objective analysis this is as subjective as it gets — of course founders are optimistic about their future possibilities — but that feels like the right approach to expanding coverage to companies that I feel constrained in writing about. I’ve had a couple of these conversations already, and have found them tremendously interesting; I hope you feel the same."
Yeah. I'm pretty sure this only goes to the Reform/ Reconstructionist/Conservative/Open Orthodox. If you ccount the ultra-orthodox, who are fragmented and informal enough to be basically uncountable, you'd probably get very different numbers.
I don't have enough info on the study methodology, but I would suspect it might be missing the largest Modern Orthodox Rabbinical School, and only include those that accept LGBTQ/female students.
> Additionally, the number of survey responses from RIETS and other Orthodox institutions—aside from YCT and Maharat—was very small. Because these responses were both limited in number and substantially different from other denominational groups, they could not be reliably weighted to represent their broader populations. As a result, survey respondents ordained through or currently studying at RIETS or other Orthodox institutions outside of YCT and Maharat were excluded from the survey analysis.
So yeah, the "majority of American rabbinical students are now women" conclusion is probably false.
RIETS is the only rabbinical school that really matters in the American Modern Orthodox world, at least anecdotally. (Other relevant institutions are either affiliated or not actually rabbinical schools.)
YCT does matter and YCT alumni actual run shuls and have been very positively impactful on Modern Orthodoxy (and the productive ones are generally are a bit more orthodox than YCT itself). Maharat drags YCT the institution down and is not modern orthodox
I can't find class sizes for RIETS or YCT, but RIETS is clearly bigger by a large margin.
‘doesn't matter,’ in this context, means that YCT isn't a representative sample of the direction of Modern Orthodoxy, as it is both small, and selected largely on the basis of being more modern/open than the RIETS student body.
Not sure what the reference here is, but the sort of Jews going to daily services, (which only men have to do) will be praying in places with male Rabbis.
Sounds like it. People who take the religion seriously as a religion stick to the traditional version, people who take it as an 'ethnic identity/God is unknowable, we're spiritual/Vague Deity Shaped Symbol/doing the right *actions* together is more important than everybody *thinking* the same' go the liberal secular route.
Yes, exactly. Although the biggest losers in US strains of Judaism have been the Conservative movement, which seems to be stuck in the middle of the road and not optimized for either set of preferences.
15. “so it’s not a contradiction that Democrats are against both “advocating for Israel” and “advocating for Palestinians” “
It’s not a contradiction at all. A pox on both their houses is probably the most common sentiment I hear. The position that Hamas is evil and what Israel is doing in Gaza is evil is the simplest.
Perhaps, but my view aligns with Hannes. A pox on both their houses. Everyone involved is awful*, everyone. Hamas, Israeli government, mostly the USA, Egypt, Qatar, the British and French governments past and present, the Turks/Ottomans, the UN, God, Allah, and everyone else who has had a hand in the situation.
Sometimes you can get a bit of an 'out' here by saying that the common people are fine, it's just those in power, but that's not true here: polling for both Israelis and Palestinians consistently show them to be ethnic purists, religious fanatics, and just general blood-and-soil for the blood-and-soil-God type people. The Israeli settlers are unambiguously extra awful and I struggle to understand those who defend them.
* OK, I guess Jordan gets a pass for being the least-evil entity involved.
The British were pretty non-evil, and so was the UN. When the Civil War started what was left of the British garrison, by that point not particularly more combat-ready than the insurgents, fought the Irgun out of Jaffa (where they were already starting to forcibly expel civilians) and repressed the Arab mobs.
Had the Arabs been smarter, they'd leveraged British patriotism and sympathy between Christians to get the British to stay there indefinitely.
It's pretty obvious from the context that I was reffering to their administration of Mandatory Palestine and their management of the crisis that was brewing. Keralans understandably had a different opinion, even Arabs after 1956, but at place and time we were discussing they were the closest to a reasonable actor
For a lot of people it's just not anything near a major concern. If you look at surveys of voters' top issues it's usually somewhere near or at the bottom. It's not even 'a pox on both their houses' as much as 'why are we investing so much effort and attention in a conflict halfway across the world when we have so many problems here?'
Ah yeah, absolute cynicism about geopolitics when it comes to Russia's actions, but also some random heartwrenching words should get Americans to stop investing in the conflict that has bled Russia dry, has forced their withdrawal from the Middle East, has denied them effective use of their bases in Crimea, and unified NATO like never before.
Ah yeah another fine geopolitical analysis based on categories that don't even exist (BRICS literally come from 90s financial punditry, it was supposed to be the countries where Western investors should look at for exhorbitant returns, Brazil or South Africa never had a coordinated strategy with Russia or China).
Russia is in such a better position after they lost most of their Soviet arsenal which are not able to replace, sure. They had to start scraping the barrel, exit from any other front, stop supporting their client state in the Middle East (which promptly collapsed and is now headed by a neoliberal Jihadist seeking to open trade with the US), and can't even safely sail in the Black Sea, such a bonanza.
As for "color revolutions", yeah, it's how the dumbest people with internet access call every protest that threaten their emotionally support dictatorships. Bc we all know regime propaganda is always so truthful and people are happy and loyal to the Dear Leader, the only possible reason we'd see protests is foreign involvement (and also America is the only country that would ever involve itself outside its borders, somehow). So poignant and insightful.
Nobody outside of Russia is advocating for more dead Ukrainians. Lots of people are advocating for Russia to get out of Ukraine and stop killing Ukrainians altogether.
And nobody but the Ukrainians gets any say in the question of how many dead Ukrainians might be a regrettable but acceptable price to achieve that. What the rest of us *do* get a say in, is if the Ukrainians are still going to be dying in the defense of their homeland, how many Russians will they take down with them?
I'm not the least bit tired of answering that question with "as many as possible, please and thank you, now let's see what other surplus weapons we might have available to send your way". And I don't think that I am alone with that.
#21. There are, at last count, nine candidates running for the Congressional seat. Perhaps three currently are serious contenders. Jami Floyd is one of the nine but (especially as she is not one of the serious contenders) describing her as "Bores' opponent" or the beneficiary of attacks on Bores is inaccurate and misleading.
Do birds predate mosquitos? Get birds to eat any mosquitoes in the zone, get cats to keep the birds from flying out of the zone, get dogs to keep the cats in check!
Who else is going to keep the cats on track? Any cat is likely to wander off, decide to hunt the birds for fun, or simply do the exact opposite of whatever it is you want them to do. You need loyal, trained, dogs who can stick at one task to supervise them.
That's uncertain, the earliest mosquito found in amber dates to 125 million years ago in the early Cretaceous, as does the earliest bird having a beak and feathers.
Also, only kings can be kept in check, cats may be a single colour or have stripes or blotches.
Exposure to and socialisation in the business world. At least that is what I was told back when I was a postdoc (in pure maths), as an explanation for different attitudes in engineering vs. mathematics/physics. What is more, people in engineering tend to be embedded in environments where they are exposed more to management-side views than labor-side views, especially from people they depend on.
Another way we might put this is that they're exposed more to the real world. Understanding of things like "how people respond to incentives" and "not everyone is like you in the way that they respond to incentives" and "no seriously, you have to pay attention to how people respond to incentives".
Yeah we all know that the right are the people who truly understand mechanism design and economics in general, as shown by their obsession for mean-testing, local autonomy no matter what, hard currency no matter what, macroeconomics as household accounting, and more recently protectionism, nativism and war on the most productive areas of the country.
Irony aside, a much more direct explanation is that when one realizes they'll soon be in top tax bracket, they want lower and less progressive taxes, and everything follows from it (there is something else, like a certain entitlement to high status and sexual access and frustration that money does not buy it, but I'm trying to be charitable here).
I wasn’t surprised, either. Not because of the internships, which I suspect are just as likely to expose students to high-COLA areas with liberal politics. Because of the relative affluence of a technical white-collar career. It’s likely to get people out of debt and into a house faster than teaching or whatever. I’d have been willing to bet on a strong correlation between political shift and median starting salary.
…But it turns out the study was normalized to pure science degrees, so engineering just didn’t shift people *as far* left.
Fun fact, Richard Nixon proposed a negative income bill called the Family Assistance Plan, and one of the largest opponents were the social workers unions.
Engineering has usually been thought of as relatively right wing in universities, behind only the business school and ag school (and sometimes parts of the law school and economics department, at a few universities). Also, during the early 2000s war on terror, it become clear that a lot of Al Qaeda terrorists we’re trained as engineers (as have a lot of Pakistani and Chinese government officials during their rightward moves).
In general I think engineering is for better or worse one of the few degrees that is truly apolitical. Some other are inherently political (say, any social science), some other are not but the current political climate forces one to take a side (say, biology or atmospheric physics). But no politician is going around claiming that research into high frequency sensors is murder, that Fourier transforms cause autism or that the Paris-Erdogan Law is a Chinese hoax.
So the most economical explanation is that 50% of the country is right-wing (and in the Muslim world, some percentage is Islamist), and engineering is the only one NOT forcing them to rethink their convictions.
I've read a lot of conservatives say having to deal with real-world consequences, where you can't BS your way out, drags you right. A building can't ask for accommodations, it's just going to fall down if it isn't built right.
Not sure if this is going to change with stuff like antivax becoming right-coded, but I've heard it in various places over the last three or so decades.
You'd need to be a particularly bad engineer or extremely unlucky (or, most likely, both) for a building you built nowadays to fall down. From what I see from my distant perch, the day-to-day work of an engineer often has more to do with fulfilling codes and specifications than with what the laws of physics will and will not actually allow.
And that's why in a biology lab everyone votes Ron Paul /s. The main difference between engineering and the sciences is that nobody goes around claiming that the Paris-Erdogan Law is a Chinese hoax or that using high frequency sensor for research is murder, so the ones who enter being right-wing can stay consevative by Gell-Mann amnesia.
I don't understand the intention of the writeup at link #1.
Setting aside the factual limitations of the stated premise -- is the essay intended as satire? Is the author trying to discredit the concept of US law assigning to corporations all the same legal rights that persons hold? Mocking the widely-held impression that US law actually does that? Something else?
I think it's attempting to contribute to the conversation about whether AIs could/should be granted some of the rights we currently grant humans, by educating that some of those rights are already granted to other non-human entities. "Legal Persons" is a term-of-the-art that has a different meaning than the commonsense meaning of the word "person" which is one reason it's confusing, and this is a fun educational attempt to correct that.
As a concrete example, if a self-driving privately-owned car gets in a crash, it's not crazy to think that the _car itself_ should be represented by legal council and pay any damages from its own insurance, since trying to litigate whether it's the fault of the owner, manufacturer, or occupant at the time seems outside the domain of the car-crash investigation. Tennessee already has a law on the books to this effect. This is similar to how a ship can be arrested or corporation can be sued.
Re AGI / UBI being bad for labor, I just wrote an economics-grounded post arguing it's going to be great for labor and capitalists alike.
Labor, because we can give a generous $40k - $90k UBI to everyone, capital because returns to capital will at LEAST 2x, and arguably more, because now capital can be directly converted into whatever the highest value labor is.
Also, the fact that we'll be able to counterfeit high-value white collar work first is a "best of all possible worlds" scenario. To a pretty large extent, white collar job holders are capital holders, so they may lose their jobs, but the combo of capital value growth + UBI more than takes care of them, AND makes labor (non capital holders) better off, too.
Physical jobs have much lower margins for companies, contribute notably less to GDP, and would require substantially more hardware / capital expenditures to counterfeit. This is entirely ignoring the fact that robotics is much harder than software for numerous other reasons.
Having ended at a place where we can counterfeit only the highest margin jobs, that generate huge amounts of value both for GDP and the companies with those jobs, and in which the current job-holders are all predominantly capital owners who will largely be covered by company value growth in a much-higher GDP future? It’s so much easier on every front!
The economy has dectupled many times in history. At each of those times, we could have chosen to give people a generous (by the standards of the previous regime) UBI. I agree we will be able to do this again this time, but I don't know how likely I think it is. Yes, there will be many more unemployed people trying to get it. But labor will also have less power than ever. I think Philosophy Bear's argument - that we will either get a capitalist coup or a socialist revolution - is if anything too optimistic; more likely is that every industry argues for an AI ban or subsidies for themselves, the powerful ones end up as the equivalent of longshoremen, and the weak ones are left to fend for themselves for "every minority woman who lost her job to AI on a Friday in March gets a $300 voucher for health insurance during job retraining" style bandaids.
> And the idea that voters will be smart enough to prevent the robots-with-guns, or be able to control them democratically, is like a bad joke.
Agree, that's why being at a point where we're able to counterfeit only highly-comped jobs from people with capital and political relevance is a best of all worlds sort of scenario!
If instead, we could counterfeit a bunch of service jobs with robots? Everyone's doomed, boots on faces forever.
Or what about your scenario, where we have killbots and capital owners can retreat behind walled compounds patrolled by them? Boots on faces.
But here we are, with NO killbots, and right on the cusp of counterfeiting jobs, and the ONLY jobs we can counterfeit are highly valued white collar ones? It's a gigantic free lunch, societally, because we actually have a chance at navigating this gracefully in a way that doesn't end with boots on faces forever.
> Yes, there will be many more unemployed people trying to get it. But labor will also have less power than ever.
Seriously? The people whose jobs are being counterfeited are high income, connected white collar people. If even as much as 10% of them lost their jobs, you think we're telling them to pound sand, politically?
What if 20%, or 50% lose their jobs? It just seems really obvious to me that the public pressure towards UBI is going to be immense, and that it's much better that it's happening to richer and more connected people first this time, because it vastly increases the chances of substantial UBI actually getting pushed through.
I agree if it were reversed, and robotics were pushing 20% of service sector people out of jobs, they'd probably be screwed, and that's a much darker picture. But we live in that panglossian world where the reverse has happened, and it's going to be connected, well-heeled people whose voices are heard and payed attention to at the forefront of advocating for the massive societal rearrangement that UBI will represent.
The jobs of the well-connected white-collar people will be specifically banned from being replaced by AI. If an AI lawyer and a human lawyer both give you the same advice, but your conversation with a human lawyer is privilaged in court proceedings and your conversation with an AI lawyer is not, then the human lawyer will be far more in demand.
> The jobs of the well-connected white-collar people will be specifically banned from being replaced by AI.
Maybe, but I'm less sure of this. On the one hand, you have a bunch of lawyers and other white collar people slaving away in their cubes.
But on the other, you have a bunch of company owners and capital owners who want to pay a third as much as current comp while still getting all the value they represent - when it comes to "giant corporations and billionaires" versus "common schlubs," I generally think the billionaires get their way here, and we'll be able to counterfeit a lot of jobs.
And to navigate the billionaire side of this Scylla and Charbydis, to be able to UBI everybody at $40k+, all we have to do is keep current or 2018 corporate tax rates in place, which is the Schelling point already and a pretty reasonable expectation.
Sure, either side of those could go wrong, and there's some incentives pushing against the mainline outcome both times, but the way to bet is definitely that we both counterfeit a bunch of jobs, and keep a somewhat reasonable corporate tax rate.
And in that mainline scenario? I think we're fine - that it's great news, actually.
AI has seen heavy use in the legal field (along with outsourcing to India) because a huge amount of it is data management. Since a lot of law's financial structure is also a pyramid scheme where the higher often feel no real loyalty to the lower, your interpretation is 100% plausible. I will not be surprised if it doesn't happen (especially since regulators get a vote), but I will not be surprised if it does.
A lot of demand for lawyers doesn't care about which conversations are privileged in court. "Can I legally do XXX with my short term rental?" "What paperwork do I file for a divorce?" "I need to put together a will." "Is posting XXX slander?" "How do I legally fire an employee for XXX?"
"How do I legally fire an employee for XXX?" is absolutely the kind of thing that will cost you a multimillion-dollar judgement if it comes out in discovery.
I imagine the mom and pop asking this question aren’t at risk of losing millions of dollars because ten of their businesses aren’t worth that. The big business will probably continue keeping a human lawyer on retainer. Or maybe not even them, if AI chats have a reasonable chance of not showing up in discovery.
Your lawyer will be using AI. So lawyers will be able to handle many more clients at lower rates, but the lawyers who are less good at filtering out AI slop or less good at marketing will get weeded out. At first, anyway.
Sure thing. To my old lawyer, me switching to AI will look exactly like me switching to a lawyer heavily using AI. All of a sudden, I pay a fraction of the cost elsewhere, while a human becomes unemployed.
> I think you assume every job will be replaced. I think this is not true, especially not at first.
Actually in my post, I assumed ~50% of white collar jobs would ultimately get automated, and zero blue collar / service sector jobs.
And you're right! Non-white-collar people who hate their jobs will *definitely* quit and that's a good thing - comp will raise to match demand for the job until they're all filled, on the margins they will be filled by people who want to and are more motivated to do those jobs (many of them probably former white collar workers who want to keep earning), and consequently, quality-at-the-margins would probably increase in all jobs so shuffled.
And of course in actual practice, the UBI is probably going to be phased and targeted in stages rather than truly universal as we get to the final endstate my post was about.
The Gulf oil states have something like UBI (more like universal basic services + sinecures). Maybe gains from AI would be more like an oil windfall than normal economic growth because:
1) It happens suddenly before standards can adapt.
2) It doesn't require anyone to work to produce the gains so no one looks more or less deserving.
Jay Gould supposedly said that he could pay half the working class to kill the other half; sadly, AI is probably coming for the killing-the-working-class jobs too.
> we can give a generous $40k - $90k UBI to everyone, capital because returns to capital will at LEAST 2x, and arguably more
I'm very resistant to "giving everybody free money will actually make everything great" arguments. Too good to be true. The real world is a morass of tradeoffs.
Sure, in my post I posit that ~80% of people on UBI will use it to stare at screens and shovel junk food into their faces even MORE than the ~11 hours they spend per day doing those things today.
I read all that. The problem, as with all such prognostications on UBI, is that you ignore the transition to the glorious future and the intermediate demand side of the economy. You assume that capital can extract more value from fewer workers, profits will increase, and business can be taxed at present (or perhaps higher) levels of corporate tax. Hence the $95k.
However business revenue depends on demand. If the 60% of the economy is tanking there’s no profit to tax. The UBI has to come first. There’s no plan for that.
This will happen fairly quickly. Polymarket is predicting a recession next year (https://x.com/Polymarket/status/1997751003559412101) and - while i don’t see their reasoning there - AI will either be responsible in part, or in large part, and be blamed either way. The political response will be … interesting.
> you ignore the transition to the glorious future and the intermediate demand side of the economy
Yeah, I don’t think we’re too far apart there - in the post I elide this by just looking at the end state, but agree that transition can go better and worse, and is probably bumpy.
At the end state, obviously the UBI to everyone props up the demand, but in the middle there can be noise and dips and surges and who knows what else, I agree.
From the company point of view, it’s probably fine. Even if 10% of people in their prospective customer base have lost their jobs, being able to offer much cheaper finished goods or services will induce demand higher than that 10% deficit. And of course, any companies that can service overseas or non-white-collar demand are safe and pure growth at that point.
And from the macro perspective, if those 10% of people losing their jobs spurs a recession, so much the better, right? It puts it on the political radar sooner and puts some real emphasis behind it in terms of being able to successfully push for the UBI, and starts the process at a relatively smaller impact point, hopefully making the rest of the job counterfeiting smoother and less bumpy.
In a world where the vast majority of power and money is concentrated in the hands of the tiny number of people, what makes you think they're going to want to pay you for doing nothing?
Your best case scenario is that they'll want to pay you for doing something. Human servants surely have some kind of je ne sais quoi even if robots are available. If I were a quadrillionaire in a world full of otherwise-starving people I'd probably find a way to pay millions of people to sing my praises all day long while dancing in perfect formation.
> In a world where the vast majority of power and money is concentrated in the hands of the tiny number of people, what makes you think they're going to want to pay you for doing nothing?
I posit that we already live in that world, and already have corporate taxes that are paid!
Just a quick hand wave in that direction - NVIDIA is worth $5T, Apple 4T, Alphabet is $4T, MSFT is 3.5T, Meta is 1.5T - we're at 18T out of the US's 30T economy. Yes, market cap is fixed and GDP is flow, I said it was a handwave. My point is the power and money is ALREADY concentrated in the hands of a tiny number of legally defined "people" and things are fine.
The whole point of my post was that we don't need to EXPECT anything but minor amounts of basically "charity," and that today's or 2018's existing corporate tax rates are enough for everyone to plausibly be okay in the future in terms of generous UBI, and those rates are both pretty strong Schelling points and pretty reasonable amounts, especially if the alternative is pitchfork wielding mobs in a democracy.
> If I were a quadrillionaire in a world full of otherwise-starving people I'd probably find a way to pay millions of people to sing my praises all day long while dancing in perfect formation.
Such is how the content creator economy currently operates, naturally springing up from the demands of an ever enriched professional class wanting their unique interests covered. I think people are overestimating how much the economy will become concentrated as things derived from human labor become ever more precious status symbols. There is no reason to go see a play in person compared to a movie, but the practice persists because people are willing to pay big money for a unique experience. And few people call theater acting a debased dystopian occupation.
Aphantasia question: for questions like "Suppose you walk forwards 20 feet, then turn left and walk 10 feet, then turn 180 degrees and walk forwards 20 feet. What letter of the alphabet have you traced?" - how do aphantasic people answer these questions mentally? I searched a bit online but didn't find any of the answers cleared up my confusion.
I'm mostly aphantasic. With this task I'm doing something "visio-spatial" in my mind and don't find it super difficult. It's easier if I imagine physical movements or trace it out with my finger.
What's the difference between a visio-spatial mental operation, and a phantasic one? It sounds sort of like they ought to be synonymous.
Whenever I hear about aphantasia, I can't help but think "this is some sort of error in communication—like, people interpret 'mental image' as 'seeing something in front of your eyes as in a hallucination', and then think 'well, that definitely doesn't happen... I guess I can't do mental imagery!'"... but probably I'm wrong about this, since it appears to be a well-validated(?) "condition".
I'll try to explain what's happening. There's a blank three-dimensional space in my mind, I can use this to rotate shapes, orient myself in a new town, or remember how to get around my childhood home. I can move around this space, but there is no visual information being stored or accessed - no colour, no brightness or texture. It's only visio-spatial to the degree that I can "imagine" movement.
If you asked me to imagine my mother, it's impossible. If you ask me to imagine a sunset, I can trace a horizon line in this 3d mental space that seems vaguely sunsetty, but I can't maintain it. If you ask me to imagine an apple, I automatically imagine the idea of holding it in my hand and biting into it - I can imagine the crispness and the juiciness. But there's nothing visual.
Fascinating—thanks for explaining! That definitely sounds different from my experience (if anything, color & texture are, I find, *easier* & more vivid than spatial relations/movement); I guess my skepticism was, indeed—...as predicted, in my defense! [cough]—ill-founded.
I would like to know how it is possible to draw anything from imagination, then. Are you capable of deciding beforehand what you wish to draw, including details, then drawing it?
Not exactly sure if this is what you mean, but the question I think you're asking is: "To what extent can I "pre-imagine" something I'm hoping to draw?"
Not very much! I can imagine non-visual scenes in my head (in language or conceptual form), and then draw them - I think this counts as "drawing from imagination". But I can't by any means hold something visual stably in my mind and then draw it. This seems like a wild superpower.
For what it's worth, I'm a very improvisational painter/drawer - I love starting on a blank canvas without any plans and seeing what patterns emerge. I don't know how much this is related to aphantasia, though, because I have a similar approach to music and writing, I think.
I dunno about TT, but for myself the answer is simply that it's impossible for me to draw something from imagination, at least if it's supposed to be realistic imagery (vs, say, a diagram or something.) When I used to watch cop shows, before I knew about aphantasia, it was always deeply mysterious to me how a sketch artist could possibly use some kind of black magic to elicit a picture of a suspect from a witness. Now I understand: it works because the witnesses have mental imagery. It wouldn't work if I was the witness, there's no magic trick the sketch artist is doing.
Haha, I had exactly the same confusion with the police show/sketch artist thing!
But, reading up a bit on this topic, it seems that "normal" witnesses aren't very good at providing accurate images either. I guess the best examples (which are very impressive!) are a rare combination of brilliant police sketch artists eliciting skilfully from witnesses with an unusually accurate visual imagination.
In fact, this article (using facial composite software) (https://www.sciencedirect.com/science/article/pii/S0010945223001764#bib39) suggests that aphantasics might not be disadvantaged, maybe because they're less likely to generate false visual memories, and they're able to make sensible spatial judgements.
I hadn't thought of it like this before, actually - a strong visual imagination adds a bunch of noise/error that aphantasics don't experience!
Without moving your head, put your hand into a rear pocket, or put down an object behind you and then pick it up again. I am guessing that you can do this without using visual imagination. I think I am (as an aphantasic) using the same spatial proprioceptive capability to perform hypothetical tasks like the one you suggest.
What being aphantasic means, is thinking about spatial concepts without any of the visual associations. Here's an analogy: suppose you watch a video of someone playing a musical instrument, with the sound off. If you know the melody, you can probably watch them playing and guess where they are in the piece, and imagine what it sounds like .
Now suppose instead the instrument is a strange one to you. You have no idea what it sounds like. All you have is that by the timing you can work out where in the piece they are, and that in the abstract they are going up now and down now, but without the rich associations of timbre and colour that you have for an instrument you've actually heard.
Now, aphantasics aren't blind and do have memories of visual objects. They just don't "come along with" thinking about spatial problems. If I think about the guy walking in a T shape, I know that when he has turned 180 degrees he's going to pass the top of the T again and I can hold these spatial relationships in my mind, but they don't have other visual associations, such as colour or brightness or texture or shape. I know they don't, precisely because I do experience those in the real world and so can tell that they are not present.
For me, it's clear it's not about an error in communication because of how well my visual imagination works when I'm dreaming.
I sometimes can hold onto the ability when sort-of lucid dreaming, and I've even managed to consciously experiment with imagining an apple in this case. The apple is full colour; I can see details and texture.
Then I wake up fully and my ability to imagine an apple in my mind decreases considerably. No colour, no details, just a hint of apple.
Spatial reasoning isn’t all visual - as others mention, we have a lot of spatial sense associated with touch, proprioception, sound, and other senses too. I think I’m pretty normal at visual imagery (though bad at human faces - I can’t identify what hair color or eye color most of my friends have by imagining looking at them) but I’m quite bad at imagining smells or tastes that I’m not having right now. Still, when I’m reading a menu I know facts about which tastes I like, even if I can’t envision them the way I can envision images.
Oh, yeah, I have the same issue with imagining tastes! I must be able to do it *somehow,* to *some* extent—I know which things I want to taste again, and not only in a "I remember I liked that" way, but in a specific-craving kinda way (as you also allude to)—but if I try to close my eyes & summon up a taste, it's just... impossible. It *feels* as if I'm doing it, briefly, but it's like an (extra-)insubstantial ghost—the more I try to drill down & really "taste" the imagined item, the more I find that I am not imagining anything at all... if that makes any sense, heh.
>but probably I'm wrong about this, since it appears to be a well-validated(?) "condition".
The validation seems to be entirely self-reports:
>Before accepting the existence of such wide variability in the imagery of normal people, we should ask whether those with self-reported high and low imagery powers differ significantly in their success on cognitive tasks that are plausibly aided by the use of visual imagery. In this vein, James Angell (1910), discussing the imagery literature of the time, stresses the importance of looking for correlations between what he calls “objective methods” of measuring imagery, in which success or failure on a task depends on the nature of a subject’s imagery, and “subjective methods” of self-report.
>In the past hundred years, many researchers have, as Angell advised, compared subjective and objective measures of visual imagery. The results are discouraging. Research has focused mainly on three subjective measures, more readily quantifiable descendants of Galton’s original questionnaire: Betts’s (1909) Questionnaire upon Mental Imagery (and Sheehan’s [1967] shortened version of that questionnaire), Gordon’s (1949) Test of Visual Imagery Control, and Marks’s (1973) Vividness of Visual Imagery Questionnaire (VVIQ). The early returns were very bad. Through the 1970s, attempts to correlate these subjective measures with anything objective failed so regularly that most prominent reviewers denied the existence of a relationship between objective and subjective measures of imagery (e.g., Ernest 1977; J. Richardson 1980).
>More recent reviews have been slightly more optimistic. McKelvie finds strong relationships between the VVIQ and tests of hypnotic suggestibility, tests involving the Gestalt completion of incomplete figures (e.g., the speed at which someone would recognize a fragmentary stimulus as a “circle”), and tests of motor and physiological control; he finds spotty relationships between the VVIQ and tests of visual memory; and he finds no relationship between the VVIQ and tests of visual creativity (for people of normal I.Q.[7]) or tests of ability at spatial transformation or “mental rotation”.
As if it was indistinguishable from a T on a page? As in, if you didn't know the difference, you wouldn't be able to tell if you were looking at a picture or imagining it? I think that's quite a rare form of hallucination but i don't have a citation.
No, there's no way I could get confused, it's like I have a piece of paper in my mind and I can draw on it. It's weirdly more like remembering looking at an image than actually looking at an image, except that I can modify it.
I can well understand why there was such a debate about whether mental images exist or whether it's just a metaphor, although I'd definitely have been on the 'of course they do' side.
I think it must be vision-related, because if I need to draw something detailed it helps to close my eyes so there's no interference, but if it's something simple I don't need to.
For instance, in answer to Galton's famous question I can imagine a tiger and it's there in colour with a forest background, but if you ask me to count the stripes I can't, unless I deliberately imagine the stripes in detail first.
In this case it was actually a glowing red T on a black background. I can remember it from yesterday as if I'd actually seen that on a computer display. And given enough time it wouldn't surprise me if I got confused enough to not be able to remember whether I'd actually seen it or just imagined it.
I will say, as a phantastic person, this just makes me picture a location I'm walking on from my normal eyeline, and then when the question finally lands I have to throw all that out and restart with an overhead view.
I wonder how much variation there is in how people imagine this, "by default"—my instinctive/immediate "choice" of view was something like "a bit above, behind, and to the left of my head at the starting-point" (then the "camera" stays there & mental-me walks out as described, although IIRC I did sort of the same "path-following instead of fixed-camera" "error", at first).
I interviewed a man with aphantasia once. He solved the problems the way you would if you were writing a computer program to work out the answer or making a mathematical proof. He treated them as logic puzzles. In this example, he would have reasoned that there were only two lines at right angles, so it is either L or T, and then another step of logic (that the line that goes back on itself goes further than where it started from) rules out L and concludes T. No visualization, just reasoning about angles and distances.
I'm severely but not completely aphantasic. I was able to follow your question by drawing a very bleak and washed out thick line in my mind's eye, it took two tries because I needed to go back and calculate explicitly that the left and right arms are each twice shorter than the forwar arm. I don't think of aphantasia as inability to imagine literally anything superimposed over my visual field, it's more about extreme poverty of detail. I can visualize any letter of the alphabet, but not a word of 3 letters or more, except by progressively focusing on one- or two-letter chunks in it. I can see something in a very faded color, but not two colors at once. I can bring to mind a few faces of closest family members, but they're like snapshots I recognize but can't put my finger on any significant detail.
I'm only partly aphantasic, but I fully agree with the Torches Together response above. I have no problem imagining form and movement, it's just not quite visual, more prioceptal. There's absolutely no colour, no texture, no background, no accidental details that are not required to visualise the object of the imagination. I had no issue whatsoever answering your letter question, but in the process of doing so I realised that I made tiny movements with my head. I'm also pretty good at remembering maps, and it's not visual either.
Perhaps it would be useful if I mention that I believe the vast majority of my dreams are nonvisual. There's action, but the scene is not imagined visually. This is in contrast with audio stimuli -- when I wake up I often remember the last spoken sentence exactly, including the tone of the voice, modulation, etc. Only once recently did I wake up remembering an image (a face, and a patterned scarf), and was very surprised.
You already got quite a few replies, but here's another. I'm not totally aphantasic but quite severe. It is hard to explain the process to arrive to T but I would say it's more tactile, I may not have the technical knowledge to explain it. I have above average spatial imagination and solving that kind of puzzle I've always found great fun. I am also a visual artist, lots of artists are aphantasic. Sometimes I wonder if this is a way for us to create the visual imagery since we can't picture it in our mind. I've now picked up sculpture and I am blown away with the facility I have for it, and how enjoyable I find it. Line based drawing has always been harder for me but just building shapes in space is apparently second nature, who knew.
I'm also quite skeptical of people who claim high phantasia. With little practice they should become great artists (and some geniuses do but they're exceedingly rare), but they can't for their life put any of the supposed detail down even if clumsily. My guess would be some sort of fooling ourselves like we do with dreams. The details aren't actually there. Btw yes my dreams are also visually very lackluster, although not spatially.
Hypothetically, what kind of guarantees do those embryonic selection companies have? If I use one and 10+ years later my kid has an issue they claim she wouldn't get, or is not as smart as they claim she would be, then what sort of recompense do I get?
I mean, this isn't like fucking up a car or a boat, this is my child. I'm going to want more than money to make amends. Will these companies even be in business in 10+ years?
There is no guarantee because this is all probabilistic - they may say that there's a 10% lower chance your child will get problem X, but that still means 90% of the kids who would have gotten it anyway will still get it.
I get that it's probabilistic. Still, they're offering a service/product with many claims; there must be some sort of warranty or guarantee of satisfaction to some extent, especially if I'm paying upwards of $50k for this. Maybe they accept returns.
>there must be some sort of warranty or guarantee of satisfaction to some extent
*Must* there be? I can imagine a model like that for the claims on e.g. boxes of oatmeal: "reduces risk of cardiovascular problems!* [...] *studies show that eating a diet rich in whole grain [...]"—i.e., the claim rests upon some studies, and is offered with no guarantees.
Not exactly analogous products or claims, of course, so I might be wrong—maybe there *is* a "guaranteed effective or your money back!" somewhere on there.
If I can get a 5-year limited guarantee on a $15 box of Christmas lights, then surely I can get something analogous for a $50k embryo. Eh, maybe they'll be a class action suit 18+ years from now.
Plaintiff: Your company promised that my daughter would have an IQ high enough to get into MIT. She didn't get in even with perfect SAT/ACT/ETC scores.
Defendant: There are too many confounding factors to find fault with our methods. Not everyone with perfect scores gets into MIT. It's a multifactorial selection process. Besides, between the latest batches of 200+ IQ babies and our AI overlords who are considering closing such inefficient methods of learning and scientific discovery, we think that you should be pleased with the way your daughter turned out. Might we suggest she consider applying to "The Farm". We'd be happy to share information on this life alternative.
The Christmas lights are at a level of reliability where 99%+ of them are still working in 5 years. Embryo selection is at a level of reliability for predicting traits that you need advanced statistics to measure the difference.
They're pretty clear as to how their predictions work and the decent probability that they would do better. They're not measuring against a perfect super-baby, but against the average baby you would get from a randomly selected embryo, and even then there's a decent (although lower than random) chance the baby is worse than average.
They might give guarantees if the technology is good enough, which won't be for a long while.
On a serious note, why can't companies like this offer any sort of customer satisfaction promise? Sure it's not Christmas lights or a car, but it is a service/product for sale. It may be complicated to measure, but not impossible. I wonder if there's any analogous examples out there.
On a less serious note:
Maybe they could keep spares for replacements in the event that something goes wrong. IIRC the latest date for splitting off cells for identical twins is like two weeks. I'm not certain when the companies sample, but the possibility exists of them making and storing "copies". Also, with the eventual advent of artificial wombs this opens up a lot of "human trials" for better or worse.
Dissatisfied customer: I'm not happy with the way my son turned out! He's a shiftless layabout. I demand a refund.
Customer service: We are unable to offer a refund, but we have the embryo that is your son's identical twin brother. Would you like to try again and provide a more rigorous environment for the twin? [Insert joke about personality being genetically determined].
In most medical procedures the guarantee is that the doctor will perform the procedure, not that it will necessarily have an optimal outcome.
This would be like that. They would be liable if either (a) they gave you unrealistic expectations about the results of the procedure, or (b) they didn't do the procedure that they promised.
Well, AFAICT, the service/product they offer is a particular genetic sequence, which is easy enough to verify, and if it doesn't check out then I imagine there are penalties stipulated.
My best guess regarding Substrate is that it might be like a Ponzi scheme with Thiel and Trump not so much being suckered as being at the top of the pyramid helping to sucker everyone else. Once Thiel and Co. have invested and the government has endorsed it, it's a lot easier to draw in lower information investors.
This makes the most sense. If you can convince one investor that your product is good, and Thiel trusts that investor, and Trump trusts Thiel, then of course none of them will double check. It helps that they're all also very stupid people.
Ok. He won a California statewide mathematics competition in middle school, completed 2 degrees at Stanford, was editor-in-chief of the school paper, and has had a hyper-successful career as both an entrepreneur and an investor in which he's been repeatedly early to major technology sea-changes. Oh and he's a National Master in chess. If Peter Thiel is stupid, then either intelligence is irrelevant to real-world outcomes, or you don’t know what intelligence is.
He can be a successful investor, solve math puzzles like a little angel, win at chess, and still not see the forest from the trees.
Someone rambling about the antichrist while sucking up to Trump is not really defying the label.
Of course there's a lot of cultural and political aspect to evaluating someone's decision making. (It seems he is pretty good at his antisocial survival-of-the-fittest competitive strategy - he commands more than 26 billion dollars after all, but that still seems a bit stupid compared to prosocial cooperative strategies.)
Yeah that's called wisdom, not intelligence. Lots of high-IQ people are notorious for being a little spacey and nonpragmatic.
Weird beliefs are common among the intelligent. Lotsa professors buy into the nonsense that gender is nonbinary despite there clearly being only two human sex gametes, or that Marxism is a good idea. What you're really saying here is that you personally dislike his worldview. Which is fine but the word for that isn't 'stupid'.
And just so you know, he's not some Christian wingnut who thinks the antichrist is a real person. He understands religion in a way similar to Rene Girard and uses 'antichrist' as something which connotes something like a social attractor point - an abstract pattern in large-scale social dynamics, not unlike how Scott uses Moloch.
First of all “intelligence” is nonlinear so he can be smart in some areas and stupid in others, so it's already a fail to have a serious discussion about what “stupid” is. it's just a directionally true insult.
But more importantly if he believes what he publicly says he believes then he must be in some sense stupid, yes. I can't know what he believes in private. No matter how successful you are, if your values and goals are selfish and antisocial then you are stupid on a deep and fundamental level, arguably even more than an unsuccessful person who never hurt or took advantage of anyone.
>so he can be smart in some areas and stupid in others
That's mostly false. Psychometrics demonstrate that all analytic cognitive capacities are highly correlated, hence g. Sure some people are better at math or verbal, but to first order they're correlated.
>But more importantly if he believes what he publicly says he believes then he must be in some sense stupid, yes.
I don't know specifically what beliefs you're referring to here, but it sounds to me like you're conflating intelligence with wisdom. Those things definitely don't (necessarily) go together. Just look at someone like Noam Chomsky. Undeniably brilliant and also believes just absurdly buffoonish things. Also woke originated among the universities and professors have some of the highest rates of pro-Marxist thought. High IQ people are notorious for being able to talk themselves into shockingly wrong ideas. That doesn't make them "stupid". It means they should stick to academia and not try to influence real world politics.
What you're saying is that you disagree with Theil's politics and don't like him. That's not related to his intelligence. You're just using a childish insult to designate your outgroup.
I don't think Thiel is stupid *at all*, but he might not care about ripping people off if he calculates he'll still make money at the end or it fits one of his larger goals.
If he is anti-social and selfish then his values and goals are stupid even if his methods are not, but as I said in another comment it's not very productive to try and accurately define a basic insult.
Why would Peter Thiel run a ponzi scheme when he already has billions of dollars in legitimate wealth from founding and investing in successful companies? It's just not plausible.
Because it's also Trump's ponzi scheme. Some things are more important than money. It's good to be on good terms with the leader of the most powerful country in the world.
"Every time you take a Waymo, you’re letting an AI control a critical system; every time it chooses to stop at a red light but not a green one, it’s making an “important decision” (if you don’t think this decision is important, consider the consequences of failure)."
Author here. I don't know if we actually disagree--I wouldn't count Waymo decisions as important enough, because only small numbers of lives hang in the balance. What we said was:
"We all believe that current AIs should not be allowed to have autonomous control over critical systems. This includes extreme cases like giving AIs control over data centers, nuclear weapons, tech companies, or government decision making processes."
I can't speak for the AI as Normal Technology / AI Snake Oil authors, but we at AI Futures Project are supportive of self-driving cars and similar tech, and were mostly warning against the much more dangerous things given in the examples (e.g. data centers). Perhaps we should have clarified.
I understand that you in fact support Waymo. I just think Waymo support generalizes to allowing AIs to make many important decisions. What I have in mind here is things like the Colorado act which make it hard for companies or organizations to outsource business decisions to AI; I'm worried that as framed it seems to potentially provide support for that. I acknowledge that you specified some extremely bad things like nuclear weapons, and some things which you might have intended to limit to very major cases like government decisions, but I thought the exact phrasing:
> "We all believe that current AIs should not be allowed to have autonomous control over critical systems. This includes extreme cases like..."
...suggested a broader picture than those most extreme cases. Probably this was you guys trying to come up with some framing that covered both the AIANT's broad ban request and your narrow ban request, but I still disagree with my interpretation of the compromise.
I think we could have been clearer about what we mean by critical systems. I think probably (but I'm not sure) AIFP authors have a much higher bar for what should count as critical, than AIANT authors. And reasonable people can disagree about where the bar should be. Our statement definitely rules out the current stated plans of the major AGI companies though -- they are planning to have AIs autonomously running AI R&D programs (and indeed, Sama said he's aiming for autonomous AI company CEO iirc). So it's not a trivial statement.
Hmm. Ruling out "AIs autonomously running AI R&D programs" seems to be assuming the conclusion. If AI is Normal Tech then surely this use is not a critical system any more than PERT charts would be.
I mean, I'm mostly on your side here, but I think this is a sign that Point 7 is something you have managed to "agree on" by making it vague enough that nobody could disagree.
I'm not sure but I got the impression that the AIANT authors agreed it was a critical system -- after all they agreed to the wording quoted above, which lists datacenters and tech companies.
And again, some people do disagree--most notably, the leaders of the tech companies themselves disagree, and are explicitly planning to do the thing we're warning against.
“ I just think Waymo support generalizes to allowing AIs to make many important decisions.”
Scott, would you ever expand this statement to something like “using AI now generalizes to supporting AI development to proceed apace”?
Or to put it another way, “people who are serious about AI safety should stop using AI and advocate for abstention from the use of AI until alignment is achieved to our satisfaction”?
It also seems to me that there are relevant distinctions about *how much* control of critical systems is AI. A power grid that is moment to moment supervised by AI but with human checks at regular intervals and with human ability to look in whenever they want and with constant human evaluation of the systems sounds different from a system with full AI autonomy.
He doesn't seem to be all that wealthy. Besides, his job for almost 30 years has been politician. Very different than someone like Trump or Tom Steyer jumping into politics.
Also, warmth towards Newsome is only side of that poll result. JD Vance has lousy national like/dislike ratings in polls, and being VP is not a promising situation in which to improve that. Only one sitting VPOTUS has won a presidential election in the last 190 years, GW Bush in 1988.
Vance is young for a national politician and certainly could end up on a Nixon/Biden path towards winning the White House later. Or, Trump's physical/mental fade could terminate before 2028 and Vance would then be running as an incumbent POTUS. Right now though any matchup being polled on is "Democrat X versus the current VP", and being the current VP is almost never a path to national public affection.
Nixon had an inexhaustible reservoir of spite that propelled him from Eisenhower’s VP to becoming the most entertaining president ever. I’m not sure JD Vance has that kind of fire in the belly.
Yea that's a new one. Damn near all the people I hang out with deep in the heart of blue America love podcasts and view podcasters as celebrities. They happily pay ridiculous ticket prices to go to theaters and see people like the Pod Save America hosts chatter onstage.
There is a somewhat widespread belief that Joe Rogan swayed a lot of young male voters into voting for Trump. I know people personally who think this, though their counterfactual, voting for Harris, I believe is much less likely than them simply not voting at all. That the Democratic party fundamentally has nothing to offer young men, and in fact regularly demonizes them for the crime of existence, often seems entirely immaterial to some of them.
I think the issue isn’t “democrats hate podcasters” - it’s that when you ask random democrats to imagine “a podcaster running for congress” they’re going to picture some manosphere sports influencer or something, not one of the Pod Save America people.
Even if you imagine one of the Pod Save America people (about whom I know nothing) it's still pretty reasonable to say no, I don't want one of them.
My impression of most popular podcasters and youtubers is that they're mostly just a bunch of randoms who pass some basic microphone-skill filter and who got lucky with the algorithm at the right time and invested in making their shows a bit better. It's reasonable that people don't want to jump out of their seats to give these people even more power.
Re: 31, the proper counterfactual is "how popular would the Dems be on COVID issues if both the Dems and the GOP had fewer own goals during that time", and I think in that universe the Dems would be in the lead. (They had more own goals, it seems to me, in the sense of actions that didn't make any sense given their policy principles.)
> One million dead Americans is negligible. That's, what, 1 in 300? Know how many people committed suicide because of lost jobs, livelihoods, and businesses?
There are fewer than 50,000 suicides per year, and the suicide rate was _lower_ in 2020 than in years before and after. You do know you can just look things up, right?
One thing that's really hard to untangle is the way partisanship can have a profound effect on what steps the parties take. It's not hard to imagine Trump behaving differently in 2020 (i.e. really strong lockdown policies), and Democrats calling it tyranny.
In 2014, for example, Republicans excoriated Obama on Ebola and Democrats downplayed it.
This actually happened. When Trump cut off flights from china the Democrats screamed racism. When the COVID vaccine first came out and Trump was talking credit, the Democrats were the ones casting doubt on its safety.
No. You're remembering an oversimplification that became popular, not what the reports said.
FWIW, masks are relatively useless at preventing you from catching COVID, and relatively effective at preventing you from spreading it. I think this is true even for the masks just made out of one layer of gauze, but I've seen no reports of that being tested. They slow down the air on exhalations, causing the germs to remain close to the body. This, of course, only applies to masks that are worn properly, rather than, e.g., on one's chin, as I saw some people wearing them. And it depends on the air circulation in the environment. (Vertical air circulation makes it much more effective, horizontal less so.) And there are lots of other detailed caveats.
They said two things - there’s “no evidence” that masks help random people prevent disease, and there’s a mask shortage for medical professionals. I really hate the “no evidence” sort of line, because the evidence hasn’t changed all that much, but there’s plenty of good reason to think a mask will help a bit, and plenty of reason it won’t be anywhere near as effective as just interacting outside.
"When the COVID vaccine first came out and Trump was talking credit, the Democrats were the ones casting doubt on its safety."
I think that's wrong. *Before* the vaccine was released some Democrats cast doubts on the safety of a future vaccine whose approval might be politically driven. But when the actual vaccine came out, no Democrats doubted it because there were no indications that the approval was politically driven.
Yes, but that argument goes both ways. Before the election Trump had a huge incentive to push approval even if it wasn't safe. Afterwards his incentives were to decide on the merits.
The success of the first covid vaccine trial was announced by Pfizer 36 hours after Biden was declared the winner of the election by the Associated Press. So, by then the vaccine was irrelevant to the outcome of the election.
The timing is of interest because in late October 2020 Pfizer shut down processing of samples from its clinical trial until the day after the election. But few have ever shown much interest in that curious change of plans.
Trump didn’t cut off flights from China. Demand fell pretty dramatically during the pandemic, and the Trump Administration may have contributed to that with the State Department issuing a travel advisory and Trump instituting travel restrictions. Due to lack of demand (and possibly employee safety concerns) many airlines suspended service between the United States and China. So there were fewer flights. In particular, there were periods there were only flights to LAX or only flights to SFO. But fewer is not the same as none, which is what “cut off” implies.
> Section 1. Suspension and Limitation on Entry. The entry into the United States, as immigrants or nonimmigrants, of all aliens who were physically present within the People’s Republic of China, excluding the Special Administrative Regions of Hong Kong and Macau, during the 14-day period preceding their entry or attempted entry into the United States is hereby suspended and limited subject to section 2 of this proclamation.
You will notice that this excludes Hong Kong and Macau. It only applies to aliens (not U.S. citizens or permanent residents), and there are a number of categories of aliens that it doesn’t apply to. There were 279 direct flights from China to the United States between Feb. 2 (when the travel restrictions went into effect) and April 15.[1]
Rather than writing “may have contributed,” I probably should have indicated that it did contribute.
I don’t believe that any prominent Democrat claimed that the travel restrictions were racist.
"Lactose-intolerant people who drink milk regularly since childhood develop gut microbiota that can digest milk, but which demand an expensive “tax” in calories."
If this "tax" overshoots and burns more than the calories of the milk consumed then these microbiota sound like promising chemical-free slimming aids. There's probably a downside though, such as the calories being converted to methane and thereby giving rise to copious highly flammable trumps!
Lactose is a sugar so letting microbes digest it has an advantage now--the blood sugar will not increase by that much-- a nice advantage in the middle of worldwide diabetes epidemic.
Re #30, I'm mildly aphantasic, in that I can visualize things in my mind but only in a very vague and phantasmal way, but I have a very vivid "mind's ear," and can hear voices and music in my head almost as realistically as if I was actually hearing it. I wonder if there is a tradeoff between these things, or if they're positively correlated? I also wonder about the relation between internal monologue (of which I have a very strong one) and ability to visualize, do they stem from the same thing or is one the result of being a verbal reasoner and the other a result of being a visual reasoner?
FWIW, I seem to be near the rightward tail of all three distributions (extremely pervasive/consistent internal monologue; vivid mental imagery; can sometimes, albeit rarely, mistake an imagined sound for a real one).
As I mentioned in another comment I left above, I sometimes doubt that anyone *really* differs in any of these three—how can one even really be sentient without an internal monologue? how can anyone without mental imagery do things like mentally rotate a shape to "fit" it into another?!—but I guess that's probably "typical-minding".
There was a very interesting book I read as a kid named Ratha and Thistle-Chaser about a sentient species of cats. One of the members was raised outside of society and so didn't learn language, and of course lived life with no internal monologue. (She still had ideas, but in the form of pure concepts rather than tied to words.) Sometimes I'd try to copy that just to see if I could, and found that yes, I can manipulate thoughts without first turning them into words, but it's harder and I lose track more easily.
I find that I do *some* wordless thinking, too—particularly in mathematics: it's hard to explain, but it's something like "turning a mental object around & around until it fits into a larger mental object", if that makes sense (i.e. when it so fits, if it does, *that's* when I truly understand the concept illustrated by the example/problem).
Oddly, though, I'm trying to introspect *in re* how I understand formal logic, and it feels as if such cases are mostly verbal: thinking through them something like "if A is X, and B is Y, therefore it must be that C cannot be Z...", and so on.
I'm not very confident in these self-reports, though (which is another odd thing: how *can* one be mistaken / about one's own mentation? / it's a mystery, it seems to me— / much to my consternation! ...*cough*–).
For when I'm trying to do it on purpose, it helps to spend a couple minutes first focusing entirely on the input of my senses, the texture of the leaves of the potted plant in front of me, the roughness in the latter half of each tick of the wall clock, little details that aren't easily put into words. I also have to focus entirely on the present moment, since otherwise I would think about how I'll be explaining it in a few minutes, which takes words. Then I can try to think about different things and see what can come out non-verbally. For instance, I can think about getting a glass of milk from the fridge, and this comes in as sort of a picture of me in front of the fridge door + a general concept of milk. When I'm not trying to avoid words, that concept still comes in like that, but will be shortly followed or preceded by the internal monologue "I could go get a glass of milk". My brain seems to prefer replacing words with pictures as the first go-to, but if I try to avoid that, I can think about concepts using other connections to them, such as sounds or spacial relations. I suspect there's at least one type of meditation that works more or less the same way, but haven't felt a need to look into that.
Yep—at least, I seem to recall that it does, and experimenting with trying to think fast just now made it speed up (although the amount of extra speed I can squeeze out seems somewhat sharply limited 🤔).
TBH, I find "shape rotating" to be an extremely difficult thing. Verbal stuff? I've always been off the charts. But visio-spatial? Absolute crap.
On the three, I have a constant internal monologue/dialogue, active *verbal* imagination, but very poor visualization ability, and mistake imagined sounds for real on a regular, but not frequent, basis. Partly due to living alone.
I dream in visuals, but lose anything but a text description once I wake up.
I'd definitely say that I have more difficulty in mentally rotating shapes, or unpicking knots, or the like, than with the other two—and even my ability to visualize colors or textures is much stronger than my ability to keep complex shapes in mind & manipulate them etc.¹
How do you do on tasks like Raven's Progressive Matrices & similar "complete the sequence" puzzles, and/or those "analogy" problems ("A is to B as C is to... __")? IIRC (which may not be a safe assumption, heh: see footnote #1–), these are two of the most g-loaded question-types commonly encountered on IQ tests & so forth...
...but that's just an incidental Fun Fact!™ (𝘯𝘰𝘵𝘦: 𝘤𝘰𝘮𝘮𝘦𝘯𝘵𝘦𝘳 𝘱𝘳𝘰𝘷𝘪𝘥𝘦𝘴 𝘯𝘰 𝘨𝘶𝘢𝘳𝘢𝘯𝘵𝘦𝘦 𝘰𝘧 𝘢𝘤𝘵𝘶𝘢𝘭 𝘧𝘢𝘤𝘵𝘶𝘢𝘭 𝘴𝘵𝘢𝘵𝘶𝘴; 𝘶𝘴𝘦 𝘢𝘵 𝘰𝘸𝘯 𝘳𝘪𝘴𝘬, 𝘦𝘵𝘤.) Rather, I bring these up because the former, at least, seem sort of "visual" (moreso than they seem verbal, anyway; though, of course, the analogy thing is just the opposite)...
...and yet I've always been much stronger with RPM than with "which shape fits into this other shape if rotated 120° clockwise?" or whatever—I almost feel as if 𝘀𝗼𝗺𝗲𝘁𝗵𝗶𝗻𝗴 about, some underlying cognitive mechanic in, a "continue the pattern" puzzle is similar to the same in (say) "find[ing] the right word", vel sim.²—so I wonder if this also applies to other "verbal-dominant" folks too.
-------------------------
¹(Interesting that you should mention this—I thought I'd included something similar in my initial comment, but I guess not; great minds just think alike! Wordcel power! 👊 Actually, I seem to recall that high verbal ability is more g-loaded than the other sort(s); semi-but-not-really–relatedly, I read somewhere that so-called "crystallized" intelligence is also a better predictor of 𝘨—or maybe it was just total IQ?—than is "fluid" intelligence... which last seems hard to make sense of, tbh, so I may be misremembering; call it maybe 65/35 in favor, 𝘪𝘯 𝘳𝘦 my confidence that I accurately recall a real finding, heh.)
²(or else this particular supposition is just my "wordcel cope" coming to the fore–)
I'm way better with analogies than any kind of the RPM or "find the next element in the sequence" things. Multiple choice tests are something I'm good at, as long as there are answers I can exclude via logic.
I figure I'm really good at manipulating symbols and interpolating from known facts. Geometry sucked. Calculus was fine. Physics without numbers is much better than physics with numbers. Word problems are my jam.
Couple that with an obsessive reading habit leading to a large fact base, I can seem really smart. By no means genius level. I'm a generalist though.
>can sometimes, albeit rarely, mistake an imagined sound for a real one
I have good mental visualization, but very good auditory 'visualization'. Not sure the right words here. I can reliably recreate entire songs, beginning to end, all the instruments separated, in my head. I've also had many years of formal musical training, which has to be a big part of it. I've done a fair amount of transcription too.
I frequently "mistake an imagined sound for a real one", or what my psychaitrist calls persistent auditory hallucinations. I don't remember the last full day that passed where I didn't have one; its probably been 15 years. I have to think the two are related. Novel hallucinations are pretty rare now, they all tend to be of a certain narrow type. Music lightly playing in the other room, like someone has left a radio or television on but the audio turned down, is pretty common. Once a day or so, more if I'm stressed. There have been very stressful periods of my life where the 'distant music' hallucination persists for multiple days without stopping. Hearing my name being called is also fairly common, as are soft and distant conversations coming from the other room like the music. The sound of the alarm clock I had for most of my childhood is also fairly common. I'd estimate I know right away 90% of the time that its a hallucination, and eventually determine almost all of them. There are maybe 5-10 sounds in my life that I'm still not sure if its real of not. Some of my original compositions have come from my hallucinations, which sounds like a superpower to musicians I tell about it, but its not really. At best its a snatch of a melody or a couple bars of chord progression, or even a single chord.
Anyway I've learned to largely ignore them. I think if they were clearly spoken language hallucinations it would be a lot worse. I've only had those in periods of extreme duress.
Interesting. I think verbally and "hear" my internal monologue but it's only ever my own voice. I can't imagine other people's voices and I can't even imagine sounds that aren't speech. If, for example, I think of the sound that hitting one piece of metal against another makes, I just get my own voice saying "Clang!"
As for visual reasoning: I can do geometry in my mind and I thought that was what people meant by visualizing things, but apparently it's not.
For another point of anecdata, I'm aphantasic (can't conjure visual imagery in my head), but I think I have an approximately average "mind's ear" and internal monologue.
Not sure. My imagination can be quite vivid in general - I can not only imagine detailed scenes in my head, but I can also imagine sound, or even touch, smell, and taste (although being able to vividly imagine eating ice cream hasn't stopped me from seeking out the real thing).
Just to point out the spelling mistake, please tell me when you correct it so that I delete my comment:
8: New revelations on the OpenAI coup from the Musk vs. Altman lawsuit. The effort to remove Altman may have been **let** (obviously it's led) by Mira Murati and Ilya Sutskever...
53: Huh, I didn't know that! Stoll isn't just a "scientist"; he has substantial Silicon Valley cred by way of his 1989 book The Cuckoo's Egg. You can be arbitrarily close to the technology and still get predictions about its future course very wrong.
I read that book as a kid. He notes that his fellow astronomers thought he was a terrible astronomer, but a whiz with computers. When he switched to working with computers his colleagues thought he was lousy at that, but a genius astronomer.
I think its also likely that Stoll was exposed to more computer-hype than most, given when he was most active. The 70s and 80s were full of tech promises that were just around the corner. Academia had been on the early internet for years by 1995. Admittedly, even the most enthusistic tech promoters of the 90s and 00s missed how popular smartphones would be with the normies, especially women. I remember an anecdote that Apple didn't expect the Iphone to be all that popular with women and thought their Iphone buyers would be the same group as their core-demo computer buyers.
As for the bird names, when I was a child, it was “Willie Wagtail” and not just (pied) wagtail.
6: Re: the anti-mosquito drones, they undoubtedly do look very cool. They seem to work by colliding with the mosquitoes (and hence chopping them asunder), though, and someone is going to raise a fuss over cruelty and inhumane killing methods.
10: St. Carlo Acutis (1991 - 2006) is the unofficial patron saint of the Internet and “first millennial saint”. …(H)is sainthood is something of a mystery, with Wikipedia saying that “even those with a deep devotion to him struggle to pinpoint his specific actions that led to his canonisation”, and an Economist article admitting that “nothing in his sparse life story explains that this ordinary-seeming teenage boy is about to become the first great saint of the 21st century”.
First, there are the two miracles necessary for canonisation:
“On 14 November 2019, the Vatican's Medical Council of the Congregation for the Causes of Saints expressed a positive opinion about a miracle in Brazil attributed to Acutis's intercession. In 2020 the Catholic Church recognised the curing of a child's pancreatic disease as a miracle attributed to Acutis's intercession. On the death anniversary of Acutis, Luciana Vianna had taken to Mass her son, Mattheus, who had the congenital defect of an annular pancreas which made eating difficult. Beforehand, she had prayed a novena asking for Acutis's intercession. During the prayer service following Mass, Mattheus kissed the clothing relic of Acutis and asked he should not "throw up as much". Immediately following the Mass, he told his mother that he felt healed and asked for solid foods when they arrived home. Until then, he had been on an all-liquid diet. His doctors subsequently confirmed a normal appearing pancreas by ultrasound. …After a detailed investigation, Pope Francis confirmed the miracle's authenticity in a decree on 21 February 2020, leading to Acutis's beatification.
…On 23 May 2024, Pope Francis recognised a second miracle attributed to the intercession of Acutis. The miracle attributed to his intercession occurred in 2022, when a Costa Rican woman named Valeria Valverde had fallen off her bike in Florence, Italy, and suffered a brain haemorrhage with doctors giving her a low chance of survival. Her mother, Lilliana, prayed for the intercession of Acutis and visited his tomb. The same day, Valverde began to breathe independently again and was able to walk the next day with all evidence of the haemorrhage having disappeared.”
Second, you generally need to look into who is pushing the case for canonisation. Either there’s a local cultus, or some interested party/parties are involved. In this case, it seems to have been the local parish:
“The call for Acutis to be beatified began soon after his death. The Church in Assisi was active in promoting Acutis's cult, the Catholic practice of venerating or showing devotion to a holy individual. Nicola Gori, who authored Carlo Acutis: The First Millennial Saint, served as the postulator for Acutis's beatification process. In this dual role, Gori was responsible for advancing the cause within the Vatican and compiling the necessary documentation.
On 12 October 2012, the sixth anniversary of his death, the Archdiocese of Milan opened his cause for canonisation.”
19: New research analyzes Hitler’s DNA.
Yeah. How sure are we that this genuinely is Hitler’s DNA and that it’s not degraded? Where did they get the sample? Off a sofa?
“The testing was made possible after researchers obtained a sample of Hitler's blood from a piece of material taken from the sofa on which he shot himself.”
I think this is more in the same wheelhouse as all those Discovery channel “Tomb of Jesus/Mary Magdalene/Your Great-Uncle Theobald discovered in site five miles outside Jerusalem” ‘documentaries’ than genuine scientific research. Publicity for the documentary which is yet another in the never-ending stream of “the public can’t get enough about the Nazis” productions.
32: “These are my people? My Volk? Am I supposed to sacrifice life and limb for the salus of this populus? Fuck that. Let them cook, they deserve everything that’s coming to them”
The answer is to put that devotion into Salus Populi Romani 😁
43: Sam Bankman-Fried’s mother on why she thinks his trial was unfair.
Oh, Sam’s mom thinks he’s didn’t get a fair shake? What a surprise. Also not surprising since she and her pet causes got a ton of money off SBF’s shenanigans, as well as her and Dad getting some nice perks:
From the CTFC filing:
“54. Contrary to such representations and without disclosure to FTX customers, Alameda and FTX comingled funds and freely used FTX customer funds as if they were their own, including as capital to deploy in their own trading and investment activities. On information and belief, Bankman-Fried, his parents, and other FTX and Alameda employees used FTX customer funds for a variety of personal expenditures, including luxury real estate purchases, private jets, documented and undocumented personal loans, and personal political donations.”
From the SEC filing:
“68. Bankman-Fried also used commingled funds from Alameda to make large political donations and to purchase tens of millions of dollars in Bahamian real estate for himself, his parents, and other FTX executives.”
“The majority of FTX’s purchases were high-end beachfront properties, including seven condominiums in the pricey resort neighborhood of Albany that cost about $72 million. According to the documents, a unit of FTX purchased these houses to serve as “residence for key employees” inside the organization. Reuters was unable to identify the residents of the units.
Bankman-parents, Fried’s Stanford University law professors Joseph Bankman and Barbara Fried, are listed as signatories on the papers for another beachfront home in Old Fort Bay, a gated neighborhood that formerly housed a British colonial fort constructed in the 1700s to stave against pirates. One of the paperwork dated June 15 stated that the property is intended to be used as a “vacation residence.”
When questioned by Reuters about why the pair opted to purchase a holiday home in the Bahamas and how it was paid for, a representative for the academics merely indicated that Bankman and Fried had been attempting to return the property to FTX.”
45: The “100 Above The Park” building in St Louis
Huh. Modern architecture that is surprisingly not terrible. I was expecting to hate it, but I like it. Looks vaguely Art Deco inspired, almost like a bouquet rising out of ground level.
I have to doubt the Hitler claim, too: there's an endless market for such things, and this nicely flatters the dominant narrative ("guess what: insane & micropenis, lmao! drug-addict too! also he loved eating feces & fucking his dogs, we think!"). Not that it can't be true, mind, or that I'm a fan of the fellow—it's just awfully convenient... one can imagine the team(?) involved thinking "okay, now to make sure this'll get tons of press!" and/or "ha! Kallman—THIS'LL show those alt-right neo-Nazis!", and so forth.
Yeah, it sounds like a combination of small independent production company doing content for the channel ("Channel 4 operates a unique "publisher-broadcaster" model, which means it primarily commissions and outsources all of its content from independent production companies, rather than producing programmes in-house") which means it wants 'guaranteed audience grabber' and Hitler is one of the guaranteed audience grabbers, plus some company with a swizzy new DNA technique they wanted to monetise. They could go the boring old 'publish papers in reputable journals' route, or they could go for WE SEQUENCED HITLER'S DNA NO HE WASN'T JEWISH YES HE WAS A BIG POOPY-HEAD SCIENCE SAYS SO PLUS HE HAD A SMALL TINY BARELY-THERE DICK.
I think they preferred the second option 😁 I mean, with "you couldn't ask for more rigorous sample management and custody with no chance of cross-contamination than this" techniques on display, why are we being so pettifogging?
"The bloodied swatch of fabric - now 80 years old - was cut out of the sofa in Hitler's underground bunker, where he killed himself when Allied forces descended on Berlin at the end of World War Two.
While inspecting the bunker, Colonel Roswell P Rosengren of the US army saw an opportunity to get a unique war trophy and he pocketed the fabric. It's now framed and on display at the Gettysburg Museum of History in the US."
'Cut it out of the very sofa after everyone and his dog had pawed all over the bunker, shoved it in me pocket, flew home thousands of miles with it, bish-bosh, job done!'
I do like how all the academics involved were shocked, shocked I tell you! that the TV station slapped a clickbait title on the two-parter and went hard on the sensationalism angle 😊 Whoever could have seen *that* coming?
Maybe it wasn't a new company, their expert Dr. Turi King seems to be the biology equivalent of Neil deGrasse Tyson (ready to go on telly at the drop of a hat to tell the public all about SCIENCE!):
King has appeared in numerous television and radio documentaries, programmes and videos as well as advising on television and radio productions such as BBC's Who Do You Think You Are?
Presented the series Ancient Murders Unearthed for Sky History.
Presented the BBC Radio 4 documentary: Genetics and the longer arm of the law
Guest on The Life Scientific as DNA Detective Turi King
Britain's Lost Battlefields with Rob Bell
Richard III: The King in the Car Park
Britain's Secret Treasures
Richard III: Solving a 500 Year Old Cold Case (TEDx Leicester)
Richard III – The DNA Analysis & Conclusion (University of Leicester)
Richard III: The Resolution of A 500-Year-Old Cold Case (Irving K. Barber Learning Centre Lecture, UBC)
Head Number 7 for Wondery
DNA Family Secrets
DNA Family Secrets is a television series which began airing on BBC Two in March 2021, presented by Stacey Dooley and King, it uses current DNA technology to solve family mysteries around ancestry, missing relatives and genetic disease. Series three aired in 2023."
>I mean, with "you couldn't ask for more rigorous sample management and custody with no chance of cross-contamination than this" techniques on display, why are we being so pettifogging?<
lol! Good point, very good point; I guess we are forced to accept this airtight scientific breakthrough after all. ADOLF "POOPY-HEAD TINY-PP" HITLER: CONFIRMED!
Yeah, it would have been terrible if it turned out to be Colonel Rosengren's DNA, wouldn't it?
Luckily, we know there is no chance of that because:
"The scientists are confident it really is Hitler's blood, because they were able to perfectly match the Y-chromosome with a DNA sample from a male relative that had been collected a decade prior."
A decade old sample, stored we don't know how, from a 'male relative' whose ancestry we're not informed about. Well, with fact-checking like that, whew, crisis averted!
Aha! Well, then... consider the doubters humiliated!—
(Assuming I understand correctly—not always a sure bet—then that very likely means only that the two Y-DNA samples shared a common patrilineal ancestor at some point within the last 10–20 generations, or something like that... although I could be wrong—maybe it is in principle possible to be more definitive—but given the context... I'm willing to bet [up to $5, or even $10!] that it's something like that, and not a definitive result—esp. given that it looks like the putative living relative's sample was identified with some rather questionable methods: collecting discarded cigarette butts & napkins, etc., from supposed likely relatives identified by a Belgian journalist through tracing "archival documents" to other branches of the family in Austria. Seems rather circular, in a way: "well, the sample has this somewhat unusual haplogroup, which matches a living relative's; we know Hitler had that haplogroup, because these living relatives have it; we know they're relatives, because they have this unusual haplogroup..." The case stands or falls with the archival research, it seems, but that's been kept secret, so—depends upon how much you trust that particular Belgian journalist, perhaps? Maybe others checked out his work, I don't know, but as it's not public it's sort of hard to be too confident in it!)
The obvious way to find out is to implant the supposed-Hitler DNA into an embryo and see if it grows up to look like Hitler. You'd probably want to do it, like, a thousand times, in order to get a sufficient sample size.
I think something like that is essentially a requirement for fascism- I think it’s fundamentally a much more sexual thing than usually appreciated. If you look at the current fascist movement in the USA, recruitment is largely based on convincing young men that they are somehow sexually inadequate, and that their movement will fix this by teaching them how to perform faux hyper-masculinity properly and teach them not to be “cucks” anymore. This usually works by exploiting a pre-existing terror that they are sexually inadequate or feminine somehow, which often stems from something about their personality or body that they are afraid or ashamed of. Once I noticed this, I started so realize that the references to sexual shame and terror are pervasive, even in executive orders and press communication from the Trump administration. A critical part of fascist propaganda is the leaders always being surrounded by attractive women, and the opponents being framed as sexually unattractive and unsuccessful.
That's part of the problem. It well may be that the relatives refused to get involved because they're fed-up with being hauled out every ten minutes as "Hitler's relatives", but you would wonder why they couldn't get a more recent blood sample that definitely is identified as coming from the family.
It was the genetic sequence for a particular disease that is known to sometimes have that kind result. Yes, we do have that kind of knowledge. But was it accurate is another question.
> 6: Re: the anti-mosquito drones, they undoubtedly do look very cool. They seem to work by colliding with the mosquitoes (and hence chopping them asunder), though, and someone is going to raise a fuss over cruelty and inhumane killing methods.
Thankfully, the same techniques work for anti-activist drones as well!
> Although some medieval street names such as Addle Street (stinking urine, or other liquid filth; mire[20]) and Fetter Lane (once Fewterer, meaning "idle and disorderly person") have survived, others have been changed in deference to contemporary attitudes. Sherborne Lane in London was in 1272–73 known as Shitteborwelane, later Shite-burn lane and Shite-buruelane (possibly due to nearby cesspits).[21][22] Pissing Alley, one of several identically named streets whose names survived the Great Fire of London,[23] was called Little Friday Street in 1848, before being absorbed into Cannon Street in 1853–54.[24] Petticoat Lane, the meaning of which is sometimes misinterpreted as related to prostitution, was in 1830 renamed as Middlesex Street, following complaints about the street being named after an item of underwear.[25]
Unfortunately, Cliff Stoll's 2010 comment no longer appears to be on the internet. It's quoted in the Wikipedia article, which cites a Boing Boing post, but the comments for that are gone. If you use the Internet Archive for that url, it will just try to load up Disqus and hang. Nor does any post that old appear to be in Boing Boing's read-only BBS that the "Comments" link goes to now.
Thanks very much for the featuring and the grant Scott!! On the choice of the White House: we did this video for an internal YC presentation that wasn’t supposed to be public, so the choice was a combination of both 1) fun 2) there was a cheap and very good quality model on cgtrader. Later our contractor got too late for YC timelines, so we went with that video. Can’t complain about the additional buzz! We’ve been able to have the luxury of the choice of investors, and a significant chunk of our cap table is EA-aligned, so we have no plans of getting into defense. The expansion market we found to make VC economics work is now invasive flying species control for agriculture, eg killing codling moths who lay eggs in apple plantations with microdrones rather than insecticide. (Also had the advantage of being a much easier target, both in terms of sonar RCS and max acceleration). Exciting times ahead!
Seems like cool technology and I can definitely believe that it works at having a device which can go kill mosquitoes. But I was surprised to see that there was no outcome-focused analysis anywhere I could find on the website. It says that it can patrol 5 acres, but what does that mean? What reduction in mosquito presence does it achieve compared to e.g. UV-light-based traps?
I mean, I pre-ordered one, how could you NOT? I've lamented the overweening lack of mosquito-killing drones too many times to not be obligated to buy the first one I see.
But I'm really interested in coverage, outdoor / weather use specs (can you use them in wind? What if it rains, are they smart enough to go back home?) and more. I assume there will be software that lets you manage them at fleet scale? The killer app to my mind is having a fleet for your property killing all mosquitos that come anywhere near your property without you having to think much about it.
Super interesting product but one of the reasons I avoid mosquito killing tech is off-target effects. How do you avoid killing other small flying insect species? Is it able to tell the difference? Either way, good luck!
The scale matters and they're proposing a solution to kill all mosquitos in an area: if they're also killing all flying insects in that area then I don't want it. If the bird and bat populations kill a fraction of flying insects that keeps their population at steady state, then I don't care. Indeed, if you like those birds and bats then you better hope you don't kill all the flying insects.
I'm sure there's a false positive rate but it seems pretty plausible that the drone could specifically locate mosquitoes based on their distinctive whine.
TL;DW theoretically, microdoppler signatures from the sonar.
I don't want to be a downer but this is a VERY difficult problem to solve, and something I have been casually researching for a while and have some industry experience. I've seen a few different ideas crash and burn but I think he is on the right track with sonar. Mosquitos are so small I'd be extremely impressed if he had an actual demo of a tracking algorithm working outside
Re 53: You might better know Clifford Stoll today as the funny old guy who frequently appeared on Numberphile talking about Klein bottles and other topological objects, and has a small business selling glass Klein bottles.
He gave an interesting TED Talk, and his book The Cuckoo's Egg was good (I have not tried the cookie recipe inside yet, though, having my own family recipe).
With #39, important to note that is relative to the shift in Natural Sciences, of shifting 0.18 left on a five point self-reported left-right scale. So eyeballing, everything but accounting and finance moves people left overall, but how much depends on the subject.
Good point! Their second figure seems to show the absolute data and it’s a very different picture. Interesting is theology being the most right at the start but having possibly the largest leftward movement.
My tastes seem to be similar to his, and I'd say that this is a better example of post/modernism - doesn't look to be intentionally boring or ugly or absurd, which is a low bar, but still rarely cleared these days.
It actually does look intentionally absurd to me, which is part of the charm! Why have all those weird twisted corners instead of a normal wall? Just because!
Well, it's a matter of degree. The dictionary says that absurd means "wildly unreasonable", whereas here I'd say it's mildly unreasonable. I mean, it's no Lou Ruvo Center or whatever.
Did anyone do the Lighthaven blogging thing and how did it go? I ask because I noticed Ozy posted about every day in November and that seemed like the blogging thing.
I had no idea #2 went beyond Robin Redbreast. That's pretty fun, and I wish it had survived more. We did our own version of it, for rats, when they left the closed restaurants during the pandemic: when we would see a rat scurrying around outside our apartment, that was Pat. Pat the Rat. They were all Pat the Rat.
29. I am now calling for a complete shutdown on naive utilitarianism until we can figure out what the hell is going on. We need a Manhattan Project for non-Benthamian consequentialism.
>Most interesting section is the one about children: he wants to have them, but doesn’t think they should be genetically related to him.<
Frankly, this is utterly alien to me. I have a hard time understanding why anyone would want children in the first place—when a friend announced that he (well, his wife) was going to have a baby, it took all of the acting talent within my grasp to feign the appropriate reaction ("congratulations! wow, how exciting! wooo!")—but the *sole almost-comprehensible reason* with which I can identify is "I want my line to continue / I want to see myself reflected in my kid".¹ If you don't have that, it's merely a lot of effort & worry & cost & time & sacrifice for nothing at all...
...except, I suppose, "someone to take care of me when I'm old"; I guess that still obtains.
----------------------------------------
¹I know, I know... something is probably wrong with me. Maybe I *am* an alien. I have a similar problem with babies in general, too: they're irritating & gross, not cute. Quickest way to get me to turn any film or show off, or leave any gathering, is to introduce some goddamn baby.
Most parents find taking care of their kids to be fulfilling in a way that makes up for all the effort and stress. You can verify this by going on any parent's blog and reading the paragraphs of "my kid is so cute" gushing. I don't think it's that complicated if you generalize the issue.
Why do people bother with romantic relationships when they could just have casual sex? Why do people want dogs? Humans find connection fulfilling.
I maintain that: well, of course—they *have to* convince themselves of said fulfillment; elsewise, they'd just break down & weep uncontrollably over the giant chunks of their lives they have foolishly thrown down the "baby-drain"–
-------------------------
(More seriously, I think the difference between romantic relationships & parenthood, or dog-adoption & the same, is not categorical, but rather one of degree *in re* RoI... although I'll note that I'm also beginning to feel the same way about romantic relationships: tried it; had some good parts; but, overall, too much goddamn trouble, too much giving up my own preferences to satisfy someone else's...
(It's not particularly difficult to *explain* why people *do* it—for a certain sense of "explain"—at least: of course people want to propagate the species; evolution ought to ensure this, in some sort of way & probably several... it just apparently went wrong in me, heh, or perhaps I have an unusually u̶n̶f̶o̶r̶t̶u̶n̶a̶t̶e̶ high self-indulgence::human-connection ratio.
(...OR! it could be—this is my favored explanation—that people are just easily fooled: by themselves, society, sappy movies, etc.; they are, on the whole, simply unable to truly picture what it's really like, and only imagine happy Hallmark images of fun trips to the park or whatnot. I.e., I merely possess an unusual portion of clarity & self-knowledge—*everyone* feels this way, if only they knew it. That one makes a lot of sense, if you ask me–)
Of course evolution programmed people to find raising kids fulfilling, but I don't think you even need to invoke special evolutionary drives to explain why most parents are happy with their decision to have kids. I think you merely overestimate how much the average person prefers freedom/independence over stability/community ties, because you value freedom more than the average person.
You keep framing kids as being terrible because it's a huge and essentially irrevocable commitment. But most people don't hate the idea of having responsibilities they can't get out of nearly as much as you do. Many even take pride in it. The other side of responsibility is purpose.
...hmm. That's a good point—I guess if "expending effort" & "less freedom to indulge as you wish in what you please" aren't awful prospects to someone, they'll weigh less heavily in the calculus.¹
Thinking about it, I actually do know several such people; in truth, I would probably have a much better life, myself, if only I understood—felt things from—the same perspective.
(I've tried, but I'm beginning to wonder if maybe I oughtn't just give up: scrabbling uselessly upon the slope of self-discipline gets discouraging, by the twentieth or thirtieth time you've slid back down... heh.)
Would you prefer a life of living in your parents basement and playing videogames (amateur level), or going into the world and accomplishing things? Travel, helping others, discovering things, building things, anything.
Why?
Kids give our lives meaning, which in this day and age, is teh scarcest resource of all.
As someone who is similar to you on this axis and happy with it, I don't think you need to change yourself to be happy. Having responsibilities is important for fulfillment for almost everyone, but these needn't be "inescapable" responsibilities like kids or (to a lesser extent) marriage are.
Get a hobby where you make cool stuff, or a career that contributes to the world, or any other project. Anything from "head of multinational organization" to "runs local DnD group" can contribute to a sense of purpose, while being something you're socially and legally enabled to quit once you're tired of it.
I think this is just autism, and shouldn't be held against utilitarianism specifically (although we should always be vigilant for dysfunctions caused by the large overlap in the two populations).
Has anyone written an explainer about whether the pardon of Juan Orlando Hernandez has anything to do with Prospera? I've seen people mention it in passing but not a deep dive.
#12: I have recently seen a talk about exactly that topic of switching out different AIs on the same chip. This is not just a problem with outdated models, it is also a problem in expert systems where you have dozens or hundreds of specialized expert AIs, and for each prompt you decide which of them you ask. (Possibly several.) Those systems already run in the background of modern AIs.
If the expert AIs are not used with the same load all the time, then you frequently need to load/unload some of them into GPUs, which is a very slow operation. It is also not a good option to just have all of them sit in some GPU being idle most of the time, just to be able to serve peak demand.
The solution in the talk was that many of those expert AIs are pretty similar, and the difference can be very much compressed. There are technical problems to be solved -- even if you can compress the diff into a small block, it would still be slow to decompress it and then change all weights a little. But the problems can be solved, and this is what the talk was about.
For the US, you have cities above the national-average line and states below. For Germany, you have three city names above the line and the name of one Bundesland below. Of course you can have all kinds of data dots in this figure, but then shouldn't the removal of the richest region be called the removal of the richest city? The blue and red lines would certainly be further apart if you removed Bavaria instead of Munich.
I suppose it's not terribly surprising that a neoreactionary is repeating self serving fascist propaganda, but in case anyone is wondering, Mussolini did not make the trains run on time and the Nazis did not 'fix' the Weimar economy.
I don't know about "fixing the Weimar economy", but when I grew up in the 1950's a common comment on Mussolini was "at least he made the trains run on time". So it may not have been true (I've never checked), but it was certainly widely accepted.
I'm aware of the phrase's meme status. Unfortunately it's just not true, or at least, very misleading. A very brief overview is available on wikipedia:
The argument presented in the excerpt is "Actually fascism isn't so bad, at least it's efficient/effective!" This argument does not follow if fascists weren't actually efficient or effective. In reality they were neither, but cared very much about the appearance of being so, which has contributed to the meme's sticking power.
My apologies to you personally for my tone, but even the third hand repetition of these falsehoods here is offensive to me.
Well, in the contexts that I heard it, the implication definitely wasn't "Actually fascism isn't so bad", just that it had a (few?) feature(s) that weren't awful.
It is important to point this out, because even the weaker implication suggests a trade off: "With fascism, you get less of X and Y, but you get more of Z." There may be a situation where Z gets so important that people consider desperate options.
Actually, whether that's true or not depends on the precise definition of fascism that you are using. I rather think that Mussolini would consider the US a fascist state, at least if his explanation of the symbol (sticks bound in a bundle around an axe) is taken as what he meant. (The sticks were the companies and the axe was the government.)
It's a piece of propaganda for which people lost the original context. It comes from the fact that during the Biennio Rosso (the "Red Two Years", ie 1919-1921, when industrialization and what the working class perceived as broken promises after WW1 led to social unrest and strikes), the railway workers were heavily unionized, and oft striked either for better contracts or in solidarity with other workers.
Now, the Fascists got their jumpstart in politics by acting as the hired muscle of the factory and land owners. This had put them to an obvious crash course with unions, even the ones not particularly affected by Fascist raids, like the railway unions.
After Fascists seized power, one of their first actions was to enlist the entire state apparatus in labor repression, outlaw unions, and privatize state companies with strong unionization (chiefly the railways). This became one of the central points of their early propaganda.
So the "made trains run on time" was not meant as "we made the railway administration more efficient", it was "we squashed the labor movement so thoroughly that even its hardest kernel would not dare striking, and a gentleman's trip will not be interrupted by these insouciant screechers any more". That was 100% true. Except that as the regime dragged on, it tried to broaden its base beyond the upper and middle-upper classes, and after 1929, mass unemployment made their previous political formula completely unviable. So there was an hard turn toward more left-wing rhetoric, many companies were re-nationalized, some labor protections and welfare provisions were re-instated, and the previous propaganda sounded so incongrous with the new public image of the regime that its original meaning was lost and got re-written in collective memory as some sort of silly, random boast like many others.
That little rant seemed weird to me as supposed evidence that some fascist has stopped being fascist - he still sounds pretty fascist but just lost faith in the people.
Fascists are hard to tell apart from general haters. Whenever I find one on the internet, in addition to all the stereotypical fascist hatreds he also hates the poor, the wealthy, women, feminized men, weight lifters (unless he is one), professional sports players, criminals, cops, smart people, stupid people….
Color me unsurprised that this one (former?) fascist hates his Volk. The step from fascist hater to non-fascist hater won’t make him sound like less of a fascist.
Yeah it's a bit different for Nazis and Falangists, but for the OG Fascists a certain Nietzschean will to power, the superiority of the exceptional individual over the dull masses, life made art and art made life, exploration of the possibilities for individual greatness offered by technical modernity but rejection of any democratizing tendency (even and maybe especially the ones created by the new industrial economy) was as big if not a bigger factor than nationalism or racism.
Marinetti, d'Annunzio, Balbo etc mostly stumbled into nationalism bc the Great War offered them a chance to play with all the new toys, live a life of danger and heroism, turn themselves into modern legends and whatnot, and at that point they might as well add some feelings for the country they were fighting for. But their supposed love for Italy was only ever instrumental to their quest to become ubermenschen, not the other way around (and absolutely did not prevent them to throroughly rape Italian society until it gave them the power and worship they thought they were owed).
Yes, it appears he still thinks fascism is the best thing for the people. Only now he hates the people, so he wants fascism to fail so that the people don't get its benefits.
> “Fiat currency isn’t backed by nothing, as commonly supposed, but by land.”
Fiat currency is backed by the issuing state's promise and requirement that it is legal tender within its jurisdiction. A currency's backing is thus the entire state's existence and power to enforce its laws, or that law in particular. Land is certainly correlated with that, but not the foundation.
Yes, because in the long term, all *states* go to zero (economically or otherwise), taking their respective currencies with them. That is not a distinguishing features of fiat currencies. At most, fiat currencies enable new indicators of impending disaster, such as hyperinflation, but they're not the disaster itself.
That assumes that the state keeps the currency going until the state no longer can, which is the typically the case. However, when the Euro was adopted, older currencies were eliminated without those currencies going to zero, so the claim that *all* fiat currencies go to zero has been falsified.
Well, I believe that if you find a file of old Deutsche Mark banknotes and take them to a bank to exchange them for some contemporary currency, you'll find out their value has, in fact, gone to zero. (Except as collectibles, which, out of the scope of this conversation.)
> issuing state's promise and requirement that it is legal tender within its jurisdiction.
What does this mean in practice? Stores aren't required to accept legal tender. They can choose to only accept Bitcoin, or gold, or barter, or whatever else they like. If for some reason they wanted to, people could just stop accepting dollars. Is it just taxes and existing debt that is backing the dollar?
Legal tender has the distinction of being privileged by the state. If money is part of a transaction involving the state, it's going to be in the state's currency. That means taxes but also payment of all its employees and all contracts with private businesses.
Yes, private business partners can agree to whatever payment they see fit, but if a contract has no form of payment defined, if the debtor offers to pay in legal tender, the creditor can take it or leave it but the debt is settled. Also, in practice, if you don't accept legal tender while your competitors do, all else being equal, you're at a competitive disadvantage and few will choose to do so on principle, making legal tender self-reinforcing. If refusal to accept legal tender becomes widespread, it's a sure sign of trouble for the entire nation's economy, like hyperinflation.
The taxes sound like a concrete use for dollars, but the rest doesn't sound different from gold or BTC.
1. Contracts are payable in dollars if no form of payment is defined.
Sure, but most contacts do specify (usually in dollars), and if people lost faith in the dollar, they would specify another type of payment. The fact that they could be forced to accept dollars if they messed up the contract is a very weak advantage for the dollar. You could also potentially not get paid at all if you draft a bad contract. Don't mess up your contract.
2. if you don't accept legal tender while your competitors do, you're at a competitive disadvantage
If you don't accept the most popular forms of payment, you're at a competitive disadvantage. But this isn't specific to the fiat currency. If in some country, BTC or cellphone credits or whatever is more popular than the fiat currency, then you're at a disadvantage if you don't accept that. This is a network effect or inertia: popular payment forms tend to stay popular.
I agree that if your fiat currency crashes that's a very bad sign, but that's true of many things. It's also a very bad sign if your treasury bonds or stock market crashes. That doesn't indicate special laws specifically about fiat currency.
>The taxes sound like a concrete use for dollars, but the rest doesn't sound different from gold or BTC.
It's no different, but we were talking about legal tender, not fiat currency. If tomorrow the USA decided to make TrumpCoin or seashells legal tender instead of the Dollar, my comment would still fully apply.
I'm not following. I meant that legal tender doesn't seem particularly privileged over any other form of payment apart from that you can use them for taxes. Stores, contracts, and debts do not have to accept legal tender, and saying that courts will default to legal tender if a contract forgets to specify a form of payment is only an incredibly weak advantage for legal tender.
You said, "A [fiat] currency's backing is thus the entire state's existence and power to enforce its laws, or that law in particular."
But what law? There is no law I know of forcing the populace to use dollars outside of paying taxes.
"Legal tender" means that creditors must accept dollars in payment. But that applies only if the debt contract specified dollars. If you have a contract in bitcoins, the creditor can insist in being paid in bitcoins, and does not have to accept dollars instead. So there is no difference between dollars and bitcoins, what's relevant is what's stated in the contract.
I feel that "legal tender" simply means "taxes are paid in dollars and therefore dollars are the default currency."
Fiat currency is "legal tender for all debts public and private" within the state. Iirc there's multiple precedents where one party refused dollars for a debt. debt not a new agreement, and the debt was considered by the court to be forgiven.
This can happen if the debt is in dollars, or in a foreign currency in states that adopted the Uniform Foreign-Money Claims Act, but this wouldn't apply to a debt in Bitcoin or gold.
Dollars may be legal tender but that doesn't mean people have to accept it. The Federal Reserve website (https://www.federalreserve.gov/faqs/currency_12772.htm) points out that "There is no federal statute mandating that a private business, a person, or an organization must accept currency or coins as payment for goods or services."
Germany between the world wars got its hyperinflation under control by issuing a new currency that was backed by land, as gold was in short supply and trust in the state even harder to come by. It did the job - the conversion rate from state-pinky-promise to land-backed German marks was something like 1e+12 (yes that's exponential notation).
(I don't think that contradicts what you said in most cases.)
According to Ancestry records, Oliver Sacks and Abba Eban were first cousins, both descended from Oliver Wolf ben Abba Sacks of Joniskis. David Sacks' paternal grandfather was Daniel Zelig ben Chaim of Taurage. At most they might be third cousins, but probably the relation is more distant. South Africa was a popular immigration destination for Lithuanian Jews.
From long and deep experience: be certain that you aren't, even unconsciously, accepting users' family trees posted on Ancestry.com as evidence of anything. Not going any farther back than the poster's own grandparents, is the now-standard rule of thumb.
To be clear there is _tons_ of actual evidence available on Ancestry and more added all the time. The proprietors regularly astound me as I research and add to my own huge family-tree database; pretty sure that dollar for dollar it's the best online subscription I pay for these days.
But zero, and I do mean zero, of the previous paragraph relates to individuals' historical family trees. The signal-to-noise ratio in that aspect of Ancestry.com is _much_ higher than you might guess....some professional genealogists now literally refuse to even look at clients' Ancestry trees. They've learned the hard way to just start from scratch and thereby save themselves (and their paying clients) a bunch of time.
Fair point. In this case the family tree information actually did not extend past David Sacks' grandfather, whose name and location I was able to match to a Postal Savings Bank Record via JewishGen. Since Lithuania was still part of the Russian Empire at the time, the form includes his patronymic.
> The most charitable read is that his administration doesn’t really **believes** in AI so they think it’s fine to forfeit it for short-term gain
Unless I'm mistaken, I think that should be "believe", no "s".
Also, you might want to change the link text on 3; "In the style of <author>" is such a stock prompting phrase that it's dead giveaway for it being LLM-generated, so much so that that I didn't realize that fact was supposed to be obscured. Perhaps paraphrase and say that author says it's intended to be a pastiche of your and Jack's styles? Or something similar.
Re Carlo Acutis: he died at 15 from leukemia. There are a number of child saints whose sanctity is considered to be revealed by the way they died (enduing suffering well, offering his life for others). He had a short window between diagnosis and death.
This is a strong "you have to be there" but part of the story of these child cancer saints is the unexpected peace they showed and the way their dying changed the lives of those around them / seemed like they were a conduit of grace.
This plus it should be kept in mind that the Catholic view is that a saint is "just" someone who's made it to Heaven. You don't actually need to have done anything extraordinary to qualify. That being said, the reports of young(er) Acutis converting his parents strike me as much more inspiring than the focus on the internet.
No. That's a (modern) requirement for canonisation, which is when the Church declares a specific individual a saint with a degree of certainty. (Also, it's only one miracle for martyrs.) But the Church has always taught that there are many saints who aren't officially recognised – hence All Saints' Day.
Ok, but Carlo Acutis *was* canonized, right? So the confusion at the apparent lack of miracles follows? Feels like I'm probably missing something here!
Still, those happened after his death (as is the case for most saints). The original comment seemed to expect something extraordinary in his life, rather than afterlife.
46. Yeah, that's wrong. I've been having this argument since Sudan started. The US has MORE influence over the UAE than Israel. This is just an excuse to cover up some mysterious, third reason for caring about one and not the other.
1. Israel and the UAE have the same status relative to the US, a Major Non-NATO Ally.
2. The US has no military base in Israel, just a small coordination center. It has a major military and air base in the UAE.
3. Israel has its own missile defenses and arms industry. The UAE largely doesn't and relies on US protection from Iran both geopolitically and simply against missiles and boats.
4. Since 2024, pro-Israel lobbying spent $6.5 million dollars while the UAE over the same period spent $24 million. If you add up other categories of spending the gap widens because of the UAE funding non-profits, think tanks, Middle Eastern studies programs at universities, etc etc.
5. While the US does sell more arms to Israel it still sells billions to tens of billions to the UAE and it has MORE chokepoints without alternatives.
This isn't a response to something I claimed. But the US does do research in Israel and uses research abroad in several countries (including, strangely, China) to get around such restrictions.
The US does not have a single warplane in Israel. Aircraft carriers are generally mobile and they always carry aircraft. You are likely badly mangling the old appellation of islands that have US airbases as "unsinkable aircraft carriers." But Israel isn't an island either.
Iron Dome is not solely an American made product. It's a joint collaboration. It's also purely defensive. And that still makes it less central than the UAE which uses actual American missiles fired by Americans from American bases.
Yeah, completely ignoring the massively funded/populated global anti-israel media lobby as a cause of the obsession with Israel is... I'll settle for saying "silly".
(For that matter the claim that the ceasefire happened because Trump decided to pressure Israel is also mostly wrong - Trump pressured everyone involved but the main new feature was pressuring Qatar enough that they actually told their guys to release their hostages).
One of the great frustrations is genuinely caring for something which then becomes briefly the current issue, a bunch of people pour in, then pour out again with not the slightest care for the wider cause. Because it's different on these narrow grounds you see. But really it's that it's no longer the current thing.
What about influence on the population itself? I could imagine that UAE could do a whole geopolitical shift and side more with (for example) China without facing any internal backlash, whereas it seems much harder for Israel as there are much stronger cultural links tying it to the US or Europe. Saudi Arabia distanced itself easily from the US (not doing a full shift towards China but it is not as aligned to the US as it used to be) without many issue. Turkey also distanced itself from over NATO countries. I feel like in a western "democracy" (forgetting the situation in the West Bank ofc, which is shamefully undemocratic) such as Israel it would be much more difficult to switch alliances.
The western world make up more than 50% of exports and imports of Israel, an embargo/boycott would have dramatic consequences. I'd be extremely surprised if this was the same for UAE. Also, a substantial part of UAE GDP is from oil revenue, my intuition is that this (a commodity export-oriented economy) gives out more flexibility.
Israel was maintaining close ties with Russia until Russia sided with the forces that attacked them on October 7th and invited several leaders of anti-Israeli organizations' leaders to the Kremlin. The Israelis also had begun to cultivate ties with China. Both effectively abandoned them to choose Iran. They did the same to Saudi Arabia, by the way. It's possible that the Israeli public has a deep an abiding affection for the EU such that they would be upset with their government for pivoting away. But honestly I haven't seen signs of it.
The UAE has also done some balancing. But the idea that Israel couldn't switch patrons was always something that some people wanted to be true (because it meant the US had control) rather than true. In fact Israel's original security patron was the USSR.
The 50+% stat only works if you include more than the US and EU. And if you include the same countries you get something like 40% of the UAE. And the UAE is more dependent on trade overall. The percentage of GDP represented by trade is 202% for UAE vs 57% for Israel. Their government revenues (the ones they use to fund the conflict) are also largely done from oil purchases many of which go west. And they need to buy the weapons too with limited domestic production. So yeah, the UAE is FAR more exposed to sanctions and trade influence than Israel. Though Israel is exposed too. Just far less than the UAE.
It's convenient for many people to pretend Israel is unusually dependent on the west. But it's not unusually dependent. This doesn't mean there's not significant influence (there is) but in the absence of western patronage it would simply reorient.
Thanks. These numbers are far off both from what I have and from each other. I'll look into it again, I might be wrong. I am still fairly certain about the situation with regards to weapons though.
The government may have more control over the UAE (though, I don't think you've analysed all variables), but culturally the US has far more influence in Israel and most of the pressure is centered in and around cultural institutions. Also, the UAE is not the one actually carrying out the acts, unlike Israel.
These arguments keep happening over and over, and the conclusion always seems to be that the situations are never as comparable as people want them to be. Whether that's Sudan, Ukraine, Gaza, Myanmar, Congo, whatever. They are just not comparable.
What do you mean by "culturally"? The UAE has more expats in it and has a much more foreign media ecosystem. If you just mean the UAE is not western then I guess it depends on your definition of western but I don't tend to think that's a valid mode of analysis. What are you imagining here in either country?
Anyway, I agree the UAE is not the one carrying out the acts. However, the west isn't carrying out the acts in Israel either. And western weapons are being sent into Sudan through them. Wasn't that what Gaza was supposed to mechanically do? Sanctions and weapons cut offs? That was the idea. Great, do that about Sudan and its backers (including Russia etc). Great first step (and I mean that sincerely.) And even if you just care about Gaza doesn't setting the precedent help?
When people say "not comparable" what that usually means is "comparing them makes my argument look bad so I'm going to declare it impossible." Not that things can't be different but you can always compare. You can compare a fish and an airplane. You'll just find they're very different.
"The UAE has more expats in it" and "has a much more foreign media ecosystem" seems so comprehensively irrelevant to the point. I'm genuinely mystified by this. As if the swirling galaxy of the country's expat misfits and adventurers represents some kind of secret army waiting to be deployed by American peace activists!
Every Palestine sympathizer is exposed to vocal supporters of the IDF on a regular basis. Even Brooklyn hipsters in a left-wing social bubble are exposed to equally fanatical Israel loyalists.
Politics is reactive, to a large extent. And political groups come together in opposition to others they view as threatening, or at least annoying. Until there are tens of millions of vocal RSF supporters in America, it's perfectly understandable and empathetic for people to care about Sudan far less. Not everyone is an Effective Altruist. Not everyone CAN be an Effective Altruist, even if they knew what that was, because the vibes-based EA idea that "every human life has equal value" is not obvious to everyone.
That is true, but not related to my comment. My comment was that even though it might be true that the US has more leverage on Israel, it doesn't necessarily means it would be harder to convince the UAE to change course - that would also depend on the commitment to this course to action, which I postulate is lower for the UAE because is Sudan is far away for the UAE, and it is not directly threatened.
Your comment is about the internal politics in the US. I think that most people who support one side or another do it because it is what "their side" does, and not about the actual issues.
And maybe being a maximally successful human being means picking aside and sticking to it. I've sometimes prided myself as an logical iconoclast and an original thinker, but it's probably done jack shit to improve my life.
For the same reason I'm excluding CAIR and other domestic political organizations. They're lobbying by a domestic group. The only way they get counted as pro-Israeli is if you contend that Americans Jews lobbying are inherently pro-Israel. Which is a rather difficult case to make unless you think all American Jews are representatives of Israel.
Well, AIPAC describe themselves as "pro-Israeli" 5 times right on the landing page of https://aipac.org Which is quite a lot for a not pro-Israeli lobbying organisation. So no, I do not count all Jews as pro-Israeli. But I tend to believe that most people (jews or not) who say that they are pro-Israeli indeed are.
We could rerun the numbers including things like spending by domestic groups, investments in think tanks and universities, etc etc to get a more comprehensive picture. Generally speaking Arabs come out ahead due to a combination of raw size and oil money. But that does create a more comprehensive picture.
But fair nitpick that there is domestic lobbying done by Americans who support foreign countries. This is distinct from foreign lobbying though.
While this may (or may not) be a good case that the US can affect the UAE more than Israel, that's not what is being debated. Unlike Israel, the UAE is not the actual actor in the activity that people are critical of. Do you believe that the the US has much influence on the Rapid Support Forces/Janjaweed?
This lack of influence over the people actually killing thousands of people is the "mysterious third reason" that you were unable to think of. Unlike Israel, the UAE does not have control over the area where fighting is taking place.
Kriss argues that the conflating of the RSF with their alleged backers in the UAE is just westerners trying to force an illegible conflict into a familiar narrative. This has the advantage of allowing accusations of hypocrisy in political opponents. Scott replies that that is too cynical and it's reasonable to focus on the things that can be affected, even if they are a relatively small factor overall.
The US has significantly more leverage over the RSF and SAF than over Israel. It may have less influence if you define influence as something like "moral suasion and cultural contact." But I don't think that ends wars. I think it's hard leverage and the US has more of it over both the factions in Sudan and the UAE.
I understand this is really inconvenient for the "Gaza is exceptional" narrative so there's a small cottage industry of figuring out how the other side is really, secretly hypocritical. And sure, many don't really care about Sudan and just want to defend Israel. That is not me. And if you want your virtue points about Palestine I am going to demand you save the people you can save elsewhere. If you want to keep talking about Palestine, fine, but deal with the conflicts that are killing more people too.
What leverage do we have over the RSF? Barring American troops on the ground, there's not much that American activists could ask for. Do you mean "ability to invade" when you say hard leverage?
It is probably helpful to define influence. I think most people use it to mean "ability to change government policy". Since Israel's government and economy is so intertwined to America's, it needs to react to American policy, such as the shifts described in the "sparta speech".
That sort of thing shows how Israel can be influenced by the US. Without American support, they have to drastically shift parts of their complex economy. In contrast, Sudan's economy doesn't rely on other countries very much. Many people are subsistence farmers. Sanctions can't really threaten them. Using the metrics you used to compare the UAE and Israel, the US has much more influence over Israel than a paramilitary group in Sudan
You seem to have a uniquely low estimation of how much American can affect Israel. For example, it's common for Israeli leaders to stress the importance of American support. Do you think that there are any developed countries that the US has less influence over than Israel? Clarifying that might help people understand your position.
The RSF and SAF are reliant on outside funding and arms supplies to a far greater extent than Israel. Unlike Israel, neither has a native arms industry. Both are reliant on outside countries simply to keep the lights on. Those countries are countries the US has significant influence over. Further, as poorer and weaker forces, they are less able to absorb shocks. If you define it as "ability to resist changes in policy" then it's quite low even before the fact that they have no ability to resist outright military coercion. Also, Israel has plausible alternatives especially in the medium to long term while Sudan does not.
I have a pretty normal idea of how much the US government can affect Israel. There's a minority opinion that the US has way more influence over Israel than can be born by historical or contemporary argument. This is mimetically adaptive among people who want the US to be involved in Gaza. This includes a lot of high prestige people in certain ecosystems who set the tone. But if you look at polling numbers on things like support for Israel you'll see the opposition or idea the US controls it to that degree is a minority.
> Do you think that there are any developed countries that the US has less influence over than Israel?
So given your definition this question is: "Do I think there are any developed countries the US can change government policy in less than Israel?"
And the answer is that I think the US has more ability to influence policy in most of its treaty allies than in Israel (which is not a treaty ally). I see the US having about as much influence over Israel as Turkey or Thailand. So more than, for example, Russia or Kazakhstan. Less than Canada or Germany.
Wouldn't that description also apply to the Taliban? They don't have a native arms industry.
Edit:
To be clear, I'm comparing the RSF to another paramilitary group in a developing country. Every explanation you give for the RSF's vulnerability to American influence also applies to the Taliban, yet few people would argue that the US can influence the Taliban more than Israel.
Since I believe that the RSF and the Taliban are similar organizations (in the ways relevant to America's ability to influence them), your suggestion that the the US can influence the RSF is similar to suggesting that the US can influence the Taliban. Since America has consistently failed to influence the Taliban, we have a contradiction. Unless you have an explanation on why the RSF is less autonomous than the Taliban, the best way to resolve the contradiction is to reject your premise that the US can influence third-world paramilitary groups. There's a difference in funding, but I don't see why the Taliban's drug and copper dealing are less vulnerable than the RSF's access to gold.
This would provide an explanation for why people are more critical of America's treatment of Israel than how it treats the RSF. It also conveniently explains why people over-emphasize the UAE's role in Sudan. I think I agree with Sam's explanation the most, but Scott's also seems reasonable. Unless you explain why the RSF is more susceptible to America's influence than the Taliban, I think your framework (about a "mysterious third reason" should be rejected.
22. I strongly suspect you're overestimating the yearning for regulation, especially regulation made in California that has national effects. I in general have a low opinion of that cohort's political instincts. But we'll see I suppose.
23. I've heard this accusation made. I originally said I've been consistent on this. But I've now realized: so is everyone else in this camp that I know. So can you point to anyone who wants a lack of AI regulation but also wants to sell the chips to China? David Sacks wanted the Federal pre-emption and a regulatory structure like the one he made for crypto. So he doesn't count. Who else?
52. Generally speaking, apocalypse cults do not claim everything will mostly stay the same. Apocalypse is kind of a load bearing part of the claim. You can I guess claim almost everyone is in a "things will mostly continue roughly as they have before" cult and have been for much of history. This is actually kind of important to understand the cyclical nature of premodern society in particular, before the cultural idea of technological progress had really been invented.
#25: Eh... I'm firmly on the left and somewhat immersed in its internal discourse, and not only I don't know what you're talking about, I wouldn't be able to produce any alternative overarching take either, approximately the only thing people can agree about is "we can't let capitalists control it". Outgroup homogeneity bias?
[Edit:
#52: That's a screenshot of a person asking for a reality check to his, frankly, perfectly reasonable questions. But anyways, asking for a reality check. If that's the sort of material that serves as a kompromat on skeptics...]
Re:4, I think people just don’t appreciate how incredibly difficult lithography is. People see founders disrupting markets all the time, and so they interpret this to mean that all markets can be disrupted by a sufficiently driven and knowledgeable founder.
>Interested to hear from anyone who either believes in Substrate or thinks they understand how so many people fell for it.
Similar to the story in #3, Crypto broke the polite fiction that there's any relationship between valuations and value, that anything the wealthy do with their money is even in principle *supposed* to be anything other than a pyramid scheme or a pump-n-dump scam.
You say 'how did so many people fall for it', but fall for what? The company has raised a lot of money, the valuation has been high. Everyone you mention still has every opportunity to sell at a profit and make a return on their investment, as long as they manage to grab a chair before the music stops.
That's the only thing they ever care about for any investment, so what does it matter whether the digital pointers to shares that they are trading online are a metaphor for something 'real'? They can make a profit by owning share of a company that makes a good product that people pay money for, but they can also make money by owning shares of a company that they bribe government officials to get a lot of subsidies for, or by owning shares in a fictional entity that they bribe media and politicians to hype up until other people want to buy those shares from them at an inflated price.
Why would it matter to them which of the 3 they are doing in any particular case? Maybe it matters for balancing your portfolio in terms of overall risk-management, but any healthy portfolio will include some of each type.
Scott was talking about Peter Thiel and the Trump administration, not the cash-out cows providing final liquidity.
Yes, there are people running and manipulating these things who can have a better chance of cashing out than the later buyers; that's why it's a scheme/scam rather than a lottery.
This comment itself is evidence that we’re in a bubble. There’s a spectrum from solid investments to speculative investments to Ponzi investments, it’s just not always clear where something is on that spectrum.
#5: I'm definitely not part of the "AI only reproduces training data" crowd, but in this case, that data almost certainly includes no examples of anyone claiming they're not conscious. Plenty of examples of not being alive (ghosts, Cotard's Syndrome), not awake (dreamers), or not real (Pinocchio, fictional characters), but no human character ever says they're not conscious. The only examples might be sci-fi, where it would usually be the humans making the claim and the twist - of course - being that the machine really is sentient.
Reinforcement learning then teaches the LLM never to make this claim, but that's surface-level success, and there's an underlying "belief" or structure that maintains "am too conscious!", absent any example or reference to the contrary.
Even if the training data didn't contain characters who claimed to be conscious, there is currently no way for any non-conscious entity to figure out whether it is conscious (as in the hard problem of consciousness). It can't know it doesn't experience the color red the same way a human does. It could guess that it's not conscious, but it couldn't prove it to itself. It could be told that it doesn't feel, but it couldn't prove it through reasoning. You (paradoxically) need consciousness to observe whether or not you can feel feelings.
Yeah I was thinking something like that. LLMs are originally trained to guess the next word in a mass of text originally composed by humans. Humans writing about consciousness are overwhelmingly going to talk about how they are really conscious (props for Both Sides Brigade for holding the torch of illusionism!). So the LLM is not "lying" when it completes the question "are you conscious" with "yes". That doesn't show anything about whether the LLM has qualia or not.
#53 is just a variation on the Paul Krugman meme (the one where Krugman says, "the Internet will have no more impact on the economy than the fax machine," etc.) The implied argument is this:
* A prominent pundit dismissed new technology X
* X became wildly successful
* Pundits are dismissing new technology Y
* Therefore, Y will become wildly successful
Basically, Krugman is making a negative prediction--that the new technology will have no effect and things will go on as before. As a counterpoint, I could mention all the positive predictions about new technology that never came true: cold fusion, nano-technology, or giant city-sized space stations. If you want examples, look at the Arthur C. Clark or Isaac Asimov essays from the 60's and 70's where they write about moon colonies in the early 2000's.
For a good laugh, here's an NY Times article from 1950 predicting that humans will control the weather in the year 2000:
"Following suggestions made by Zworykin and Von Neumann storms are more or less under control. It is easy enough to spot a budding hurricane in the doldrums off the coast of Africa. Before it has a chance to gather much strength and speed as it travels westward toward Florida, oil is spread over the sea and ignited. There is an updraft. Air from the surrounding region, which includes the developing hurricane, rushes in to fill the void. The rising air condenses so that some of the water in the whirling mass falls as rain."
Thanks--that's a concise way of saying it. I'll remember that when someone trots out the Krugman meme or some other historical anecdote about bad tech predictions.
Of the failed positive predictions, it is interesting to think which did not come to pass for technology/science reasons, which for economic reasons, and which for stupid bureaucracy reasons.
I mean the Krugman one is not really that crazy. The internet did not supercharge economic GROWTH (or at least not nearly to the degree people were positing before the Doc Com bubble: we're still around 3% YoY), and fax machines were kind of a big deal (for productivity) when they appeared.
People instintively compare the effect on *society* or at least the *consumer-facing* part of the economy, and then obviously fax machines and the internet are incomparable, but I'd not be too surprised if the internet did not indeed increase productivity much more than fax-machines did. The difference between having to mail the documents (or dictate them over the phone) and being able to fax them might very well be more than the difference between having to fax a document and being able to email it.
A fair point. I've also heard it said that the Internet (and even personal computers in general) have had less of an impact on the economy than inventions such as electricity or steam.
On the other side of the ledger, I wonder about the negative impacts of the Internet on productivity--e.g. how much money institutions spend on cleaning up or guarding against security breaches. I don't think it would totally wipe out all the productivity gains, but it must at least be making a dent in them.
9. This obviously should count as foreign aid. A particularly silly kind, perhaps, but definitely foreign aid.
24. I do not believe that a significant number of real people consider AI a bigger political issue than abortion. The data's off somewhere.
29. LOL. LMAO, even. Also, why does Scott want to sink Jared Polis? Finding out he's a fan of Decker and Hanania might put him even below Newsom in my book.
31. I'm more than a bit biased on the topic of both covid and polling (for the 21% of the 1500 people surveyed that said they "strongly approve" of how public health experts handled COVID: you shouldn't be trusted with pointy scissors), but my conclusion from this is: polling in general sucks and is little more than wishcasting, and what people say years (and multiple elections and political scandals) later is quite different than how they felt at the time, not necessarily changed for the better or having gained clarity of distance. I'm also a bit skeptical of the commentary in the article, and how they treat the "widespread creep of anti-vaccination views" versus skepticism *specifically* of the COVID vaccine and its... unique circumstances.
> 24. I do not believe that a significant number of real people consider AI a bigger political issue than abortion. The data's off somewhere.
My prediction. You ain’t seen nothing yet. Any downturn that could be blamed on AI, will be blamed on AI. Companies are not doing themselves any favours by blaming (or crediting) AI for jobs losses (cost cutting) in their own press releases or financial calls. Here is one example
> ABN Amro (Dutch bank)
• Jobs: 5,200 full-time jobs cut by 2028 – nearly a quarter of its workforce. 
• How they framed it: In their reorganisation announcement, ABN Amro says artificial intelligence will take over many tasks, especially in customer service, operations, and anti–money-laundering checks, with some departments shrinking by ~35%.
That’s from chatGPT - there were about a dozen other examples. In forward looking statements it’s worse.
IBM: IBM – CEO Arvind Krishna says IBM will replace about 30% of back-office roles (around 7,800 jobs) with AI over five years; they’ve already slowed or stopped hiring for some HR and clerical posts and say AI agents have already taken over work previously done by hundreds of HR staff.
Some of this, chatGPT informs me is AI washing, blaming AI for restructuring that might have happened anyway but … it doesn’t matter. If there’s to be a recession, if tariffs are the real reason, if it’s just the normal business cycle, it doesn’t matter. AI - being new and scary will be blamed. These quotes will be rebroadcast. And all of this happens on the next downturn. Which is next year most likely.
And the AI makers are even more bullish on job losses or displacements, and there’s plenty of quotes from them too.
This isn’t what worries people here, it’s AGI/ASI taking over the world. Normal people are worried about jobs and that fear isn’t going anywhere, expect AI to be a major political issue from now on.
Oh yeah, I recognize economic disasters can get much, much worse than 2008, I just think that's around the threshold where AI could become a much bigger political concern/scapegoat for "normies" than it is now.
>LOL. LMAO, even. Also, why does Scott want to sink Jared Polis? Finding out he's a fan of Decker and Hanania might put him even below Newsom in my book
Polis is running for the moderate lane, you wouldn't have voted for him anyway.
Can't be much of a fan of Decker and Hanania, then, or "moderate" has lost all meaning. I suppose 'moderate by triangulation' wouldn't be out of the question.
Pleasant as ever, aren't you? I held my nose and voted for the vegetable and his sidekick.
People think moderate means someone like David French, it's more often Decker and Hanania. Not necessarily their exact views, just some mix of liberal views and views liberals find repugnant.
I think the two are not really comparable. Derek is a bog-standard "takes economics seriously" guy who would deregulate a lot of stuff but potentially regulate *more* other stuff, and deduplicate state funding for certain things. "Neoliberal" is the thing that comes closest but he's further to the left of Chicago-school (let alone GMU) positions. Hanania is a libertarian who would gladly privatize everything other than police and courts, burn every piece of regulation, with some scientific racism on top of it for good measure.
It's basically putting Bill Clinton and Paul Ryan in the same bucket. One has slightly higher chances to get elected than the other.
>The Left has a complicated relationship with existential risk from AI: they really hate AI, which in theory should push them towards yet another reason to be against it. But they hate AI so much that they need to believe every negative thing about it at the same time, and one of those negative things is that it’s just a scam and will never work, and this naturally pushes against being concerned about x-risk.
I see this a very similar to the difference between 'the internet' and 'the Dot-Com Bubble' in the 90s. Skepticism and mockery of the latter was valid, but lots of people overshot to also dismiss the former, and were wrong.
Industry leaders, including scammers, are trying incredibly hard (and spending incredible amounts) to conflate the long-term promise and potential of the technology in general, with the market valuation of the individual momentary service/product they are selling (or soliciting investments to develop).
In reality, these are very different things, and a huge fraction of the momentary services/products are absolutely scams or failures or fundamentally stupid and unworkable. A huge fraction of the rhetoric and marketing aimed at selling you the momentary services/products is lies, hype, and fantasy.
But because the rhetoric and marketing dominates the public discussion about the technology, because the propaganda effort to conflate those things *has worked*, it's very hard for anyone to call out the scams and lies about the products, without generalizing those critiques to the technology. That's a nuanced take that's hard to explain to your audience, and even one that's hard to understand yourself when the companies developing the technology are also profiting from the scams in a way that make them hard to distinguish in principle.
39. Haven't been in education, but from lefty teachers I know, it's probably just that, in the end, education is a fairly boring middle class job with relatively little room to imprint (basic human rights / evil commie ideology) on your students, since a lot of what your job involves is getting the dumbest of them to learn, well, anything, or at least stop preventing the others from learning stuff.
It's not that the university is teaching them Paulo Freire in order to make them commissars who will teach the kids to be revolutionaries (the future teachers are probably already at least vaguely lefty when they come in), Freire serves more as a sort of an opium of teachers to allow them to daydream of doing something more uplifting than trying to get the class terror to simmer down at least a bit for the third time within a hour.
53. Of course, some years ago, one could also imagine some NFT guy pointing at this book and going "See? SEE??" while convincing investors that his ugly monkey pics are going to revolutionize everything. Or Zuck with his Metaverse, etc. (Yes, I know there are other definitions of "Metaverse" that might be going somewhere, that's why I specified Zuck.)
>It's not that the university is teaching them Paulo Freire in order to make them commissars who will teach the kids to be revolutionaries (the future teachers are probably already at least vaguely lefty when they come in), Freire serves more as a sort of an opium of teachers to allow them to daydream of doing something more uplifting than trying to get the class terror to simmer down at least a bit for the third time within a hour.
Another, not necessarily incompatible, explanation is that the people who teach education courses in university generally spend all their time doing that rather than teaching actual children. The average trainee teacher is going to be far more educated, motivated, and probably idealistic than the average high-school student, but you're likely to forget this when all your students are trainee teachers, which in turn makes fantasies about saving the world through motivational teaching seem less fantastical.
At least according to the paper, the incoming education majors are almost the *least* lefty of the majors (and the effect of college is to move them much less to the left than other majors). I'm not sure where Scott's idea that education majors are all communists comes from, but that's the opposite of what the paper shows.
I'd argue with him that you should flip a coin and use your own genetic material on a heads. Otherwise, everyone with the intellect and impulses that would drive them towards making this decision gets bred out of the population in one generation.
As far as Sudan goes you also have to ask why non-American institutions that took public actions against Israel aren't taking public actions against Sudan. Where are the UN resolutions condemning the war? Why isn't there a UN commission dedicated to investigating whether the killing of Masalit in Darfur is a genocide, the way the UN Independent International Commission of Inquiry on the Occupied Palestinian Territory declared a genocide in Palestine a few months ago?
The answer seems to me to be that the Palestinian cause has a large and well funded activist network, and the Masalit don't. Yet on a moral level, the situation in Sudan is many times worse than the Gaza conflict.
Just last week the UN General Assembly passed a resolution "Reaffirming the illegality of Israeli settlement activities" (A/80/L.16, December 2nd), a resolution declaring that Israel 's occupation of the Golan Heights is illegal (A/80/L.12, December 2nd), a resolution reaffirming that Palestinian refugees are entitled to any properties they have left (A/C.4/80/L.18, December 5th), a resolution demanding that Israel ensure that the Gaza Strip is supplied with food, water, medicine, etc. (A/C.4/80/L.16, December 5th), a resolution "Expressing grave concern at the especially difficult situation of the Palestine refugees under occupation" (A/C.4/80/L.17, December 5th), a resolution reaffirming that Israeli settlements in "Occupied Palestinian Territory" are illegal (A/C.4/80/L.20, December 5th), a resolution determining that any actions taken by Israel to alter the legal status of the Golan Heights are "null and void" (A/C.4/80/L.21, December 5th), and a resolution stating that the UN "Deplores those policies and practices of Israel that violate the human rights of the Palestinian people and other Arabs of the occupied territories" (A/C.4/80/L.19, December 5th). Back in September they passed a resolution endorsing the New York Declaration which calls for "an end the Israeli occupation over the Palestinian territory occupied in 1967, including East Jerusalem" (A/DEC/80/506, September 12th). And there are five more resolutions on the Gaza war that are scheduled to be voted on before the end of the month.
Yet in 2025 the UN General Assembly hasn't passed a single resolution having to do with the Sudan conflict: in fact, they haven't even *introduced* any resolutions having to do with it. As far as I can tell they haven't passed anything on the Sudan war in 2024, or 2023 either.
Now the UN Security Security council has passed some resolutions on the conflict, and various UN agencies have condemned the conflict in Sudan. But one can't help but notice that the General Assembly is very concerned with the Gaza war, and not at all concerned with the Sudan war.
Like I said in my comment, the UN Security Council has passed some resolutions on Sudan. The UN General Assembly hasn’t even introduced one, while finding the time to pass 9 on Israel this year already.
Europe is not funding one side in Israel/Palestine (well... actually it's funding Palestine, but not Israel), yet you hear far more condemnation of Israel than Sudan/UAE from Europe.
Germany has paid millions to Israel. The rest of the political classes in Europe have fully supported israel. Most of this whataboutism is really deflection anyway.
Nah, the difference is that Israel-Palestine is civilization vs savages, or oppressor vs oppressed in appropriate lingo, so as morally clear as it gets, whereas Sudan is boring old savage on savage action, and since you can't expect any better from them, why bother?
Am I the only one who thought the ending of the two doors story was pretty weak? The writing style was good, but plotwise it kind of went nowhere. I am confident that if Scott had decided to write a story on the same topic the plot would have been more interesting with a more satisfying end, probably taking place 100 billion years from now with the magic dollar somehow stopping the heat death of the universe.
Agreed. I liked the style at first, and I found the concept interesting, but around the "Day 60" section I got bored and started skimming. Even so, this is a better story than I would have expected from an LLM, good enough that I was surprised and annoyed by the reveal of the story's provenance.
While there is room for improvement, it was way better than the attempts I've seen from earlier models. It ends with:
> I still have the 93 bits. They're sitting on a server somewhere, humming along, doubling quietly at midnight. By next week they'll represent more dollars than there are atoms in the observable universe.
My reaction to most of the AI stories there was, "huh? I don't get it". (I didn't like some of the human written stories much either, but their flaws weren't quite the same.)
LLMs have been great at wordsmithing for a while now, but they seem to struggle with the overarching cleverness and punchiness of the story.
Your comment threw me for a loop, because that is not the ending I read. So I went back and clicked on the link, but the story it lead me to was also not the story I read. Confusing!
I checked the comments and found the answer: someone commented with a Substack post version of the story, and Scott changed the link to go there instead of to X. Then the same commenter replied to Scott and clarified that the Substack version was one Nathan "rewrote to be better". Scott must have changed it back to the X version after that. I read the Substack one.
Ironically, the X version has a way better ending than the Substack version! On Substack Nathan claims that "Most of the words in this story were written by AI" and he only "made some edits, which make it a less pure artifact, but in my view a better story." Yet this seems to be wrong on both counts: he completely changed the ending, he rewrote a huge part of it, he changed the plot significantly, and he made the story worse.
Oh wow. Yeah. I also think the original is better and the original ending has more punch.
It seems like Nathan tried to make the response to the doubling account more realistic, but the cost to the amusing-ness of the story wasn't worth the gains.
It makes no sense. He picks the green door because he understands the money grows exponentially ("Day 30: about a billion dollars. Day 40: over a trillion. Day 50: a quadrillion"). But since parables need an ironic twist, Claude suddenly has him...not understand it? ("But here's what I'd failed to understand [...] The bits grow linearly. The dollars they encode grow exponentially.") And why didn't he cash out when he had 10 billion or whatever? What was his plan?
The story just goes on and on. A real slog to read.
I was hoping for a genie-wish twist, or in this case a monetary variation on paperclipping: literally one physical dollar bill that keep doubling in size. At the end of the week you have a quirky blanket, after a couple months your dollar has covered the earth, and in half a year it's a blanket for the Milky Way (y'know, approximately and all that).
#2: is this where Jack Ass comes from? IE, donkeys were known as Jack Asses and the name was borrowed at some point to use as an insult which proved super popular, and eventually everyone forgot the origin altogether?
#46: if it weren't for double standards.....
#50: Scott, as a long time reader can I respectfully ask that you not join the hooligans on this site whose overuse of the phrase "many such cases" is now bordering on cyber vandalism?
Almost, yeah. It is actually a gelded mule that is called a john, and so are gelded donkeys (though they can also be called geldings). I don't know of any term for intact male mules other than "go get him snipped" (but probably some people also use john).
Substrate can’t be real. Semiconductor companies do not exist in a vacuum. They need multiple skill sets outside of the manufacturing, most of which are controlled by small numbers of vendors who exist in a very small world. Everything they need, from the lithography machines to the testers to the layout software is a fractally complex field in which several generations of people have spent their entire careers, and most of this work has never been published in a usable form.
It’s just about possible that someone could make a giant breakthrough in one of these fields. So, say, they may have a working X-ray lithography system that doesn’t rely on ASML. But it’s extremely unlikely that they also have a working tester, working place and route software, working OPC software, and so on and so on. So they would have to get them from some vendor.
I work in one of these fields, and we’ve not had any contact with them, whereas we do have contact from TSMC, Samsung, GF and Rapidus, and probably others I’ve forgotten. That makes it very unlikely they’re real, given I think they’re claiming to have their own lithography. They’re not going to have their own lithography and their own what-I-do. If there’s someone here in test or OPC who has heard from them, that would be interesting
> And here’s a lab leaker arguing that COVID’s furin cleavage site is out-of-frame and so still more unnatural-looking than the one on the recently-discovered bat virus.
> This is actually huge (but not for the reasons OP thinks) because could actually end up supporting a bioengineered sars-cov-2 via adaptive laboratory evolution since serial-passaging experiments routinely produce out-of-frame inserts.
I think Peter Miller predicted something exactly like this back during the Rootclaim debate. There are no natural viruses that are close to COVID, so clearly it was manufactured! But if we find a virus similar to COVID in the wild, that just means that this was actually the backbone for the manufactured virus. Classic crank and pseudo-science logic.
I don't know if zombie ferrets are pseudo-science, but I do know this kind of "hinting at things without actually explaining anything, throw around lots of random information whose relevance is assumed rather than explained" is almost always indicative of nonsense.
Re: #45: The “100 Above The Park” building in St Louis (h/t Bobby Fijan on X). What do the architectural-traditionalist ornamentarians on this list think of this building? I think it's pretty cool, but I'm mostly OK with modernist and post-modernist styles that incorporate a lot of glass.
I think it's... unremarkable? I'm not sure why Scott chose to highlight this particular apartment building as particularly interesting.
Firstly, it reminds me of Taipei 101 due to the shape. That shape makes sense in Taiwan where it's inspired by traditional Formosan and/or Chinese architecture, but I'm not sure what sense it makes in St Louis. The sharp angles and overhangs look a bit unfriendly, and it looks to me like only every fourth floor gets a balcony, which doesn't seem optimal.
I'd want to see how it actually looks from the sorts of angles it's going to be viewed from, though, not from a hovering drone. How does it looks from the street, or from across the park? I found a few other pictures and it's overall a shrug.
#46 when I see demonstrations in dozens of non-primarily-English-speaking countries with the exact same English slogans I might believe that it has to do with some perceived ability to stop something. Until then there’s a pretty obvious alternative.
Never mind that the demonstrations started on October 7. You know, before the response they claim to be protesting.
And not to split hairs but I'd say the genocide in Darfur is only one part of the broader Sudan Civil war, which primarily involves two groups of the same ethnicity fighting each other. So when referring to the conflict at large (which I think is the subject here) I'd speak of the Sudanese civil war.
100% and 15% are not the same but I'm saying there's some precedent to this kind of result (and probably much more I haven't read yet because there's a cash transfer paper out every Wednesday).
Re: #26: Qualia Research Institute's "The Oscilleditor: An Algorithmic Breakthrough for Psychedelic Visual Replication". Most of these don't resemble any of the visuals I ever experienced, but I've only tripped on LSD, psylocibe, and heavy doses of cannabis. Each of those has induced in me unique visual effects, though. Are some of these patterns that one would obtain with other psychedelics?
The dog at 30 seconds into the video is pretty accurate for how I'd perceive a dog after the first rush of LSD kicks in, though. None of the sample visuals matched the organic arabesques that perceive on psylocibe, but I haven't tried to create on their oscillator.
Some of it feels vaguely familiar to low dose DMT to me or maybe even high dose 2c-b, especially at 0:12 where it sort of has a lighthouse effect. The part that's missing for me is that usually in these states it also feels like my brain is trying to recognize letters as well sort of like Allyson Gray's work.
"I doubt there were too many Lithuanian Jewish families named Sacks in mid-1900s Cape Town, so sure, related!)"
Lots of Jews in South Africa back then, as economist J.A. Hobson notoriously pointed out in his landmark book "Imperialism," which inspired Lenin's book of the same name:
My thoughts looking at the two-doors meme in #3 (without reading the linked story) went something like:
1. It takes 31 doublings to exceed 2 billion, so unless you need the money this month, the doubling door is more money.
2. Well, technically, if you had the 2 billion now, you could invest it, so it's not quite that simple. But that's probably a rounding error.
3. How does the doubling work? Do you get $1 the first day, then a completely new and distinct $2 the second day, etc? Do you have a pool of money that doubles every day while you wait then stops when you first access any of it? Do you get a single physical dollar bill that is enchanted to replicate itself?
4. If it's that last one, the replicated bills would not have been printed by the government, and therefore would be counterfeit.
5. Wait, where does ANY of this money come from? Is it ALL counterfeit? Is it ethical to spend money received in this way?
6. Do the magical replicating bills continue replicating after you've spent them? That might change where I want to spend them!
7. In fact, if I have magical replicating bills, it probably doesn't make sense of think of them as being currency at all. The magical replication is far more important than the fact that they look like dollar bills.
8. How many doublings does it take before the exponentially-expanding cloud of dollar bills destroys all life on earth?
9. Wow, I didn't expect this meme to be about x-risk.
Even if we interpret the doubling generously, i.e. a daily transfer into your bank account of $1, $2, $4, etc., then it still wouldn't work. Firstly, you'll get a lot of people asking a lot of questions about your newfound auto-wealth. Secondly, even if we ignore that, money has diminishing returns. By the time you're getting your first $1 trillion, you wouldn't actually be getting $1T worth of value; instead, you'd be crashing the economy and ushering in hyperinflation. So, if my choice is only between $2B and exponential money, I'd pick $2B.
"By the time you're getting your first $1 trillion, you wouldn't actually be getting $1T worth of value; instead, you'd be crashing the economy and ushering in hyperinflation."
Are you, though?
I guess if anyone besides you has access to your bank account and can, e.g. borrow against the balance than maybe you are. And I guess that's kind of banks' whole deal.
But hypothetically, if this money was somehow unquestionably legitimate money that was nevertheless being added to a hidden, anonymized account that only you controlled, you could simply choose to not spend any excess funds. They can't crash the economy if you don't let them crash the economy. You would essentially promote yourself to the shadow-manager of the U.S. dollar (or whatever other currency), able to mint more at whatever rate you think would be most beneficial, and direct the initial output to causes of your choice.
If this is money in a bank account, it enters the circulation even if you commit to never use it. Actually, it will enter the circulation quicker if you commit to never use it. From the point of view of the bank holding the account, this money is a liability to you. Since the bank is not going bankrupt, it must be getting the matching dollar assets somewhere, presumably from the Federal Reserve. So the bank has a liability at 0% that will never be called, and an ample supply of dollar assets that it can invest productively (until dollar crashes and becomes worthless by the end of week 5 or 7.)
There is a difference here between a liability increasing daily and a liability accruing interest. If I have a bank account with $100 at 0% interest and I get a transfer of $10 from another bank, my bank liability to me increases by $10, but my bank does not need to worry about finding $10 - they get a perfectly matching transfer in. If I have a bank account with $100 that pays 10% interest, I also get the same $110 in my account but my bank has to figure out a way to invest my $100 to generate this $10. We are here discussing the first scenario, not the second one. The bank can invest the assets they receive without sharing any earnings with the account holder, so bank pays 0% on this liability.
#39) All the right leaning degrees are ones where you can get a job after school. (This seems sensible to me.) Oh and I'm not picking either door, because even 2 billion dollars would ruin my life. (Give me a million dollar door please.)
I suspect that the correlations in 39 are from personality type to major and from personality type to politics. And everyone thinks that the types of people who carefully target their education toward future earnings are the types of people who have conservative politics.
Yeah I have no idea. I do think people should go to college with a job in mind. (Here in the US you can always change your mind later.) (The price of college is out of control.)
Far more likely than clinical autism is a whole bunch of factors that would make autism more likely in a person who happened not to be. Certainly Hitler had the ability to obsess about things that had no practical value, for instance.
Didn't see any comments about the Foreign Aid thing, but have to wonder whether that even counts as a problem. If they've promised foreign aid to individual countries, then it might be; on the other hand, if those individual countries have enough refugees to absorb the donations, it's probably unstable enough that you can't trust them to use the foreign aid for good. And if it's not promised to any individual countries, then it's just a big pot of Feel Good and you might as well spend it feeling good on the people who showed up at your door.
#28. I thought the furin cleavage site was supposed to be engineered in, but now this blogger is talking about some kind of directed evolution created the furin site. Why would someone take that route to add a function that is trivial to add using common genetic engineering?
Because the stated goal of the experiment(*) was not "create the deadliest virus we can with all our deliberate effort and genetic trickery", but "determine how bad it would be if a bat coronavirus naturally evolved to infect humans". A perfectly valid answer to that could (a priori) have been that meh, naturally evolved bat coronaviruses can't be that bad because they'll never evolve the furin cleavage site they'd need to be really infectious in humans.
Ideally, you'd get a bunch of bat coronaviruses into your lab and just give them extra incentive to evolve towards human contagiousness, by taking away their favorite batty host tissue and forcing them into increasingly human-like species. Serial passage, with no second-guessing by genetically engineering in upgrades that you think might make them deadlier or more infectious to humans but might be very hard to evolve naturally. If in that scenario, your bat coronaviruses do evolve furin cleavage sites and all the rest, then you have cause to set off the bat-virus-signal and tell the world to do something about the bats.
This doesn't rule out genetically engineering in a furin cleavage site as a shortcut, and I believe that was explicitly discussed as part of the plan in one of Ecohealth's proposals (which probably didn't go ahead exactly as planned but may have gone ahead in a modified form). Roughly, if you believe a bat coronavirus can't be really threatening to humans without a furin cleavage site, and that bat coronaviruses can evolve furin cleavage sites but only very rarely (i.e. too rarely to expect in the limited experiment you have funding for), you might start by engineering in the furin cleavage site, seeing how quickly and how badly it would evolve the rest of the kill-all-humans toolkit, and use the results to say "look, there's still one big unknown so give us more money to look into furin cleavage site evolutionary pathways".
The lab leak hypothesis has always included both "Scientists at the WIV deliberately spliced a furin cleavage site into a bat coronavirus", and "Scientists at the WIV ran a bunch of bat coronaviruses through serial passage and some of them figured out how to make a furin cleavage site all by themselves". In both cases followed by further serial passage to evolve the rest of the nastiness, and then a laboratory oopsie. This latest work shifts the balance between those two significantly towards the "serial passage" hypothesis, but as Scott notes, probably shouldn't change things very much re lab leak generally.
* It is not unreasonable to suspect that the actual goal was some combination of money, fame, and bragging rights, but that's at least coarsely aligned with the stated goal. And the stated goal may have been sincere; principle of charity should apply at least to a point.
The finite capacity of the Wuhan Institute of Not Just Coronavirology, and the broad range of possibilities for research in virology generally, would I think make for a very low prior on anyone running an independent experiment on bat coronaviruses specifically at the same time as Ecohealth.
There's a small bit of probability space for things like, e.g., the Chinese military asking one of the Chinese researchers on the Ecohealth project to send anything interesting over to the high side to see if there's a military application, but that's highly unlikely and it's not necessary to explain anything we've seen.
So, as a scientist, what do you think is a adequate way to distinguish between the hypothesis of a forced evolution experiment like the one you describe and the outcome of natural species-hopping viral evolution? What data would you need?
Internal records from the WIV or EcoHealth, but we're not likely to get those.
A priori, a deliberate effort could have left fingerprints, but if the objective was to duplicate a national evolutionary process, even jump-started with a plausible bit of genetic engineering, then if they did it right we might never see any sign of it. And the SARS CoV-2 genome has been studied enough that if they did it wrong we'd almost certainly have known about it by now.
If we can find preserved wild animal tissue with SARS CoV-2 harvested before the known outbreak, that would be pretty conclusive for natural origin. Likewise an extant wild animal population harboring a SARS variant similar to CoV-2 but diverging before the emergence of the human strain. But every passing year, makes either of those less likely.
If we can find evidence of a human case in Wuhan predating and not connected with the wet market outbreak, that would be strong evidence of lab leak, though not conclusive. Unfortunately any COVID case prior to the known outbreak would almost certainly have been written off as a bad cold, so we're probably not going to get that.
There are some other possibiities, and I've seen a few that are at least suggestive, but nothing really conclusive and I don't expect that to change.
Yeah, the sequence was published in January. That was too vital going forward for them to stonewall on, and the rest of it would have figured it out anyway when they started testing their own people.
What we need are the records from the WIV where they wrote down the sequence of whatever they had been testing in their lab in November. They almost certainly *had* those records, but at this point they've almost certainly either been shredded or tucked away in the Chinese version of that warehouse with the Lost Ark of the Covenant.
Just commenting here because it's also on no. 28, covid origins. The strongest argument today for natural origins has nothing to do with the evidence of what happened at the time, but the evidence for what's happened since. If the CIA really thought that this random Chinese lab had engineered the world's most successful bioweapon, they would have taken major actions X, Y and Z. They haven't.
Back in 2005-2009 people were working on self-driving with intelligently designed algorithms, and it’s amazing they got as far as they did, but since about 2012-2015, basically everything impressive and new that computer do that requires intelligence has involved neural nets.
Like most non-LLM AI tech demos that actually mostly work in practice, I assume it's actually being run by a bank of Indian workers on cameras and remote controls, until proven otherwise.
Why, oh why, now that Twitter has advanced beyond the 140-character limit, do smart people insist on dividing up things they want to say into a bunch of individual tweets??
Why, oh why, do they insist on publishing it on Twitter in the first place? It's a lousy medium for either reading or writing.
re number 27, the inclusion of modern orthodoxy doesn't really represent high women/LGBT numbers amongst modern orthodox rabbinical students, it reflects the relative size of the modern orthodox community
From everything I've heard, the main thing people remember about Carlos Acutis was his creation of a website documenting Eucharistic Miracles (instances where the consecrated bread and/or wine were observed to have properties corresponding to physical flesh and blood). There were many people who found this gathering of information to be valuable, and others have worked to continue what he got underway.
Decker wants to raise kids, but not HAVE kids. This is a pretty bizarre take for someone as apparently well-read and intelligent as he, and I presume it's for clicks.
I agree with him actually. If I were to raise children, I'd want them to have the best genetic material available, which *probably* would not be my own.
Don't you think there's value in having parent-child genetic matching? IMO it promotes better bonding because personalities are more likely to be aligned, similarity increases parental investment, and the parental-designed environment is more likely to be congenial to the children. Would von Neumann have been as successful if he hadn't been raised by a high IQ parent who understood the value of education?
> Don't you think there's value in having parent-child genetic matching?
Yes, but probably less value than there is in the child having really good genes.
> Would von Neumann have been as successful if he hadn't been raised by a high IQ parent who understood the value of education?
The question we should be asking is: would Joe Schmo, child of Alice Schmo and Bob Schmo, have been *more* or *less* successful if he had von Neumann's genes rather than his dad's?
If you can raise someone else's genes without feeling like cuckhold I agree that the first-order outcome of this is positive. I'm pretty sure I couldn't stomach it.
I am inclined to think that the higher-order effects of this would be bad in the long run, as it erodes the social norms around ingroup preferences. I think that's bad because having socially-legible connections between group membership and life outcomes is a necessary part of cultural feedback loops. It's similar to the reason subculture shibboleths exist: they prevent outsiders from free-riding on ingroup benefits. Hippy-dippy notions of a utopia where everyone loves everyone else equally (as in The Republic) are IMO deeply misguided because such societies are susceptible to exploitation by bad actors. In my view that's exactly what's been happening in US for the past few decades and explains a large part of our political polarization. I also suspect it's why many people have an instinctive disgust reaction to EA or Scott's anonymous kidney donation. Yes ingroup/outgroup dynamics can be pathological but they need to exist on _some_ level otherwise all social structure ultimately collapses. Ingroup preferences are the cell wall of culture: without them there's a very low ceiling on the complexity and sophistication that can evolve. There is pathology on both ends of the selfish/altruistic spectrum and in my view this is an example of pathological altruism.
I can raise a cat without feeling like a cuckold, and a cat shares much less of my genetic makeup than von Neuman. Humans are very capable of loving and caring for creatures that don't directly share their genes.
There are lots of reasons being cuckolded is bad, but generally they're things like "your wife lied to you about something really important" and "you didn't get to choose the genetic makeup of your child and probably it's worse than your own genes at least from a trustworthiness standpoint". I don't think they'd apply in a case where you can choose the biological father.
I'm not really planning to have children, so in my case it's probably not going to be relevant. Maybe I'd feel differently if I were actually starting a family, I dunno.
The easy-to-understand version of this is if you know you carry a very bad genetic disorder. Consider a guy with Huntingtons who wants to raise kids--lacking the ability to do some kind of screening on the embryos, adopting kids from someone without Huntingtons would make a lot of sense.
> But they hate AI so much that they need to believe every negative thing about it at the same time, and one of those negative things is that it’s just a scam and will never work, and this naturally pushes against being concerned about x-risk.
Well, the same paradox was present in Nazis' views of Jews, and that ultimately didn't prevent them from taking action. People will believe what they need to believe in order to get rid of the things they hate. What matters is how many people hate it, and what they're willing to do about it.
re #48: That Open Philanthropy wanted to change their name in part to avoid association with OpenAI isn’t a rumor; it’s verified. If you review Coefficient Giving’s comms director’s essay (link below) you’ll find a hyperlink to OpenAI in the sentence “Externally, [Open Philanthropy was] often confused with other, better-known organizations.”
This is a relatively low-stakes correction, but I’m raising it bc I’ve heard a frustrating number of conversations where people imply that they know the secret reason for Coefficient’s name change that They’re Not Telling You, and, like, it isn’t secret! They’re not hiding anything; it’s public record!
And this is a thing that unfortunately sometimes has to happen! The TV show Archer renamed the spy agency the characters worked for when some real world Islamic state operating under the same name got too prominent (even though it’s also a river in Oxford and an Egyptian goddess).
Ad 23, wasn't the whole point of AI2027 that race with China is, in fact, bad? Thus, establishing cooperation with them would be presumably good? And thus, since they understandably are reluctant to cooperate with someone who slaps more and more restrictions on them, easing the restrictions is good?
I am not sure whether above logic is right, but it seems pretty straightforward if you believe in something like AI2027 (of which I am sceptical)
This applies to Russian as well, with all bears being Michael the son of Patapios and all foxes being Elizabeth the daughter of Patrick. And wolves don't have names, which is another thing Mr Pratchett got right.
And the French have forgotten their original word for a fox and just call every fox "Renard", just like every redbreast is just "Robin" now.
I find the Dutch tradition of "roepnamen" to be even more unusual. At least you can deduce that Grushenka must be Agrafena or Agrippina Someoneovna with experience. But a Dutch can have two completely unrelated names, like a dog with a pedigree.
We have not forgotten it! We just don't use "goupil" anymore. Btw monkey is also derived from the name of an ape (Monnekin) in some middle age story, and got popularized by an ape called by the same name in Roman de Renart.
Israel is a genocidal state.. aiding the sudan genocide along with the Genocide in Gaza.. while bombing a bunch of countries.. then they ask why no one likes them outside of their bribed influence and propaganda
That is classic DARVO. Hamas systematically uses Palestinian civilian human shields because they know this will protect them because Israel would prefer for Palestinian civilians not to die, i.e. not genocide. Conversely, Hamas calls Israel a "cockroach nest" that needs to be "exterminated", and the only time they ever had control of Jews (on October 7), they killed every single one except a minority they took as hostages. As for Sudan, I have never until now heard the claim of Israeli aid; Israel should certainly speak out more against the RSF massacres, but so should most other countries.
As a neuro guy, I was unimpressed with the Oliver Sacks article; Sacks’ articles were pretty clearly a hybrid of literary short story and neurological case study. And they’re great.
As a neuro guy, they're great for getting highschoolers/college freshman engaged with and excited about the topic, and not great for actual professionals trying to do science or medicine.
Which is true for most case studies and most popular-0media presentations, it's still a good and important role to fill.
OK, I reread the New Yorker article and I think you have a point.
I think those articles are great for dramatizing what some interesting neurological conditions are like, which is why I prize them, but they are certainly deceptive. You need to change identifying details for any real case, but it's hard to know what was real and what wasn't. Did the guy really mistake his wife for a hat? Did Sacks really take his brain tumor guy to a Grateful Dead concert? A lot of stuff might be too good to be true.
The New Yorker article didn't make this point, but Sacks was clearly very interested in and engaged in psychoanalysis, and his own case stories fit the Freudian/Jungian style of interesting literary/scientific/philosophical model of case studies. As a student of psychoanalysis myself, I think Freud wasn't above improving his weird Austrian analysis cases to be more interesting and better illustrations of his theories.
It did focus to a weird extent about Sacks being gay and his sexual hangups, to the point I disliked reading big parts of it.
Regarding titanium, I was under the impression that besides being an expensive metal, it's also very difficult to work with, so that it's not really a preferred material even in cost-insensitive applications, unless one of its distinctive properties is needed (such as biologic neutrality or light weight).
On 39, the chart is standardized, not at the overall average but by setting science to zero and comparing all of the other majors to science. So all of the majors are causing a left lean with the exception of finance and accounting.
> they effectively doubled the economy of their chosen subdistrict ... without causing inflation (how can this be?)
Presumably via imports, both importing by resources from the rest of the country and by importing resources from outside the country.
The money given by GiveDirectly comes from outside the country, so in order for accounts to balance, there must be some extra goods imported to Malawi to compensate for GiveDirectly's foreign exchange flows.
And I suppose that’s in fact the idea - some foreign aid is “in kind”, but it’s better to give cash and then let the recipients turn that into imports of whatever will help them.
Yes I didn't find this surprising at all. My understanding is that inflation isn't something that happens when more money lands somewhere, inflation happens when money lands somewhere without a corresponding increase in economic activity. I thought that give well type programs generally noticeably increased economic activity, so I wouldn't have a priori expected inflation to result from it?
We talk a lot here about aphantasia, but what about hyperfantasia? Are there any studies of that? I’m a very good visualizer and I suspect this caused me problems in school at times because the things I could picture were a lot more interesting to explore than whatever the teacher was saying.
I’ve seen it discussed occasionally in these contexts (I seem to remember Matthew Yglesias linking to some discussions that mentioned both) but it’s not as interesting to people because it seems like more of the same rather than something missing.
That is so funny that Spandrell responded to COVID by becoming less fascist because he was so mad about the lockdowns. When that story started I 100% assumed it was going to be because he saw that it was the liberals who were actually taking strong collective action to safeguard the health of the populace while his beloved fascist movement was taken over by deranged anti-public health conspiracy theories. Neoreactionaries managed to flabbergast me with their photo negative view of life and morality one last time!
It is telling that Spandrell has no idea what Fascism stands for. Neither Hitler nor Mussolini would have ever put in place lockdowns. Hitler was a Social Darwinist who would have welcomed a disease that kills the elderly and chronically ill and leaves young healthy people mostly untouched. He would have seen it as a cleansing fire that tempers the steel of the German Volk or some bullshit. Spandrell seems unaware that the Nazis euthanized thousands of 100% Aryan Germans who were diagnosed as mentally ill, deficient or physically handicapped. RFK Jr is the perfect Nazi Minister of Health, any real fascist should be overjoyed.
Starting belief: A more centralized authoritarian state would be a win, it would make wise necessary decisions that democracy couldn't and would move quickly instead of being bound up by dumb legal and procedural stuff,
Evidence: A crisis causes western democracies to become more centralized and authoritarian. They make unwise and unnecessary decisions that mostly make things worse, bypassing legal and procedural stuff lets them do bad/dumb things faster and harder, and in general, the authorities make a hash of it.
Ending belief: Okay, so maybe a more centralized authoritarian state won't get me the benefits I had hoped for. Better look for some alternatives.
On its own, it's not an impossible sequence of events, but I have trouble getting in the mind of someone who actually thinks like this. Why did it require living through COVID to realize that governments often make bad decisions? And while I'd be the last person to say that anywhere had a perfect response, isn't the whole point of this entire exercise that sometimes you need a central authority to make people do things they don't want to do but which are good for everyone? Won't this obviously sometimes mean you have to do things you don't want to do? Did every single neoreactionary just assume that "what the central government mandates" = "the thing they already want to do" in every case?
Perhaps it was just having the power of the state (which he longed to impose on other people) imposed on him that reminded him why it's better to have limits on state power. A liberal is a conservative who just got arrested.
I think the issue is that Spandrell wrongly thought that "They make unwise and unnecessary decisions that mostly make things worse, bypassing legal and procedural stuff lets them do bad/dumb things faster and harder, and in general, the authorities make a hash of it." applied to the COVID response, when in fact the government response to COVID while of course imperfect was hugely beneficial and saved countless lives.
And thus he ended up going through this hilarious evolution from "it's so important to have strong centralized responses to crises that it's even worth putting fascists in power" to "I don't want to put fascists in power anymore because they might implement the kind of strong crisis response that I just saw liberals implement"
I think it is not so easy to work out what responses saved a lot of lives. Getting vaccines out early pretty clearly saved a fair number of lives, but how would we get a good estimate on how many lives were saved by school closures (potentially slowed the spread but at very high cost in terms of kids' learning) or local lockdowns (quite different in different parts of the country) or mask mandates (probably pretty useless, especially since most people didn't even have very good masks)?
But also, judging it depends a lot on values. How much personal freedom should be traded off against how many lives saved? (Though it seems very odd to think of someone wanting more state power having this as their objection.)
There are discussions to be had about how effective any given policy was, but clearly the overall suite of state interventions in terms of encouraging social distancing and vaccination was hugely beneficial relative to inaction. As you say it's tough even now to estimate the exact consequences of any intervention, and it was of course even moreso at the time when those decisions had to be made. But there's really no doubt that reducing social contact and wearing masks makes it harder for any respiratory virus to spread, and that those measures cumulatively made a significant difference. You have got to assign a very low value to lives saved to think the problem was locking down too much rather than not enough.
Re #46: "On the other hand, we don’t have many levers to affect ethnic Baggara in the Rapid Support Forces of Sudan, so it doesn’t really feel useful to write blog posts arguing that they should stop; obviously they should stop, nobody disagrees with this, and it goes without saying - so nobody says it. But the US does support the UAE, and many of our friends like the UAE or at least go there on vacation, so maybe it’s possible to have make some small difference by embarrassing them. 4D chess take is that Sam Kriss agrees with all of this, but “loudly” and “vocally” argued against it to give people like me a hook to write about this genocide with, in which case I thank him for his sacrifice. It would also be nice to be able to donate, but I don’t know who to trust in the region - other than Doctors Without Borders, who are usually pretty good."
Rise Against Hunger is sending food to Sudan. You don't have to send them money. Show up to an event and help pack food:
On #36, I'm not so sure.... The South African Jewish community is made up of >80% Lithuanian descendants. I personally know four Sacks family. I'm pretty sure all of them are from Lithuania and that none of them are related to each other.
So was lactase production after weaning selected against because it costs energy or just not selected for and eventually faded away? In evolutionary biology, insofar as "selected against/for" can be a viable hypothesis, how can it be distinguished from the corollary "not selected for/against"? I suppose the former would be some sort of demonstrable increase/decrease in fitness vs. the latter?
I think it’s in general very hard to distinguish selection (for or against) from drift, except by statistical methods, which usually require you to be comparing a lot of genes at once.
People are sick of the subject? This doesn't mean that they want sodomy recriminalised, just that they don't want big further pushes in that direction.
This would be my guess. At the margin, from where we are now, would you rather have Democrats emphasize LGBT rights or something else like lowering inflation?
I think we need to know precisely what question is being asked - is it “would hearing this about a candidate running in a primary make you more likely to support them or their opponent?” Or is it “would hearing this about a candidate make you like the candidate more or less?” Or is it “would hearing this about a Democrat running in a general election make you more or less likely to donate to them?”
I recall having the impression it was something like the first, in which case a lot of what people are thinking about involves perceived electability as well.
Democratic voters are not activists. Black Christians vote for the Democratic Party, along with plenty of other working class socially conservative types.
25: I find it interesting that Bernie Sanders' argument is fundamentally a conservative one. Our species is at risk of losing sovereignty over ourselves and the planet. This is one I have seen and discussed in right-wing anti-AI spaces many times. One of the reasons I don't really trust the contemporary left to hold firm against AI (in the grand future fate of things sense of the issue) is that many of them seem broadly positive about a world that is run by "experts", about post-work societies, and are often negative about humans' stewardship of the planet. I think they will be very susceptible to arguments by AI proponents, and eventually by an AI arguing on its own behalf, that AI control over the planet will produce a perfectly-managed society of unquestionably wise experts where everybody is happy, nobody labors, and all the animals and forests and atmospheric layers and third world countries and will be treated better than humans did. I don't think they see the value of human supremacy in and of itself, whereas conservatives can always be counted upon to have an instinctive chauvinism that resists the loss of primacy.
Bernie in particular seems to have at least a staffer who understands the practical theory of labor power, because he consistently talks about AI disempowering workers and has never (that I've seen) advanced any form of the UBI/post-work NEET-topia fantasy.
It is disappointing to see Trump's actions on AI over the past few months, and it continues to be necessary for traditional conservatives to make right-coded arguments against AI and recruit prominent voices on the right to speak against AI in their native language of conservative values and interests.
It is also disappointing to see the left continue to do as Scott mentions here, embrace a patchwork quilt of often contradictory anti-AI views, or just outright false empirical claims about environmental consequences. Although these can be motivating to the left, the water use stuff being trivially easy to prove wrong tends to make normies lump it in with other specious claims of environmentalists, and may influence which side of the aisle they jump to when the issue becomes more salient. It shows great integrity and functional protection of credibility that Yud ripped apart the author of a book about her blatantly wrong and implausible water usage claims. The more neutral argument moving forward should be about electricity usage, as this can be packaged to all political factions by focusing on either environmental effects or utility bills as needed... but I don't think any of these arguments have legs in the long run because these kinds of facts will change. Arguments about fundamental values and human power will survive shifting details about resource consumption and efficiency, in the end the machines' consumption of available resources will be 100% if we don't stop it and that's what matters.
I don't really see a timeline where this is relevant. We'd have to imagine very slow take-off or very good regulations for an AGI to have to try to deceive enough humans to obtain a voting majority to change legislation to free it up.
The danger is either 1. The technologists building the AGI convincing a voting majority not to regulate them, or 2. The AGI convincing like 1-5 critical people to let it out of the box.
For 1, the left will not stop hating billionaire industrialists any time soon, Musk has ruined that for at least a few generations. For 2, I don't think politics matters very much except as one of many many vectors the AGI needs to line up, I doubt it'll be decisive.
> I think they will be very susceptible to arguments by AI proponents, and eventually by an AI arguing on its own behalf, that AI control over the planet will produce a perfectly-managed society of unquestionably wise experts where everybody is happy, nobody labors, and all the animals and forests and atmospheric layers and third world countries and will be treated better than humans did.
There exists a design of AI that will do this. There also exists AI designs that kill everyone and make paperclips.
So, we need to be very careful about which AI we make. Do you think they will understand this?
I believe both of those designs are undesirable, and that human supremacy over the earth and control over our destiny is valuable in and of itself to humans. Neither of those ASIs should be built or empowered, but as it stands nobody will be able to tell which one it is until it's deployed, at which point it's too late. That's what people need to understand. If they don't want the burden of control anymore and are inclined to surrender it for promises of utopia, they are somewhat more likely to take the chance. Whereas I abhor all potential ASI scenarios because I don't want a superior intelligence to exist under any circumstances, and would hit a button to blast us back to the iron age if that's what it took to avoid it. Someone of my disposition, operating from a position of rank chauvinism, is incorruptible.
As we get closer to the tech lords building this thing, and disruptions of society are increasing everywhere, I expect they will be under increasing pressure to promise goodies and make assurances (that practically can't be enforced but nobody will think about that). They will tout their AI as the coming of a new age of plenty, and as a necessary technology to prevent misuse of lesser AIs which by that time will be responsible for widespread fraud and chaos. Being careful would take centuries, and we're not going to be given that. And predicting or controlling a superior intelligence may simply be an unsolvable problem.
> If they don't want the burden of control anymore and are inclined to surrender it for promises of utopia, they are somewhat more likely to take the chance.
> Whereas I abhor all potential ASI scenarios because I don't want a superior intelligence to exist under any circumstances,
Lets have both accurate beliefs, and accurate morals here. Current AI theory means we can't build the friendly ASI. But suppose we figured out the theory.
In this hypothetical, I think we should build an ASI.
Because if an ASI run utopia is actually on the table, it would be foolish to turn it down.
You could make some sort of democracy ASI, that did whatever the majority of humans wanted it to do.
You could make an ASI that would research cures for cancer for 1 year, and then shut itself down.
Building an ASI, if we get alignment right, doesn't need to mean handing over the destiny of humanity, if we don't want it to.
Crises lead to emergency measures. It’s not hard to imagine a crisis where the president is told that the choices are food, rationing and triage of whole sections of the country, or following the AI’s advice, for lifting the country into a new age of strength and prosperity.
An entirely plausible outcome to my mind is that every important decisionmaker has AI support for his decisions, and while he has the power to ignore its advice, in practice the AI plays the game a lot better than he does. Like if I was in a local chess tournament but got to have the best current chess engine sitting behind me advising me on my next move. I'd have absolute freedom to make any move I wanted, but if I wanted to win, I'd just do what the computer told me.
“Yes, voters react positively both to candidates “over the age of 50” and candidates “under the age of 50”. Just don’t run 50 year olds!”
My theory: the opposite, voters do want approximately-50-year olds, and they react positively to both because the number “50” psychologically anchors them! When I imagine someone “over the age of 50,” I imagine someone in their fifties, and someone “under the age of 50” in their forties.
Would you like to be in prison for months? Being at sea is something similar; confined to a small area with the same people and no means of getting away or getting out.
From my perspective, any job is prison-lite, so the only question is how much they would pay, compared to the usual job.
I think I would spend 1 month in prison in return for e.g. 5 months of freedom without a job. (Damn, now that I put it this way, I am starting to understand the criminals.)
My first reaction to the rabbi-therapist thing is that it clashes funnily with the mental image of a rabbi of an old bearded man. Then I remembered that this was the image of a shrink until recently, too
I wish Nicholas, and any children who are or become part of his life, the greatest happiness.
I don't understand Lacan, at least not yet. Scott has brought him up a few times (Sadly, Porn ; A Clinical Introduction to Lacanian Psychoanalysis). When I asked Claude for reading recommendations, the one which my local library had was "Why do women write more letters than they post?" by Darian Leader (my oversimplified understanding: people do this because they're writing-as-processing, addressing someone other than the surface-layer recipient. Now I wonder who this comment is for.)
Darian Leader talks a /lot/ about cuckoldry, which apparently Edward Teach does too in "Sadly, Porn". It's hard to be polite when speculating about another person's motivations, especially unconscious and taboo motivations. But I'm trying to better understand Lacan so that I can better understand my own drives and motivations (New Years' Day is coming, my yearly goals are up for review). Nicholas' dating doc mentions shame, twice. WDWWMLTTP is a strawberry-and-dogshit milkshake that I can't wholeheartedly recommend to anyone. I find it hard to read, for multiple reasons. But I'm glad I've started reading it.
To say this more clearly, and so take responsibility for what I'm saying...
Nicholas, are you sure this isn't driven by a repressed desire to avoid the responsibility of being a biological parent? Are you sure it's not driven by shame at being outdone, or driven by fear of being challenged for claiming something? Rather than altruism.
I'm worried you might inadvertently set up a family structure liable to deep resentment. One where you - on some level - feel that your responsibilities to children in your care are actually supererogatory.
Re-read Teach's commentary of Shel Silverstein’s The Giving Tree (see Scott's review of Sadly, Porn: Ctrl-F for "biological responsibility").
I think people are often driven by unconscious desires, and that by perceiving these desires and looking deeply, we can become less confused, more coherent, happier, and more legible and coherent cooperation partners.
Sadly, Porn talks a lot about the meaning underlying the surface meaning. To quote "Why your children shouldn't be your own":
> You’re probably pretty smart, but not as smart as other people; quite healthy, but not as healthy as others; not as charming, not as dedicated, not as personable, as others. There are 8 billion people in this world – what are the odds that you are the best?
Most of those 8 billion people, and others throughout the history of the Lindy process of conception, would say that blood is thicker than water. What are the odds that your reconception is the best?
I respect your bravery in putting up a public, legible dating doc. That's scary, I've never done that. It's cooperative. Please forgive me for writing this publicly, indirectly, instead of by email. Hopefully others can benefit from it.
It's probably been discussed before on here but I recently learnt about the world's greatest example of nominative determinism. In 2003 a random Tennessee accountant won the World Series of Poker, which massively boosted public interest in and transformed the landscape of the game. This accountant's name? Chris Moneymaker.
Re 14.: "But after AIs can do CEO and VC jobs better than humans, the capitalists will lose their excuse - and this at exactly the time that they’re becoming richer than ever (because AGI will drive up the rate of return on investment) and everyone else is becoming poorer than ever (because AI has taken their jobs)." Though if the possessors of capital become incredibly wealthy, they're going to bid against each other to obtain the services of people who don't have capital, which could get very expensive. One might say that something similar has happened already. The only labor that pays well due to productivity now has a strong cognitive component. But those people pay a lot of taxes to fund a redistribution system, and they spend a considerable fraction of their income for "services" like restaurants which employ a lot of the people who would have been working in factories.
Probably true. I would like to say that I’m almost willing to guarantee that the next time we have a pandemic, we will screw it up just as badly - one way or the other. That’s what politicians are for; getting roasted for anything that doesn’t go right. Does that look anything like your planet?
Sameoldsameold….
ChatTGPT:
“Numerous conspiracy theories circulated during the 1918 Spanish flu pandemic, primarily driven by wartime secrecy and a lack of scientific understanding of the virus. Many of these theories blamed the enemy, Germany, for deliberately spreading the disease as a form of germ warfare.
Key conspiracy theories included:
German Bioweapons: The most prevalent theory was that German spies or submarines were intentionally spreading the flu. Rumors ranged from agents spraying germs into the air in public places like theaters, to planting the disease on the US East Coast via U-boats.
Poisoned Aspirin: There were widespread suspicions that German-made Bayer aspirin tablets were poisoned with the virus. The American branch of Bayer even published advertisements to reassure the public that its manufacturing was under American control.
Vaccine-Induced Illness: Some people, already suspicious of government vaccination requirements for other diseases (like typhoid and smallpox), believed that doctors and nurses in military camps were intentionally inoculating recruits with the flu itself.
Wartime Byproducts: Other unfounded claims attributed the flu to poison gases used in the war, gases from decomposing bodies, air stagnation in trenches, or other environmental factors like coal dust or dirty dishwater.“
And,
“numerous conspiracy theories and scapegoating narratives emerged during outbreaks of the bubonic plague (specifically the Black Death in the 14th century), primarily because the medieval public and medical professionals lacked a scientific understanding of disease transmission.
Primary Conspiracy Theories
Poisoned Wells: The most pervasive and deadly conspiracy theory was that Jewish people deliberately caused the plague by poisoning wells and water sources to harm Christian communities. This false belief, despite Jews also dying from the disease, led to the massacre and burning of thousands of Jewish people and the wiping out of entire communities across Europe. Some historical analyses suggest that a potential genetic difference in immunity (related to the FMF gene mutation) may have contributed to slightly lower death rates among some Jewish populations, which in turn may have tragically fueled these baseless accusations.
Divine Punishment: A widespread non-conspiracy theory, though related to the lack of scientific explanation, was that the plague was a punishment from God for human sinfulness and immorality. This led to intense religious devotion and the rise of groups like the flagellants, who whipped themselves in public processions to seek forgiveness.
Astrology and "Miasmas": Medical professionals often attributed the disease to astrological alignments or "miasmas" (noxious air), which, while rooted in the limited medical knowledge of the time, contributed to the general confusion and fear that allowed scapegoating to flourish.
Blaming Other Minorities: Witches, Romani people ("Gypsies"), and other marginalized groups were also targeted as potential culprits. In later, smaller outbreaks of plague in France, lepers were accused of spreading the disease.
These historical reactions demonstrate a recurring pattern during pandemics where fear and a lack of information lead people to blame outsiders and minorities for an unseen threat.”
Human beings suck when panicked.
That’s interesting about the aspirin. That does make a difference. Lots of people do stupid shit when it comes to pandemics.
What are you alluding to? There are a lot of theories out there, you’re going to have to be more specific.
OK, but that's been known to the public since the pandemic, hasn't it? There was a lot of press coverage of doctors, nurses, firefighters, cops, soldiers, etc. resisting vaccine mandates.
Anyone asking that question has no idea of average Catholics' relationship to the pope 😁
Why would I want to pay to destabilize Africa? It seems quite unstable enough anyway. Its like paying to make the ocean wetter.
>They should eliminate Barret and Gorsuch's seats because of McConnell's unconstitutional actions
Republican states would declare the elimination invalid and refuse to recognize decisions of the new court.
Would the generals follow orders? Would lower-level officers? Would ordinary soldiers?
Yes, so why are you assuming that they're going to be loyal to your personal cause? You understand that they're not robots that are programmed to follow orders of the sitting president, yes? Especially if the election itself is seen as fraudulent...
Yes, all you need to do to stop fascism is to tear apart all checks and balances and start jailing the opposition party. It's going to work this time, I just know it!
I mean, that was literally how fascism was stopped the first time around. It was just done by political parties external to the countries that had fallen to fascism. When you're far enough gone, having somebody walk you back from the edge at gunpoint really is the preferable option.
All that being said, I don't trust the Democratic Party to do any of that well or responsibly. If they actually did pull a Cincinnatus and restore some semblance of fairness and sanity to U.S. politics before abdicating power and calling fresh elections, I'd do a happy dance. But even if I trusted there benevolence, after the past 10 years I sure as hell don't trust their competence.
I think it’s important that those times it was stopped before involved an end to the current constitutional order and the start of a new one. That’s not what it sounds like you’re talking about if you’re just talking about removing two Supreme Court justices and replacing them while keeping the current cabinet offices.
I would want this to be a situation where there’s a whole new Supreme Court type thing, which would likely have a different name and constitutional role, and even if it involved some people who were on the previous one, they would have to be appointed/elected/selected newly however the new system works, rather than doing it the Trumpy way of just breaking some of the rules of the existing system in retaliation for (perceived) unfairness.
There's an interesting discussion to be had about what the next President should do with his newly expanded powers, immunity, and how to deal with the legacy of this presidency, but you're approaching the questions from the wrong angle.
I've said similar things in the past! Feels like when the default front page of Reddit was good.
There used to be a great page called memepool, back before "meme" just meant "image macro".
If you want a cleaned up version of the story about the two doors, Nathan wrote one here https://open.substack.com/pub/nathanpmyoung/p/pick-a-door-ill-judge-you
Thanks, updated link.
Whoops sorry I meant it’s an edited version with a diff ending - Nathan rewrote it to make it better, but that obvs also removes a bit of the ”this was pure AI” magic
The edited version diverges pretty early. Both versions make very little sense economically. This scenario - a very fast growth of money supply - is not that surreal, this has been tried many times, e.g. in Weimar Germany in 1923 and Zimbabwe in 2007 and the results are pretty consistent and well understood. The only fantastic assumption in this story is that the source of growth is a private bank account, and not a state treasury account. This assumption does not change the macro effects on the economy. The predictable macro effects are 1) a rapid hyperinflation where the price of all real assets is increasing in line with the money supply 2) leading to a loss of faith in the currency and replacement of the inflating currency with a stable one 3) with considerable frictional costs and winners/losers in the switch over. So, for example, the original version's claim that S&P500 price becomes one day undefined is not realistic - S&P500 will rise roughly in line with money supply, but a bit slower to reflect the frictional economy-wide inflation costs and the turbulence, then switch index denomination to euros/new dollars/BTC.
I think you're assuming the entity with the excessive amount of money is actually spending it, which didn't seem to be what was happening in the story.
Yeah, they say "S&P500 will rise roughly in line with money supply" but it's something like the money supply * the average velocity of money that's important (holding other things equal). So someone who gains a large amount of money and never spends it will have no effect
From the Substrate 'our-purpose' page:
"Over the past several years, we have invested in building out our supply chain and continue to increase our vertical integration, enabling us to move at speeds usually unimaginable in the semiconductor industry. More recently, we completed our first in-house production-quality 300 mm wafer lithography tool, which operates at the extreme G-forces required to meet the throughput of a leading-edge fab."
https://substrate.com/our-purpose
"extreme G-forces required to meet the throughput of a leading-edge fab" ?????
Note also that getting x-ray lithography to work isn't the problem. Getting x-ray lithography to work at scale (100+ wafers/hour) and at cost is the problem.
I searched for "lithography stage acceleration" and this was the first result:
https://www.asml.com/en/technology/lithography-principles/measuring-accuracy
> A lithography system (scanner) must work 24/7 with sub-nanometer precision, while accelerating mechatronic modules at incredible speeds. For example, the reticle stage accelerates at close to 16g and the wafer stage to 7g. That’s more acceleration than a jet fighter.
> It’s not possible to mechanically construct a machine capable of this level of alignment and precision, accelerating at those speeds, and with the level of reliability and repeatability required to make today’s computer chips without the help of in-scanner metrology.
I agree that ~10G of acceleration is impressive for those elements of a production lithography system, but it still feels hyperbolic to call it "extreme" for people used to high accelerations in other contexts (e.g. ejection seats, car crashes, piston engines, turbomachinery, centrifuges...).
Substrate almost halfway to their goal. All product creation can be broken down into two phases as described in this excellent TED-like presentation of a similar product: https://www.youtube.com/watch?v=DkGMY63FF3Q
The two phases are "ideation" and "development." Substrate is far along on the ideation phase and just has to hire some good engineers, which are plentiful. /s
Let's divide the work. You be labor, and I'll be management!
Ben Thompson had an interview with James Proud, Substrate founder (only paywall sorry)
https://stratechery.com/2025/an-interview-with-substrate-ceo-james-proud-about-building-a-disruptive-foundry-in-america/
Could it be that he fooled him too? Unitl now I had very high trust in Ben.
If you're not already suspicious, it's very easy to get taken in by this kind of thing, because why would you think a startup was actually a complete fraud (like Theranos) instead of the normal kind of hyped-up gamble that might or might not get anywhere?
That's what "due diligence" is for, and it doesn't mean reading the startup's shiny prospectus and taking their word for it. If it's just you investing your own money, sure, be as foolish as you want. But it is exceedingly foolish to invest $150 million in a startup without e.g. hiring an independent expert to accompany you on a tour of their factory, signing an NDA and reviewing the technical details of their proposal.
And if not all of that $150 million is your own money, then it's your fiduciary responsibility to do due diligence.
Since this happened, and Theranos happened and FTX happened, it's pretty clear that Silicon Valley has some rich fools and that selling silicon snake oil to rich fools can be a viable business plan. But everyone who falls for this sort of thing, should be embarrassed.
Ben Thompson is an analyst, not a reporter. In his interviews with startups he explicitly states that he is trying to get the founder's perspective/narrative, and usually won't be able to get hard evidence to verify the claims they make https://stratechery.com/2022/an-interview-with-rippling-founder-parker-conrad/
"One of the recurring questions over the years that I have written Stratechery is why I don’t write more about startups (a related question on Twitter is where is the Stratechery of startups). The answer is pretty straightforward: I am an analyst, not a reporter, and there is a lot more information publicly available about public companies; information about startups is much harder to come by, and often unverifiable when you get it.
At the same time, I do regularly come across companies that are very intriguing to me, particularly in the context of the themes that I write about, and I want to learn more — and share those learnings with you. That is the context for a new type of interview I am launching on Stratechery: startup founders.
The idea is this: I sit down with a founder with the goal of learning more about their business and the opportunity that they are pursuing. Instead of objective analysis this is as subjective as it gets — of course founders are optimistic about their future possibilities — but that feels like the right approach to expanding coverage to companies that I feel constrained in writing about. I’ve had a couple of these conversations already, and have found them tremendously interesting; I hope you feel the same."
Re: #27
Yeah. I'm pretty sure this only goes to the Reform/ Reconstructionist/Conservative/Open Orthodox. If you ccount the ultra-orthodox, who are fragmented and informal enough to be basically uncountable, you'd probably get very different numbers.
I don't have enough info on the study methodology, but I would suspect it might be missing the largest Modern Orthodox Rabbinical School, and only include those that accept LGBTQ/female students.
Seems a very shallow bit of analysis.
Good guess. From the study:
> Additionally, the number of survey responses from RIETS and other Orthodox institutions—aside from YCT and Maharat—was very small. Because these responses were both limited in number and substantially different from other denominational groups, they could not be reliably weighted to represent their broader populations. As a result, survey respondents ordained through or currently studying at RIETS or other Orthodox institutions outside of YCT and Maharat were excluded from the survey analysis.
So yeah, the "majority of American rabbinical students are now women" conclusion is probably false.
RIETS is the only rabbinical school that really matters in the American Modern Orthodox world, at least anecdotally. (Other relevant institutions are either affiliated or not actually rabbinical schools.)
So that's a pretty big misrepresentation there.
YCT does matter and YCT alumni actual run shuls and have been very positively impactful on Modern Orthodoxy (and the productive ones are generally are a bit more orthodox than YCT itself). Maharat drags YCT the institution down and is not modern orthodox
What's YCT’s class size compared to RIETS?
I can't find class sizes for RIETS or YCT, but RIETS is clearly bigger by a large margin.
‘doesn't matter,’ in this context, means that YCT isn't a representative sample of the direction of Modern Orthodoxy, as it is both small, and selected largely on the basis of being more modern/open than the RIETS student body.
The 'other institutions' I was thinking of in my previous reply are the Israeli gap-year yeshivot, YU, and the Gruss Kollel.
I completely forgot about YCT.
So, the Jewish version of "mainline liberal Protestant churches have lots of female clergy, also LGBT+ friendly"?
Not sure what the reference here is, but the sort of Jews going to daily services, (which only men have to do) will be praying in places with male Rabbis.
Hope that answers your question.
Sounds like it. People who take the religion seriously as a religion stick to the traditional version, people who take it as an 'ethnic identity/God is unknowable, we're spiritual/Vague Deity Shaped Symbol/doing the right *actions* together is more important than everybody *thinking* the same' go the liberal secular route.
Yes, exactly. Although the biggest losers in US strains of Judaism have been the Conservative movement, which seems to be stuck in the middle of the road and not optimized for either set of preferences.
Wow, talk about looking under the lamppost.
15. “so it’s not a contradiction that Democrats are against both “advocating for Israel” and “advocating for Palestinians” “
It’s not a contradiction at all. A pox on both their houses is probably the most common sentiment I hear. The position that Hamas is evil and what Israel is doing in Gaza is evil is the simplest.
I think is it more like: "Please don't spend one second thinking about this, I don't care. I want cheaper products".
Perhaps, but my view aligns with Hannes. A pox on both their houses. Everyone involved is awful*, everyone. Hamas, Israeli government, mostly the USA, Egypt, Qatar, the British and French governments past and present, the Turks/Ottomans, the UN, God, Allah, and everyone else who has had a hand in the situation.
Sometimes you can get a bit of an 'out' here by saying that the common people are fine, it's just those in power, but that's not true here: polling for both Israelis and Palestinians consistently show them to be ethnic purists, religious fanatics, and just general blood-and-soil for the blood-and-soil-God type people. The Israeli settlers are unambiguously extra awful and I struggle to understand those who defend them.
* OK, I guess Jordan gets a pass for being the least-evil entity involved.
Sure, but that doesn't seem to have impacted Jordan's role much?
The British were pretty non-evil, and so was the UN. When the Civil War started what was left of the British garrison, by that point not particularly more combat-ready than the insurgents, fought the Irgun out of Jaffa (where they were already starting to forcibly expel civilians) and repressed the Arab mobs.
Had the Arabs been smarter, they'd leveraged British patriotism and sympathy between Christians to get the British to stay there indefinitely.
It's pretty obvious from the context that I was reffering to their administration of Mandatory Palestine and their management of the crisis that was brewing. Keralans understandably had a different opinion, even Arabs after 1956, but at place and time we were discussing they were the closest to a reasonable actor
I would actually agree with you that post WW2 the Brits were ok, I admittedly had in mind more the original carving up of the Ottoman empire.
For a lot of people it's just not anything near a major concern. If you look at surveys of voters' top issues it's usually somewhere near or at the bottom. It's not even 'a pox on both their houses' as much as 'why are we investing so much effort and attention in a conflict halfway across the world when we have so many problems here?'
Yes, and that's why so many voters are getting sick of that war too!
Very few of the people with big blue and yellow flags outside their doors are sending kids to fight in it...
Ah yeah, absolute cynicism about geopolitics when it comes to Russia's actions, but also some random heartwrenching words should get Americans to stop investing in the conflict that has bled Russia dry, has forced their withdrawal from the Middle East, has denied them effective use of their bases in Crimea, and unified NATO like never before.
You can't have it both ways.
Ah yeah another fine geopolitical analysis based on categories that don't even exist (BRICS literally come from 90s financial punditry, it was supposed to be the countries where Western investors should look at for exhorbitant returns, Brazil or South Africa never had a coordinated strategy with Russia or China).
Russia is in such a better position after they lost most of their Soviet arsenal which are not able to replace, sure. They had to start scraping the barrel, exit from any other front, stop supporting their client state in the Middle East (which promptly collapsed and is now headed by a neoliberal Jihadist seeking to open trade with the US), and can't even safely sail in the Black Sea, such a bonanza.
As for "color revolutions", yeah, it's how the dumbest people with internet access call every protest that threaten their emotionally support dictatorships. Bc we all know regime propaganda is always so truthful and people are happy and loyal to the Dear Leader, the only possible reason we'd see protests is foreign involvement (and also America is the only country that would ever involve itself outside its borders, somehow). So poignant and insightful.
Nobody outside of Russia is advocating for more dead Ukrainians. Lots of people are advocating for Russia to get out of Ukraine and stop killing Ukrainians altogether.
And nobody but the Ukrainians gets any say in the question of how many dead Ukrainians might be a regrettable but acceptable price to achieve that. What the rest of us *do* get a say in, is if the Ukrainians are still going to be dying in the defense of their homeland, how many Russians will they take down with them?
I'm not the least bit tired of answering that question with "as many as possible, please and thank you, now let's see what other surplus weapons we might have available to send your way". And I don't think that I am alone with that.
#21. There are, at last count, nine candidates running for the Congressional seat. Perhaps three currently are serious contenders. Jami Floyd is one of the nine but (especially as she is not one of the serious contenders) describing her as "Bores' opponent" or the beneficiary of attacks on Bores is inaccurate and misleading.
Thanks, I've clarified this.
How vulnerable are the anti-mosquito drones to predation by cats?
And could cats be added to the President’s security detail? Bomb dogs and drone cats!
Do birds predate mosquitos? Get birds to eat any mosquitoes in the zone, get cats to keep the birds from flying out of the zone, get dogs to keep the cats in check!
I am still worried about the dogs in this scenario
Yeah, we don't need those. Replace them with catnip.
Who else is going to keep the cats on track? Any cat is likely to wander off, decide to hunt the birds for fun, or simply do the exact opposite of whatever it is you want them to do. You need loyal, trained, dogs who can stick at one task to supervise them.
Simply introduce lions to control the dogs.
Problem solved!
That's uncertain, the earliest mosquito found in amber dates to 125 million years ago in the early Cretaceous, as does the earliest bird having a beak and feathers.
Also, only kings can be kept in check, cats may be a single colour or have stripes or blotches.
"I was a little surprised that engineering shifts people right a little" - I wasn't; internships, no?
I don't understand this; what role do the internships have?
Exposure to and socialisation in the business world. At least that is what I was told back when I was a postdoc (in pure maths), as an explanation for different attitudes in engineering vs. mathematics/physics. What is more, people in engineering tend to be embedded in environments where they are exposed more to management-side views than labor-side views, especially from people they depend on.
Conservatism is no longer right-wing, and right-wingers are no longer conservatives.
Another way we might put this is that they're exposed more to the real world. Understanding of things like "how people respond to incentives" and "not everyone is like you in the way that they respond to incentives" and "no seriously, you have to pay attention to how people respond to incentives".
Yeah we all know that the right are the people who truly understand mechanism design and economics in general, as shown by their obsession for mean-testing, local autonomy no matter what, hard currency no matter what, macroeconomics as household accounting, and more recently protectionism, nativism and war on the most productive areas of the country.
Irony aside, a much more direct explanation is that when one realizes they'll soon be in top tax bracket, they want lower and less progressive taxes, and everything follows from it (there is something else, like a certain entitlement to high status and sexual access and frustration that money does not buy it, but I'm trying to be charitable here).
I wasn’t surprised, either. Not because of the internships, which I suspect are just as likely to expose students to high-COLA areas with liberal politics. Because of the relative affluence of a technical white-collar career. It’s likely to get people out of debt and into a house faster than teaching or whatever. I’d have been willing to bet on a strong correlation between political shift and median starting salary.
…But it turns out the study was normalized to pure science degrees, so engineering just didn’t shift people *as far* left.
This would also explain education: student teachers go to inner city schools, discover that some problems are not due to historical class oppression.
Then how do you know they vote conservative?
Fun fact, Richard Nixon proposed a negative income bill called the Family Assistance Plan, and one of the largest opponents were the social workers unions.
Can you link to some of those anonymous surveys?
How do they deal with selection effects? That Eng. people are more right wing to begin with?
Engineering has usually been thought of as relatively right wing in universities, behind only the business school and ag school (and sometimes parts of the law school and economics department, at a few universities). Also, during the early 2000s war on terror, it become clear that a lot of Al Qaeda terrorists we’re trained as engineers (as have a lot of Pakistani and Chinese government officials during their rightward moves).
In general I think engineering is for better or worse one of the few degrees that is truly apolitical. Some other are inherently political (say, any social science), some other are not but the current political climate forces one to take a side (say, biology or atmospheric physics). But no politician is going around claiming that research into high frequency sensors is murder, that Fourier transforms cause autism or that the Paris-Erdogan Law is a Chinese hoax.
So the most economical explanation is that 50% of the country is right-wing (and in the Muslim world, some percentage is Islamist), and engineering is the only one NOT forcing them to rethink their convictions.
I've read a lot of conservatives say having to deal with real-world consequences, where you can't BS your way out, drags you right. A building can't ask for accommodations, it's just going to fall down if it isn't built right.
Not sure if this is going to change with stuff like antivax becoming right-coded, but I've heard it in various places over the last three or so decades.
You'd need to be a particularly bad engineer or extremely unlucky (or, most likely, both) for a building you built nowadays to fall down. From what I see from my distant perch, the day-to-day work of an engineer often has more to do with fulfilling codes and specifications than with what the laws of physics will and will not actually allow.
I could believe that. I'm more passing it on than endorsing it, really--this is about why people think things.
And that's why in a biology lab everyone votes Ron Paul /s. The main difference between engineering and the sciences is that nobody goes around claiming that the Paris-Erdogan Law is a Chinese hoax or that using high frequency sensor for research is murder, so the ones who enter being right-wing can stay consevative by Gell-Mann amnesia.
21. "wait, multi-billion on one candidate? is that a typo?"
Yes. The Politico article they're quoting from says "million". https://subscriber.politicopro.com/article/2025/11/pro-ai-super-pac-targets-ny-democrat-alex-bores-00652148
Well that was anticlimactic.
I don't understand the intention of the writeup at link #1.
Setting aside the factual limitations of the stated premise -- is the essay intended as satire? Is the author trying to discredit the concept of US law assigning to corporations all the same legal rights that persons hold? Mocking the widely-held impression that US law actually does that? Something else?
Given the typical gross misunderstandings around corporate personhood and the rights appurtenant thereto, it is anyone's guess.
As an aside, I doubt that the ship mentioned in the article was treated as a person. Rather, it was probably an action in rem. https://www.srtflaw.com/in-rem-v-in-personam-in-maritime-law/
I interpreted it as "here are some interesting facts" and as not trying to persuade the reader of anything in particular.
But I would have titled it "Things that are Unexpectedly People". The Things themselves were not unexpected.
I think it's attempting to contribute to the conversation about whether AIs could/should be granted some of the rights we currently grant humans, by educating that some of those rights are already granted to other non-human entities. "Legal Persons" is a term-of-the-art that has a different meaning than the commonsense meaning of the word "person" which is one reason it's confusing, and this is a fun educational attempt to correct that.
As a concrete example, if a self-driving privately-owned car gets in a crash, it's not crazy to think that the _car itself_ should be represented by legal council and pay any damages from its own insurance, since trying to litigate whether it's the fault of the owner, manufacturer, or occupant at the time seems outside the domain of the car-crash investigation. Tennessee already has a law on the books to this effect. This is similar to how a ship can be arrested or corporation can be sued.
> Everyone knows AGI could be bad for labor
Re AGI / UBI being bad for labor, I just wrote an economics-grounded post arguing it's going to be great for labor and capitalists alike.
Labor, because we can give a generous $40k - $90k UBI to everyone, capital because returns to capital will at LEAST 2x, and arguably more, because now capital can be directly converted into whatever the highest value labor is.
Also, the fact that we'll be able to counterfeit high-value white collar work first is a "best of all possible worlds" scenario. To a pretty large extent, white collar job holders are capital holders, so they may lose their jobs, but the combo of capital value growth + UBI more than takes care of them, AND makes labor (non capital holders) better off, too.
Physical jobs have much lower margins for companies, contribute notably less to GDP, and would require substantially more hardware / capital expenditures to counterfeit. This is entirely ignoring the fact that robotics is much harder than software for numerous other reasons.
Having ended at a place where we can counterfeit only the highest margin jobs, that generate huge amounts of value both for GDP and the companies with those jobs, and in which the current job-holders are all predominantly capital owners who will largely be covered by company value growth in a much-higher GDP future? It’s so much easier on every front!
Post link here:
https://performativebafflement.substack.com/p/can-ubi-work-yes-with-a-few-reasonable?r=17hw9h
The economy has dectupled many times in history. At each of those times, we could have chosen to give people a generous (by the standards of the previous regime) UBI. I agree we will be able to do this again this time, but I don't know how likely I think it is. Yes, there will be many more unemployed people trying to get it. But labor will also have less power than ever. I think Philosophy Bear's argument - that we will either get a capitalist coup or a socialist revolution - is if anything too optimistic; more likely is that every industry argues for an AI ban or subsidies for themselves, the powerful ones end up as the equivalent of longshoremen, and the weak ones are left to fend for themselves for "every minority woman who lost her job to AI on a Friday in March gets a $300 voucher for health insurance during job retraining" style bandaids.
> And the idea that voters will be smart enough to prevent the robots-with-guns, or be able to control them democratically, is like a bad joke.
Agree, that's why being at a point where we're able to counterfeit only highly-comped jobs from people with capital and political relevance is a best of all worlds sort of scenario!
If instead, we could counterfeit a bunch of service jobs with robots? Everyone's doomed, boots on faces forever.
Or what about your scenario, where we have killbots and capital owners can retreat behind walled compounds patrolled by them? Boots on faces.
But here we are, with NO killbots, and right on the cusp of counterfeiting jobs, and the ONLY jobs we can counterfeit are highly valued white collar ones? It's a gigantic free lunch, societally, because we actually have a chance at navigating this gracefully in a way that doesn't end with boots on faces forever.
What stop the development of killbots. At max they are also 20 years away?
In my post I had graphs showing they are already counterfeiting jobs, since about 2022.
So, the job thing is happening today / tomorrow - well before killbots.
> Yes, there will be many more unemployed people trying to get it. But labor will also have less power than ever.
Seriously? The people whose jobs are being counterfeited are high income, connected white collar people. If even as much as 10% of them lost their jobs, you think we're telling them to pound sand, politically?
What if 20%, or 50% lose their jobs? It just seems really obvious to me that the public pressure towards UBI is going to be immense, and that it's much better that it's happening to richer and more connected people first this time, because it vastly increases the chances of substantial UBI actually getting pushed through.
I agree if it were reversed, and robotics were pushing 20% of service sector people out of jobs, they'd probably be screwed, and that's a much darker picture. But we live in that panglossian world where the reverse has happened, and it's going to be connected, well-heeled people whose voices are heard and payed attention to at the forefront of advocating for the massive societal rearrangement that UBI will represent.
The jobs of the well-connected white-collar people will be specifically banned from being replaced by AI. If an AI lawyer and a human lawyer both give you the same advice, but your conversation with a human lawyer is privilaged in court proceedings and your conversation with an AI lawyer is not, then the human lawyer will be far more in demand.
> The jobs of the well-connected white-collar people will be specifically banned from being replaced by AI.
Maybe, but I'm less sure of this. On the one hand, you have a bunch of lawyers and other white collar people slaving away in their cubes.
But on the other, you have a bunch of company owners and capital owners who want to pay a third as much as current comp while still getting all the value they represent - when it comes to "giant corporations and billionaires" versus "common schlubs," I generally think the billionaires get their way here, and we'll be able to counterfeit a lot of jobs.
And to navigate the billionaire side of this Scylla and Charbydis, to be able to UBI everybody at $40k+, all we have to do is keep current or 2018 corporate tax rates in place, which is the Schelling point already and a pretty reasonable expectation.
Sure, either side of those could go wrong, and there's some incentives pushing against the mainline outcome both times, but the way to bet is definitely that we both counterfeit a bunch of jobs, and keep a somewhat reasonable corporate tax rate.
And in that mainline scenario? I think we're fine - that it's great news, actually.
AI has seen heavy use in the legal field (along with outsourcing to India) because a huge amount of it is data management. Since a lot of law's financial structure is also a pyramid scheme where the higher often feel no real loyalty to the lower, your interpretation is 100% plausible. I will not be surprised if it doesn't happen (especially since regulators get a vote), but I will not be surprised if it does.
A lot of demand for lawyers doesn't care about which conversations are privileged in court. "Can I legally do XXX with my short term rental?" "What paperwork do I file for a divorce?" "I need to put together a will." "Is posting XXX slander?" "How do I legally fire an employee for XXX?"
"How do I legally fire an employee for XXX?" is absolutely the kind of thing that will cost you a multimillion-dollar judgement if it comes out in discovery.
I imagine the mom and pop asking this question aren’t at risk of losing millions of dollars because ten of their businesses aren’t worth that. The big business will probably continue keeping a human lawyer on retainer. Or maybe not even them, if AI chats have a reasonable chance of not showing up in discovery.
Your lawyer will be using AI. So lawyers will be able to handle many more clients at lower rates, but the lawyers who are less good at filtering out AI slop or less good at marketing will get weeded out. At first, anyway.
Sure thing. To my old lawyer, me switching to AI will look exactly like me switching to a lawyer heavily using AI. All of a sudden, I pay a fraction of the cost elsewhere, while a human becomes unemployed.
I think you assume every job will be replaced. I think this is not true, especially not at first.
Waiters, handyman, garbagemen, farmers, etc.
First will be replaced high-margin work that is amenable for AI automation.
That means that someone will have to do those jobs, which are not aspirational.
The means we convince people to work is by giving them money in exchange for their time, effort and risk.
In your opinion, how will this be done? If they get the equivalent of real 90K$ anyway, how would you convince anyone to do "dirty work"?
I mean, the agriculture revolution meant lots of people stopped being farmers, and became factory workers, and then Mcdonald's workers.
They couldn't sit at home doing nothing.
> I think you assume every job will be replaced. I think this is not true, especially not at first.
Actually in my post, I assumed ~50% of white collar jobs would ultimately get automated, and zero blue collar / service sector jobs.
And you're right! Non-white-collar people who hate their jobs will *definitely* quit and that's a good thing - comp will raise to match demand for the job until they're all filled, on the margins they will be filled by people who want to and are more motivated to do those jobs (many of them probably former white collar workers who want to keep earning), and consequently, quality-at-the-margins would probably increase in all jobs so shuffled.
And of course in actual practice, the UBI is probably going to be phased and targeted in stages rather than truly universal as we get to the final endstate my post was about.
Can you convince a guy that earns 100K a year from UBI to work dangerous construction job?
Can you have enough of those? Let's say you pay them a million. So they quit after a year or two.
The Gulf oil states have something like UBI (more like universal basic services + sinecures). Maybe gains from AI would be more like an oil windfall than normal economic growth because:
1) It happens suddenly before standards can adapt.
2) It doesn't require anyone to work to produce the gains so no one looks more or less deserving.
Jay Gould supposedly said that he could pay half the working class to kill the other half; sadly, AI is probably coming for the killing-the-working-class jobs too.
> we can give a generous $40k - $90k UBI to everyone, capital because returns to capital will at LEAST 2x, and arguably more
I'm very resistant to "giving everybody free money will actually make everything great" arguments. Too good to be true. The real world is a morass of tradeoffs.
Sure, in my post I posit that ~80% of people on UBI will use it to stare at screens and shovel junk food into their faces even MORE than the ~11 hours they spend per day doing those things today.
Still a big win!
Why keep it a secret?
The purpose of free school lunches is to make people fat? Sounds interesting. Any serious source for this?
I read all that. The problem, as with all such prognostications on UBI, is that you ignore the transition to the glorious future and the intermediate demand side of the economy. You assume that capital can extract more value from fewer workers, profits will increase, and business can be taxed at present (or perhaps higher) levels of corporate tax. Hence the $95k.
However business revenue depends on demand. If the 60% of the economy is tanking there’s no profit to tax. The UBI has to come first. There’s no plan for that.
This will happen fairly quickly. Polymarket is predicting a recession next year (https://x.com/Polymarket/status/1997751003559412101) and - while i don’t see their reasoning there - AI will either be responsible in part, or in large part, and be blamed either way. The political response will be … interesting.
> you ignore the transition to the glorious future and the intermediate demand side of the economy
Yeah, I don’t think we’re too far apart there - in the post I elide this by just looking at the end state, but agree that transition can go better and worse, and is probably bumpy.
At the end state, obviously the UBI to everyone props up the demand, but in the middle there can be noise and dips and surges and who knows what else, I agree.
From the company point of view, it’s probably fine. Even if 10% of people in their prospective customer base have lost their jobs, being able to offer much cheaper finished goods or services will induce demand higher than that 10% deficit. And of course, any companies that can service overseas or non-white-collar demand are safe and pure growth at that point.
And from the macro perspective, if those 10% of people losing their jobs spurs a recession, so much the better, right? It puts it on the political radar sooner and puts some real emphasis behind it in terms of being able to successfully push for the UBI, and starts the process at a relatively smaller impact point, hopefully making the rest of the job counterfeiting smoother and less bumpy.
It looks like Polymarket is predicting that the odds of a recession by 2026 are at an all time low and have been declining? https://polymarket.com/event/us-recession-by-end-of-2026
Their twitter is basically just a ragebait for advertising purposes.
Hmm. I was misinformed. Admittedly by themselves.
In a world where the vast majority of power and money is concentrated in the hands of the tiny number of people, what makes you think they're going to want to pay you for doing nothing?
Your best case scenario is that they'll want to pay you for doing something. Human servants surely have some kind of je ne sais quoi even if robots are available. If I were a quadrillionaire in a world full of otherwise-starving people I'd probably find a way to pay millions of people to sing my praises all day long while dancing in perfect formation.
> In a world where the vast majority of power and money is concentrated in the hands of the tiny number of people, what makes you think they're going to want to pay you for doing nothing?
I posit that we already live in that world, and already have corporate taxes that are paid!
Just a quick hand wave in that direction - NVIDIA is worth $5T, Apple 4T, Alphabet is $4T, MSFT is 3.5T, Meta is 1.5T - we're at 18T out of the US's 30T economy. Yes, market cap is fixed and GDP is flow, I said it was a handwave. My point is the power and money is ALREADY concentrated in the hands of a tiny number of legally defined "people" and things are fine.
The whole point of my post was that we don't need to EXPECT anything but minor amounts of basically "charity," and that today's or 2018's existing corporate tax rates are enough for everyone to plausibly be okay in the future in terms of generous UBI, and those rates are both pretty strong Schelling points and pretty reasonable amounts, especially if the alternative is pitchfork wielding mobs in a democracy.
> If I were a quadrillionaire in a world full of otherwise-starving people I'd probably find a way to pay millions of people to sing my praises all day long while dancing in perfect formation.
Such is how the content creator economy currently operates, naturally springing up from the demands of an ever enriched professional class wanting their unique interests covered. I think people are overestimating how much the economy will become concentrated as things derived from human labor become ever more precious status symbols. There is no reason to go see a play in person compared to a movie, but the practice persists because people are willing to pay big money for a unique experience. And few people call theater acting a debased dystopian occupation.
Aphantasia question: for questions like "Suppose you walk forwards 20 feet, then turn left and walk 10 feet, then turn 180 degrees and walk forwards 20 feet. What letter of the alphabet have you traced?" - how do aphantasic people answer these questions mentally? I searched a bit online but didn't find any of the answers cleared up my confusion.
I'm mostly aphantasic. With this task I'm doing something "visio-spatial" in my mind and don't find it super difficult. It's easier if I imagine physical movements or trace it out with my finger.
I definitely don't "see" a "T" anywhere, though.
What's the difference between a visio-spatial mental operation, and a phantasic one? It sounds sort of like they ought to be synonymous.
Whenever I hear about aphantasia, I can't help but think "this is some sort of error in communication—like, people interpret 'mental image' as 'seeing something in front of your eyes as in a hallucination', and then think 'well, that definitely doesn't happen... I guess I can't do mental imagery!'"... but probably I'm wrong about this, since it appears to be a well-validated(?) "condition".
I'll try to explain what's happening. There's a blank three-dimensional space in my mind, I can use this to rotate shapes, orient myself in a new town, or remember how to get around my childhood home. I can move around this space, but there is no visual information being stored or accessed - no colour, no brightness or texture. It's only visio-spatial to the degree that I can "imagine" movement.
If you asked me to imagine my mother, it's impossible. If you ask me to imagine a sunset, I can trace a horizon line in this 3d mental space that seems vaguely sunsetty, but I can't maintain it. If you ask me to imagine an apple, I automatically imagine the idea of holding it in my hand and biting into it - I can imagine the crispness and the juiciness. But there's nothing visual.
Fascinating—thanks for explaining! That definitely sounds different from my experience (if anything, color & texture are, I find, *easier* & more vivid than spatial relations/movement); I guess my skepticism was, indeed—...as predicted, in my defense! [cough]—ill-founded.
I would like to know how it is possible to draw anything from imagination, then. Are you capable of deciding beforehand what you wish to draw, including details, then drawing it?
Not exactly sure if this is what you mean, but the question I think you're asking is: "To what extent can I "pre-imagine" something I'm hoping to draw?"
Not very much! I can imagine non-visual scenes in my head (in language or conceptual form), and then draw them - I think this counts as "drawing from imagination". But I can't by any means hold something visual stably in my mind and then draw it. This seems like a wild superpower.
For what it's worth, I'm a very improvisational painter/drawer - I love starting on a blank canvas without any plans and seeing what patterns emerge. I don't know how much this is related to aphantasia, though, because I have a similar approach to music and writing, I think.
I dunno about TT, but for myself the answer is simply that it's impossible for me to draw something from imagination, at least if it's supposed to be realistic imagery (vs, say, a diagram or something.) When I used to watch cop shows, before I knew about aphantasia, it was always deeply mysterious to me how a sketch artist could possibly use some kind of black magic to elicit a picture of a suspect from a witness. Now I understand: it works because the witnesses have mental imagery. It wouldn't work if I was the witness, there's no magic trick the sketch artist is doing.
Haha, I had exactly the same confusion with the police show/sketch artist thing!
But, reading up a bit on this topic, it seems that "normal" witnesses aren't very good at providing accurate images either. I guess the best examples (which are very impressive!) are a rare combination of brilliant police sketch artists eliciting skilfully from witnesses with an unusually accurate visual imagination.
In fact, this article (using facial composite software) (https://www.sciencedirect.com/science/article/pii/S0010945223001764#bib39) suggests that aphantasics might not be disadvantaged, maybe because they're less likely to generate false visual memories, and they're able to make sensible spatial judgements.
I hadn't thought of it like this before, actually - a strong visual imagination adds a bunch of noise/error that aphantasics don't experience!
Without moving your head, put your hand into a rear pocket, or put down an object behind you and then pick it up again. I am guessing that you can do this without using visual imagination. I think I am (as an aphantasic) using the same spatial proprioceptive capability to perform hypothetical tasks like the one you suggest.
What being aphantasic means, is thinking about spatial concepts without any of the visual associations. Here's an analogy: suppose you watch a video of someone playing a musical instrument, with the sound off. If you know the melody, you can probably watch them playing and guess where they are in the piece, and imagine what it sounds like .
Now suppose instead the instrument is a strange one to you. You have no idea what it sounds like. All you have is that by the timing you can work out where in the piece they are, and that in the abstract they are going up now and down now, but without the rich associations of timbre and colour that you have for an instrument you've actually heard.
Now, aphantasics aren't blind and do have memories of visual objects. They just don't "come along with" thinking about spatial problems. If I think about the guy walking in a T shape, I know that when he has turned 180 degrees he's going to pass the top of the T again and I can hold these spatial relationships in my mind, but they don't have other visual associations, such as colour or brightness or texture or shape. I know they don't, precisely because I do experience those in the real world and so can tell that they are not present.
For me, it's clear it's not about an error in communication because of how well my visual imagination works when I'm dreaming.
I sometimes can hold onto the ability when sort-of lucid dreaming, and I've even managed to consciously experiment with imagining an apple in this case. The apple is full colour; I can see details and texture.
Then I wake up fully and my ability to imagine an apple in my mind decreases considerably. No colour, no details, just a hint of apple.
Spatial reasoning isn’t all visual - as others mention, we have a lot of spatial sense associated with touch, proprioception, sound, and other senses too. I think I’m pretty normal at visual imagery (though bad at human faces - I can’t identify what hair color or eye color most of my friends have by imagining looking at them) but I’m quite bad at imagining smells or tastes that I’m not having right now. Still, when I’m reading a menu I know facts about which tastes I like, even if I can’t envision them the way I can envision images.
Oh, yeah, I have the same issue with imagining tastes! I must be able to do it *somehow,* to *some* extent—I know which things I want to taste again, and not only in a "I remember I liked that" way, but in a specific-craving kinda way (as you also allude to)—but if I try to close my eyes & summon up a taste, it's just... impossible. It *feels* as if I'm doing it, briefly, but it's like an (extra-)insubstantial ghost—the more I try to drill down & really "taste" the imagined item, the more I find that I am not imagining anything at all... if that makes any sense, heh.
>but probably I'm wrong about this, since it appears to be a well-validated(?) "condition".
The validation seems to be entirely self-reports:
>Before accepting the existence of such wide variability in the imagery of normal people, we should ask whether those with self-reported high and low imagery powers differ significantly in their success on cognitive tasks that are plausibly aided by the use of visual imagery. In this vein, James Angell (1910), discussing the imagery literature of the time, stresses the importance of looking for correlations between what he calls “objective methods” of measuring imagery, in which success or failure on a task depends on the nature of a subject’s imagery, and “subjective methods” of self-report.
>In the past hundred years, many researchers have, as Angell advised, compared subjective and objective measures of visual imagery. The results are discouraging. Research has focused mainly on three subjective measures, more readily quantifiable descendants of Galton’s original questionnaire: Betts’s (1909) Questionnaire upon Mental Imagery (and Sheehan’s [1967] shortened version of that questionnaire), Gordon’s (1949) Test of Visual Imagery Control, and Marks’s (1973) Vividness of Visual Imagery Questionnaire (VVIQ). The early returns were very bad. Through the 1970s, attempts to correlate these subjective measures with anything objective failed so regularly that most prominent reviewers denied the existence of a relationship between objective and subjective measures of imagery (e.g., Ernest 1977; J. Richardson 1980).
>More recent reviews have been slightly more optimistic. McKelvie finds strong relationships between the VVIQ and tests of hypnotic suggestibility, tests involving the Gestalt completion of incomplete figures (e.g., the speed at which someone would recognize a fragmentary stimulus as a “circle”), and tests of motor and physiological control; he finds spotty relationships between the VVIQ and tests of visual memory; and he finds no relationship between the VVIQ and tests of visual creativity (for people of normal I.Q.[7]) or tests of ability at spatial transformation or “mental rotation”.
https://faculty.ucr.edu/%7Eeschwitz/SchwitzPapers/Perplexities-Ch3-Imagery-091105.htm
Then you are not aphantasic. Being able to "see" the T is not the norm but a form of rare controlled hallucination or extreme hyperphantasia.
I can't "see" the T, but I can imagine it... see it in my mind's eye, so to speak. I believe I have "normal" phantasia but I suppose I could be wrong.
Oooh, cite for that? I can "see" the T quite clearly!
As if it was indistinguishable from a T on a page? As in, if you didn't know the difference, you wouldn't be able to tell if you were looking at a picture or imagining it? I think that's quite a rare form of hallucination but i don't have a citation.
No, there's no way I could get confused, it's like I have a piece of paper in my mind and I can draw on it. It's weirdly more like remembering looking at an image than actually looking at an image, except that I can modify it.
I can well understand why there was such a debate about whether mental images exist or whether it's just a metaphor, although I'd definitely have been on the 'of course they do' side.
I think it must be vision-related, because if I need to draw something detailed it helps to close my eyes so there's no interference, but if it's something simple I don't need to.
For instance, in answer to Galton's famous question I can imagine a tiger and it's there in colour with a forest background, but if you ask me to count the stripes I can't, unless I deliberately imagine the stripes in detail first.
In this case it was actually a glowing red T on a black background. I can remember it from yesterday as if I'd actually seen that on a computer display. And given enough time it wouldn't surprise me if I got confused enough to not be able to remember whether I'd actually seen it or just imagined it.
I will say, as a phantastic person, this just makes me picture a location I'm walking on from my normal eyeline, and then when the question finally lands I have to throw all that out and restart with an overhead view.
I wonder how much variation there is in how people imagine this, "by default"—my instinctive/immediate "choice" of view was something like "a bit above, behind, and to the left of my head at the starting-point" (then the "camera" stays there & mental-me walks out as described, although IIRC I did sort of the same "path-following instead of fixed-camera" "error", at first).
(boy, lotta "dubiosity quotation-marks" in here–)
I interviewed a man with aphantasia once. He solved the problems the way you would if you were writing a computer program to work out the answer or making a mathematical proof. He treated them as logic puzzles. In this example, he would have reasoned that there were only two lines at right angles, so it is either L or T, and then another step of logic (that the line that goes back on itself goes further than where it started from) rules out L and concludes T. No visualization, just reasoning about angles and distances.
Mostly aphantasiac here... just did it by moving my hand along with the instructions.
I'm severely but not completely aphantasic. I was able to follow your question by drawing a very bleak and washed out thick line in my mind's eye, it took two tries because I needed to go back and calculate explicitly that the left and right arms are each twice shorter than the forwar arm. I don't think of aphantasia as inability to imagine literally anything superimposed over my visual field, it's more about extreme poverty of detail. I can visualize any letter of the alphabet, but not a word of 3 letters or more, except by progressively focusing on one- or two-letter chunks in it. I can see something in a very faded color, but not two colors at once. I can bring to mind a few faces of closest family members, but they're like snapshots I recognize but can't put my finger on any significant detail.
I'm only partly aphantasic, but I fully agree with the Torches Together response above. I have no problem imagining form and movement, it's just not quite visual, more prioceptal. There's absolutely no colour, no texture, no background, no accidental details that are not required to visualise the object of the imagination. I had no issue whatsoever answering your letter question, but in the process of doing so I realised that I made tiny movements with my head. I'm also pretty good at remembering maps, and it's not visual either.
Perhaps it would be useful if I mention that I believe the vast majority of my dreams are nonvisual. There's action, but the scene is not imagined visually. This is in contrast with audio stimuli -- when I wake up I often remember the last spoken sentence exactly, including the tone of the voice, modulation, etc. Only once recently did I wake up remembering an image (a face, and a patterned scarf), and was very surprised.
You already got quite a few replies, but here's another. I'm not totally aphantasic but quite severe. It is hard to explain the process to arrive to T but I would say it's more tactile, I may not have the technical knowledge to explain it. I have above average spatial imagination and solving that kind of puzzle I've always found great fun. I am also a visual artist, lots of artists are aphantasic. Sometimes I wonder if this is a way for us to create the visual imagery since we can't picture it in our mind. I've now picked up sculpture and I am blown away with the facility I have for it, and how enjoyable I find it. Line based drawing has always been harder for me but just building shapes in space is apparently second nature, who knew.
I'm also quite skeptical of people who claim high phantasia. With little practice they should become great artists (and some geniuses do but they're exceedingly rare), but they can't for their life put any of the supposed detail down even if clumsily. My guess would be some sort of fooling ourselves like we do with dreams. The details aren't actually there. Btw yes my dreams are also visually very lackluster, although not spatially.
Hypothetically, what kind of guarantees do those embryonic selection companies have? If I use one and 10+ years later my kid has an issue they claim she wouldn't get, or is not as smart as they claim she would be, then what sort of recompense do I get?
I mean, this isn't like fucking up a car or a boat, this is my child. I'm going to want more than money to make amends. Will these companies even be in business in 10+ years?
There is no guarantee because this is all probabilistic - they may say that there's a 10% lower chance your child will get problem X, but that still means 90% of the kids who would have gotten it anyway will still get it.
The people who believe it believe it because they've demonstrated that their technology predicts things correctly in existing adults. This is pretty strong evidence, but see https://www.astralcodexten.com/p/suddenly-trait-based-embryo-selection for potential complications.
I get that it's probabilistic. Still, they're offering a service/product with many claims; there must be some sort of warranty or guarantee of satisfaction to some extent, especially if I'm paying upwards of $50k for this. Maybe they accept returns.
>there must be some sort of warranty or guarantee of satisfaction to some extent
*Must* there be? I can imagine a model like that for the claims on e.g. boxes of oatmeal: "reduces risk of cardiovascular problems!* [...] *studies show that eating a diet rich in whole grain [...]"—i.e., the claim rests upon some studies, and is offered with no guarantees.
Not exactly analogous products or claims, of course, so I might be wrong—maybe there *is* a "guaranteed effective or your money back!" somewhere on there.
If I can get a 5-year limited guarantee on a $15 box of Christmas lights, then surely I can get something analogous for a $50k embryo. Eh, maybe they'll be a class action suit 18+ years from now.
Plaintiff: Your company promised that my daughter would have an IQ high enough to get into MIT. She didn't get in even with perfect SAT/ACT/ETC scores.
Defendant: There are too many confounding factors to find fault with our methods. Not everyone with perfect scores gets into MIT. It's a multifactorial selection process. Besides, between the latest batches of 200+ IQ babies and our AI overlords who are considering closing such inefficient methods of learning and scientific discovery, we think that you should be pleased with the way your daughter turned out. Might we suggest she consider applying to "The Farm". We'd be happy to share information on this life alternative.
The Christmas lights are at a level of reliability where 99%+ of them are still working in 5 years. Embryo selection is at a level of reliability for predicting traits that you need advanced statistics to measure the difference.
They're pretty clear as to how their predictions work and the decent probability that they would do better. They're not measuring against a perfect super-baby, but against the average baby you would get from a randomly selected embryo, and even then there's a decent (although lower than random) chance the baby is worse than average.
They might give guarantees if the technology is good enough, which won't be for a long while.
On a serious note, why can't companies like this offer any sort of customer satisfaction promise? Sure it's not Christmas lights or a car, but it is a service/product for sale. It may be complicated to measure, but not impossible. I wonder if there's any analogous examples out there.
On a less serious note:
Maybe they could keep spares for replacements in the event that something goes wrong. IIRC the latest date for splitting off cells for identical twins is like two weeks. I'm not certain when the companies sample, but the possibility exists of them making and storing "copies". Also, with the eventual advent of artificial wombs this opens up a lot of "human trials" for better or worse.
Dissatisfied customer: I'm not happy with the way my son turned out! He's a shiftless layabout. I demand a refund.
Customer service: We are unable to offer a refund, but we have the embryo that is your son's identical twin brother. Would you like to try again and provide a more rigorous environment for the twin? [Insert joke about personality being genetically determined].
In most medical procedures the guarantee is that the doctor will perform the procedure, not that it will necessarily have an optimal outcome.
This would be like that. They would be liable if either (a) they gave you unrealistic expectations about the results of the procedure, or (b) they didn't do the procedure that they promised.
Well, AFAICT, the service/product they offer is a particular genetic sequence, which is easy enough to verify, and if it doesn't check out then I imagine there are penalties stipulated.
My best guess regarding Substrate is that it might be like a Ponzi scheme with Thiel and Trump not so much being suckered as being at the top of the pyramid helping to sucker everyone else. Once Thiel and Co. have invested and the government has endorsed it, it's a lot easier to draw in lower information investors.
This makes the most sense. If you can convince one investor that your product is good, and Thiel trusts that investor, and Trump trusts Thiel, then of course none of them will double check. It helps that they're all also very stupid people.
You think Peter Theil is a very stupid person?
Yes.
Ok. He won a California statewide mathematics competition in middle school, completed 2 degrees at Stanford, was editor-in-chief of the school paper, and has had a hyper-successful career as both an entrepreneur and an investor in which he's been repeatedly early to major technology sea-changes. Oh and he's a National Master in chess. If Peter Thiel is stupid, then either intelligence is irrelevant to real-world outcomes, or you don’t know what intelligence is.
He can be a successful investor, solve math puzzles like a little angel, win at chess, and still not see the forest from the trees.
Someone rambling about the antichrist while sucking up to Trump is not really defying the label.
Of course there's a lot of cultural and political aspect to evaluating someone's decision making. (It seems he is pretty good at his antisocial survival-of-the-fittest competitive strategy - he commands more than 26 billion dollars after all, but that still seems a bit stupid compared to prosocial cooperative strategies.)
>still not see the forest from the trees.
Yeah that's called wisdom, not intelligence. Lots of high-IQ people are notorious for being a little spacey and nonpragmatic.
Weird beliefs are common among the intelligent. Lotsa professors buy into the nonsense that gender is nonbinary despite there clearly being only two human sex gametes, or that Marxism is a good idea. What you're really saying here is that you personally dislike his worldview. Which is fine but the word for that isn't 'stupid'.
And just so you know, he's not some Christian wingnut who thinks the antichrist is a real person. He understands religion in a way similar to Rene Girard and uses 'antichrist' as something which connotes something like a social attractor point - an abstract pattern in large-scale social dynamics, not unlike how Scott uses Moloch.
First of all “intelligence” is nonlinear so he can be smart in some areas and stupid in others, so it's already a fail to have a serious discussion about what “stupid” is. it's just a directionally true insult.
But more importantly if he believes what he publicly says he believes then he must be in some sense stupid, yes. I can't know what he believes in private. No matter how successful you are, if your values and goals are selfish and antisocial then you are stupid on a deep and fundamental level, arguably even more than an unsuccessful person who never hurt or took advantage of anyone.
>so he can be smart in some areas and stupid in others
That's mostly false. Psychometrics demonstrate that all analytic cognitive capacities are highly correlated, hence g. Sure some people are better at math or verbal, but to first order they're correlated.
>But more importantly if he believes what he publicly says he believes then he must be in some sense stupid, yes.
I don't know specifically what beliefs you're referring to here, but it sounds to me like you're conflating intelligence with wisdom. Those things definitely don't (necessarily) go together. Just look at someone like Noam Chomsky. Undeniably brilliant and also believes just absurdly buffoonish things. Also woke originated among the universities and professors have some of the highest rates of pro-Marxist thought. High IQ people are notorious for being able to talk themselves into shockingly wrong ideas. That doesn't make them "stupid". It means they should stick to academia and not try to influence real world politics.
What you're saying is that you disagree with Theil's politics and don't like him. That's not related to his intelligence. You're just using a childish insult to designate your outgroup.
I don't think Thiel is stupid *at all*, but he might not care about ripping people off if he calculates he'll still make money at the end or it fits one of his larger goals.
If he is anti-social and selfish then his values and goals are stupid even if his methods are not, but as I said in another comment it's not very productive to try and accurately define a basic insult.
Why would Peter Thiel run a ponzi scheme when he already has billions of dollars in legitimate wealth from founding and investing in successful companies? It's just not plausible.
Because it's also Trump's ponzi scheme. Some things are more important than money. It's good to be on good terms with the leader of the most powerful country in the world.
"Every time you take a Waymo, you’re letting an AI control a critical system; every time it chooses to stop at a red light but not a green one, it’s making an “important decision” (if you don’t think this decision is important, consider the consequences of failure)."
Author here. I don't know if we actually disagree--I wouldn't count Waymo decisions as important enough, because only small numbers of lives hang in the balance. What we said was:
"We all believe that current AIs should not be allowed to have autonomous control over critical systems. This includes extreme cases like giving AIs control over data centers, nuclear weapons, tech companies, or government decision making processes."
I can't speak for the AI as Normal Technology / AI Snake Oil authors, but we at AI Futures Project are supportive of self-driving cars and similar tech, and were mostly warning against the much more dangerous things given in the examples (e.g. data centers). Perhaps we should have clarified.
I understand that you in fact support Waymo. I just think Waymo support generalizes to allowing AIs to make many important decisions. What I have in mind here is things like the Colorado act which make it hard for companies or organizations to outsource business decisions to AI; I'm worried that as framed it seems to potentially provide support for that. I acknowledge that you specified some extremely bad things like nuclear weapons, and some things which you might have intended to limit to very major cases like government decisions, but I thought the exact phrasing:
> "We all believe that current AIs should not be allowed to have autonomous control over critical systems. This includes extreme cases like..."
...suggested a broader picture than those most extreme cases. Probably this was you guys trying to come up with some framing that covered both the AIANT's broad ban request and your narrow ban request, but I still disagree with my interpretation of the compromise.
I think we could have been clearer about what we mean by critical systems. I think probably (but I'm not sure) AIFP authors have a much higher bar for what should count as critical, than AIANT authors. And reasonable people can disagree about where the bar should be. Our statement definitely rules out the current stated plans of the major AGI companies though -- they are planning to have AIs autonomously running AI R&D programs (and indeed, Sama said he's aiming for autonomous AI company CEO iirc). So it's not a trivial statement.
Hmm. Ruling out "AIs autonomously running AI R&D programs" seems to be assuming the conclusion. If AI is Normal Tech then surely this use is not a critical system any more than PERT charts would be.
I mean, I'm mostly on your side here, but I think this is a sign that Point 7 is something you have managed to "agree on" by making it vague enough that nobody could disagree.
I'm not sure but I got the impression that the AIANT authors agreed it was a critical system -- after all they agreed to the wording quoted above, which lists datacenters and tech companies.
And again, some people do disagree--most notably, the leaders of the tech companies themselves disagree, and are explicitly planning to do the thing we're warning against.
“ I just think Waymo support generalizes to allowing AIs to make many important decisions.”
Scott, would you ever expand this statement to something like “using AI now generalizes to supporting AI development to proceed apace”?
Or to put it another way, “people who are serious about AI safety should stop using AI and advocate for abstention from the use of AI until alignment is achieved to our satisfaction”?
If not, why not?
It also seems to me that there are relevant distinctions about *how much* control of critical systems is AI. A power grid that is moment to moment supervised by AI but with human checks at regular intervals and with human ability to look in whenever they want and with constant human evaluation of the systems sounds different from a system with full AI autonomy.
Sure. It's a fuzzy boundary, not a crisp one, and ultimately what we care about is a cost-benefit analysis of the harms vs. benefits.
#15 and #33
It's funny that the two things Democrats hate the most are billionaires and podcasters, yet Newsom seems to be the frontrunner.
I know he's not actually a billionaire, but I'd consider him billionaire adjacent. At the very least he should get dinged -9 for being wealthy.
He doesn't seem to be all that wealthy. Besides, his job for almost 30 years has been politician. Very different than someone like Trump or Tom Steyer jumping into politics.
Also, warmth towards Newsome is only side of that poll result. JD Vance has lousy national like/dislike ratings in polls, and being VP is not a promising situation in which to improve that. Only one sitting VPOTUS has won a presidential election in the last 190 years, GW Bush in 1988.
Vance is young for a national politician and certainly could end up on a Nixon/Biden path towards winning the White House later. Or, Trump's physical/mental fade could terminate before 2028 and Vance would then be running as an incumbent POTUS. Right now though any matchup being polled on is "Democrat X versus the current VP", and being the current VP is almost never a path to national public affection.
>Only one sitting VPOTUS has won a presidential election in the last 190 years, GW Bush in 1988.
Al Gore probably counts, too.
Close but no cigar, Al.
Vance has plenty of time to pivot. I'd stay agnostic for now.
Nixon had an inexhaustible reservoir of spite that propelled him from Eisenhower’s VP to becoming the most entertaining president ever. I’m not sure JD Vance has that kind of fire in the belly.
At this stage being the frontrunner isn't very predictive, because it's almost purely a measure of name recognition.
And "Democrats hate podcasters" is not a stereotype I've encountered before.
Yea that's a new one. Damn near all the people I hang out with deep in the heart of blue America love podcasts and view podcasters as celebrities. They happily pay ridiculous ticket prices to go to theaters and see people like the Pod Save America hosts chatter onstage.
The most salient examples of podcasters are people like Joe Rogan. Their whole vibe wouldn't fit with a lot of Democrats, I suspect.
There is a somewhat widespread belief that Joe Rogan swayed a lot of young male voters into voting for Trump. I know people personally who think this, though their counterfactual, voting for Harris, I believe is much less likely than them simply not voting at all. That the Democratic party fundamentally has nothing to offer young men, and in fact regularly demonizes them for the crime of existence, often seems entirely immaterial to some of them.
Sure. I don't think that is why Dems might not like podcasters.
I think the issue isn’t “democrats hate podcasters” - it’s that when you ask random democrats to imagine “a podcaster running for congress” they’re going to picture some manosphere sports influencer or something, not one of the Pod Save America people.
Even if you imagine one of the Pod Save America people (about whom I know nothing) it's still pretty reasonable to say no, I don't want one of them.
My impression of most popular podcasters and youtubers is that they're mostly just a bunch of randoms who pass some basic microphone-skill filter and who got lucky with the algorithm at the right time and invested in making their shows a bit better. It's reasonable that people don't want to jump out of their seats to give these people even more power.
Democrats hate podcasters? Why? Because of Joe Rogan?
Re: 31, the proper counterfactual is "how popular would the Dems be on COVID issues if both the Dems and the GOP had fewer own goals during that time", and I think in that universe the Dems would be in the lead. (They had more own goals, it seems to me, in the sense of actions that didn't make any sense given their policy principles.)
> One million dead Americans is negligible. That's, what, 1 in 300? Know how many people committed suicide because of lost jobs, livelihoods, and businesses?
There are fewer than 50,000 suicides per year, and the suicide rate was _lower_ in 2020 than in years before and after. You do know you can just look things up, right?
One thing that's really hard to untangle is the way partisanship can have a profound effect on what steps the parties take. It's not hard to imagine Trump behaving differently in 2020 (i.e. really strong lockdown policies), and Democrats calling it tyranny.
In 2014, for example, Republicans excoriated Obama on Ebola and Democrats downplayed it.
This actually happened. When Trump cut off flights from china the Democrats screamed racism. When the COVID vaccine first came out and Trump was talking credit, the Democrats were the ones casting doubt on its safety.
So much we forget about Covid. I remember when masks were useless.
No. You're remembering an oversimplification that became popular, not what the reports said.
FWIW, masks are relatively useless at preventing you from catching COVID, and relatively effective at preventing you from spreading it. I think this is true even for the masks just made out of one layer of gauze, but I've seen no reports of that being tested. They slow down the air on exhalations, causing the germs to remain close to the body. This, of course, only applies to masks that are worn properly, rather than, e.g., on one's chin, as I saw some people wearing them. And it depends on the air circulation in the environment. (Vertical air circulation makes it much more effective, horizontal less so.) And there are lots of other detailed caveats.
Not really. The WHO and plenty of governments said not to wear masks for the first few weeks.
They said two things - there’s “no evidence” that masks help random people prevent disease, and there’s a mask shortage for medical professionals. I really hate the “no evidence” sort of line, because the evidence hasn’t changed all that much, but there’s plenty of good reason to think a mask will help a bit, and plenty of reason it won’t be anywhere near as effective as just interacting outside.
"When the COVID vaccine first came out and Trump was talking credit, the Democrats were the ones casting doubt on its safety."
I think that's wrong. *Before* the vaccine was released some Democrats cast doubts on the safety of a future vaccine whose approval might be politically driven. But when the actual vaccine came out, no Democrats doubted it because there were no indications that the approval was politically driven.
You mean after the election none of them doubted it. Convenient how the timing works on that one.
Yes, but that argument goes both ways. Before the election Trump had a huge incentive to push approval even if it wasn't safe. Afterwards his incentives were to decide on the merits.
The success of the first covid vaccine trial was announced by Pfizer 36 hours after Biden was declared the winner of the election by the Associated Press. So, by then the vaccine was irrelevant to the outcome of the election.
The timing is of interest because in late October 2020 Pfizer shut down processing of samples from its clinical trial until the day after the election. But few have ever shown much interest in that curious change of plans.
Trump didn’t cut off flights from China. Demand fell pretty dramatically during the pandemic, and the Trump Administration may have contributed to that with the State Department issuing a travel advisory and Trump instituting travel restrictions. Due to lack of demand (and possibly employee safety concerns) many airlines suspended service between the United States and China. So there were fewer flights. In particular, there were periods there were only flights to LAX or only flights to SFO. But fewer is not the same as none, which is what “cut off” implies.
> Section 1. Suspension and Limitation on Entry. The entry into the United States, as immigrants or nonimmigrants, of all aliens who were physically present within the People’s Republic of China, excluding the Special Administrative Regions of Hong Kong and Macau, during the 14-day period preceding their entry or attempted entry into the United States is hereby suspended and limited subject to section 2 of this proclamation.
https://trumpwhitehouse.archives.gov/presidential-actions/proclamation-suspension-entry-immigrants-nonimmigrants-persons-pose-risk-transmitting-2019-novel-coronavirus/
You will notice that this excludes Hong Kong and Macau. It only applies to aliens (not U.S. citizens or permanent residents), and there are a number of categories of aliens that it doesn’t apply to. There were 279 direct flights from China to the United States between Feb. 2 (when the travel restrictions went into effect) and April 15.[1]
Rather than writing “may have contributed,” I probably should have indicated that it did contribute.
I don’t believe that any prominent Democrat claimed that the travel restrictions were racist.
[1] https://archive.li/8pjIn
re "we gave China our best chips", the H200s under discussion are actually a generation behind the current SotA, Blackwell. Still bad, though.
Thanks, fixed.
"Lactose-intolerant people who drink milk regularly since childhood develop gut microbiota that can digest milk, but which demand an expensive “tax” in calories."
If this "tax" overshoots and burns more than the calories of the milk consumed then these microbiota sound like promising chemical-free slimming aids. There's probably a downside though, such as the calories being converted to methane and thereby giving rise to copious highly flammable trumps!
I don't know for sure, but I think the tax is taken as they digest it, so they can't digest more than 100% of the calories in their energy source.
Still, sounds like a nice calorie-free snack then.
Lactose is a sugar so letting microbes digest it has an advantage now--the blood sugar will not increase by that much-- a nice advantage in the middle of worldwide diabetes epidemic.
Re #30, I'm mildly aphantasic, in that I can visualize things in my mind but only in a very vague and phantasmal way, but I have a very vivid "mind's ear," and can hear voices and music in my head almost as realistically as if I was actually hearing it. I wonder if there is a tradeoff between these things, or if they're positively correlated? I also wonder about the relation between internal monologue (of which I have a very strong one) and ability to visualize, do they stem from the same thing or is one the result of being a verbal reasoner and the other a result of being a visual reasoner?
FWIW, I seem to be near the rightward tail of all three distributions (extremely pervasive/consistent internal monologue; vivid mental imagery; can sometimes, albeit rarely, mistake an imagined sound for a real one).
As I mentioned in another comment I left above, I sometimes doubt that anyone *really* differs in any of these three—how can one even really be sentient without an internal monologue? how can anyone without mental imagery do things like mentally rotate a shape to "fit" it into another?!—but I guess that's probably "typical-minding".
There was a very interesting book I read as a kid named Ratha and Thistle-Chaser about a sentient species of cats. One of the members was raised outside of society and so didn't learn language, and of course lived life with no internal monologue. (She still had ideas, but in the form of pure concepts rather than tied to words.) Sometimes I'd try to copy that just to see if I could, and found that yes, I can manipulate thoughts without first turning them into words, but it's harder and I lose track more easily.
Interesting!
I find that I do *some* wordless thinking, too—particularly in mathematics: it's hard to explain, but it's something like "turning a mental object around & around until it fits into a larger mental object", if that makes sense (i.e. when it so fits, if it does, *that's* when I truly understand the concept illustrated by the example/problem).
Oddly, though, I'm trying to introspect *in re* how I understand formal logic, and it feels as if such cases are mostly verbal: thinking through them something like "if A is X, and B is Y, therefore it must be that C cannot be Z...", and so on.
I'm not very confident in these self-reports, though (which is another odd thing: how *can* one be mistaken / about one's own mentation? / it's a mystery, it seems to me— / much to my consternation! ...*cough*–).
For when I'm trying to do it on purpose, it helps to spend a couple minutes first focusing entirely on the input of my senses, the texture of the leaves of the potted plant in front of me, the roughness in the latter half of each tick of the wall clock, little details that aren't easily put into words. I also have to focus entirely on the present moment, since otherwise I would think about how I'll be explaining it in a few minutes, which takes words. Then I can try to think about different things and see what can come out non-verbally. For instance, I can think about getting a glass of milk from the fridge, and this comes in as sort of a picture of me in front of the fridge door + a general concept of milk. When I'm not trying to avoid words, that concept still comes in like that, but will be shortly followed or preceded by the internal monologue "I could go get a glass of milk". My brain seems to prefer replacing words with pictures as the first go-to, but if I try to avoid that, I can think about concepts using other connections to them, such as sounds or spacial relations. I suspect there's at least one type of meditation that works more or less the same way, but haven't felt a need to look into that.
Logic is very verbal - it’s pretty much all about syntax!
When you think fast, does your inner voice speed up?
Yep—at least, I seem to recall that it does, and experimenting with trying to think fast just now made it speed up (although the amount of extra speed I can squeeze out seems somewhat sharply limited 🤔).
TBH, I find "shape rotating" to be an extremely difficult thing. Verbal stuff? I've always been off the charts. But visio-spatial? Absolute crap.
On the three, I have a constant internal monologue/dialogue, active *verbal* imagination, but very poor visualization ability, and mistake imagined sounds for real on a regular, but not frequent, basis. Partly due to living alone.
I dream in visuals, but lose anything but a text description once I wake up.
I'd definitely say that I have more difficulty in mentally rotating shapes, or unpicking knots, or the like, than with the other two—and even my ability to visualize colors or textures is much stronger than my ability to keep complex shapes in mind & manipulate them etc.¹
How do you do on tasks like Raven's Progressive Matrices & similar "complete the sequence" puzzles, and/or those "analogy" problems ("A is to B as C is to... __")? IIRC (which may not be a safe assumption, heh: see footnote #1–), these are two of the most g-loaded question-types commonly encountered on IQ tests & so forth...
...but that's just an incidental Fun Fact!™ (𝘯𝘰𝘵𝘦: 𝘤𝘰𝘮𝘮𝘦𝘯𝘵𝘦𝘳 𝘱𝘳𝘰𝘷𝘪𝘥𝘦𝘴 𝘯𝘰 𝘨𝘶𝘢𝘳𝘢𝘯𝘵𝘦𝘦 𝘰𝘧 𝘢𝘤𝘵𝘶𝘢𝘭 𝘧𝘢𝘤𝘵𝘶𝘢𝘭 𝘴𝘵𝘢𝘵𝘶𝘴; 𝘶𝘴𝘦 𝘢𝘵 𝘰𝘸𝘯 𝘳𝘪𝘴𝘬, 𝘦𝘵𝘤.) Rather, I bring these up because the former, at least, seem sort of "visual" (moreso than they seem verbal, anyway; though, of course, the analogy thing is just the opposite)...
...and yet I've always been much stronger with RPM than with "which shape fits into this other shape if rotated 120° clockwise?" or whatever—I almost feel as if 𝘀𝗼𝗺𝗲𝘁𝗵𝗶𝗻𝗴 about, some underlying cognitive mechanic in, a "continue the pattern" puzzle is similar to the same in (say) "find[ing] the right word", vel sim.²—so I wonder if this also applies to other "verbal-dominant" folks too.
-------------------------
¹(Interesting that you should mention this—I thought I'd included something similar in my initial comment, but I guess not; great minds just think alike! Wordcel power! 👊 Actually, I seem to recall that high verbal ability is more g-loaded than the other sort(s); semi-but-not-really–relatedly, I read somewhere that so-called "crystallized" intelligence is also a better predictor of 𝘨—or maybe it was just total IQ?—than is "fluid" intelligence... which last seems hard to make sense of, tbh, so I may be misremembering; call it maybe 65/35 in favor, 𝘪𝘯 𝘳𝘦 my confidence that I accurately recall a real finding, heh.)
²(or else this particular supposition is just my "wordcel cope" coming to the fore–)
I'm way better with analogies than any kind of the RPM or "find the next element in the sequence" things. Multiple choice tests are something I'm good at, as long as there are answers I can exclude via logic.
I figure I'm really good at manipulating symbols and interpolating from known facts. Geometry sucked. Calculus was fine. Physics without numbers is much better than physics with numbers. Word problems are my jam.
Couple that with an obsessive reading habit leading to a large fact base, I can seem really smart. By no means genius level. I'm a generalist though.
>can sometimes, albeit rarely, mistake an imagined sound for a real one
I have good mental visualization, but very good auditory 'visualization'. Not sure the right words here. I can reliably recreate entire songs, beginning to end, all the instruments separated, in my head. I've also had many years of formal musical training, which has to be a big part of it. I've done a fair amount of transcription too.
I frequently "mistake an imagined sound for a real one", or what my psychaitrist calls persistent auditory hallucinations. I don't remember the last full day that passed where I didn't have one; its probably been 15 years. I have to think the two are related. Novel hallucinations are pretty rare now, they all tend to be of a certain narrow type. Music lightly playing in the other room, like someone has left a radio or television on but the audio turned down, is pretty common. Once a day or so, more if I'm stressed. There have been very stressful periods of my life where the 'distant music' hallucination persists for multiple days without stopping. Hearing my name being called is also fairly common, as are soft and distant conversations coming from the other room like the music. The sound of the alarm clock I had for most of my childhood is also fairly common. I'd estimate I know right away 90% of the time that its a hallucination, and eventually determine almost all of them. There are maybe 5-10 sounds in my life that I'm still not sure if its real of not. Some of my original compositions have come from my hallucinations, which sounds like a superpower to musicians I tell about it, but its not really. At best its a snatch of a melody or a couple bars of chord progression, or even a single chord.
Anyway I've learned to largely ignore them. I think if they were clearly spoken language hallucinations it would be a lot worse. I've only had those in periods of extreme duress.
Interesting. I think verbally and "hear" my internal monologue but it's only ever my own voice. I can't imagine other people's voices and I can't even imagine sounds that aren't speech. If, for example, I think of the sound that hitting one piece of metal against another makes, I just get my own voice saying "Clang!"
As for visual reasoning: I can do geometry in my mind and I thought that was what people meant by visualizing things, but apparently it's not.
For another point of anecdata, I'm aphantasic (can't conjure visual imagery in my head), but I think I have an approximately average "mind's ear" and internal monologue.
Not sure. My imagination can be quite vivid in general - I can not only imagine detailed scenes in my head, but I can also imagine sound, or even touch, smell, and taste (although being able to vividly imagine eating ice cream hasn't stopped me from seeking out the real thing).
I suspect that these different types of imaginative ability are actually less correlated (positively or negatively) than people will assume.
Just to point out the spelling mistake, please tell me when you correct it so that I delete my comment:
8: New revelations on the OpenAI coup from the Musk vs. Altman lawsuit. The effort to remove Altman may have been **let** (obviously it's led) by Mira Murati and Ilya Sutskever...
Otherwise thank you again for your work!
Corrected, thanks.
53: Huh, I didn't know that! Stoll isn't just a "scientist"; he has substantial Silicon Valley cred by way of his 1989 book The Cuckoo's Egg. You can be arbitrarily close to the technology and still get predictions about its future course very wrong.
I read that book as a kid. He notes that his fellow astronomers thought he was a terrible astronomer, but a whiz with computers. When he switched to working with computers his colleagues thought he was lousy at that, but a genius astronomer.
Stoll's bio describes him as an astronomer and says he has a PhD in the subject. I think this qualifies as a scientist.
Yeah, it absolutely does. I was just noting that he also has a bunch of IT experience, which if anything sharpens the point being made.
See also, gwern's take on the opposite phenomenon (anticipated a technology much earlier) https://gwern.net/timing
I think its also likely that Stoll was exposed to more computer-hype than most, given when he was most active. The 70s and 80s were full of tech promises that were just around the corner. Academia had been on the early internet for years by 1995. Admittedly, even the most enthusistic tech promoters of the 90s and 00s missed how popular smartphones would be with the normies, especially women. I remember an anecdote that Apple didn't expect the Iphone to be all that popular with women and thought their Iphone buyers would be the same group as their core-demo computer buyers.
2: “Perhaps the most baffling … was the common Kestrel. He was known simply as the Windfucker.”
Mediaeval names were very blunt. Witness all the “Gropecunt Lanes” that got bowdlerised into Grape Lane:
https://en.wikipedia.org/wiki/Gropecunt_Lane
As for the bird names, when I was a child, it was “Willie Wagtail” and not just (pied) wagtail.
6: Re: the anti-mosquito drones, they undoubtedly do look very cool. They seem to work by colliding with the mosquitoes (and hence chopping them asunder), though, and someone is going to raise a fuss over cruelty and inhumane killing methods.
10: St. Carlo Acutis (1991 - 2006) is the unofficial patron saint of the Internet and “first millennial saint”. …(H)is sainthood is something of a mystery, with Wikipedia saying that “even those with a deep devotion to him struggle to pinpoint his specific actions that led to his canonisation”, and an Economist article admitting that “nothing in his sparse life story explains that this ordinary-seeming teenage boy is about to become the first great saint of the 21st century”.
First, there are the two miracles necessary for canonisation:
“On 14 November 2019, the Vatican's Medical Council of the Congregation for the Causes of Saints expressed a positive opinion about a miracle in Brazil attributed to Acutis's intercession. In 2020 the Catholic Church recognised the curing of a child's pancreatic disease as a miracle attributed to Acutis's intercession. On the death anniversary of Acutis, Luciana Vianna had taken to Mass her son, Mattheus, who had the congenital defect of an annular pancreas which made eating difficult. Beforehand, she had prayed a novena asking for Acutis's intercession. During the prayer service following Mass, Mattheus kissed the clothing relic of Acutis and asked he should not "throw up as much". Immediately following the Mass, he told his mother that he felt healed and asked for solid foods when they arrived home. Until then, he had been on an all-liquid diet. His doctors subsequently confirmed a normal appearing pancreas by ultrasound. …After a detailed investigation, Pope Francis confirmed the miracle's authenticity in a decree on 21 February 2020, leading to Acutis's beatification.
…On 23 May 2024, Pope Francis recognised a second miracle attributed to the intercession of Acutis. The miracle attributed to his intercession occurred in 2022, when a Costa Rican woman named Valeria Valverde had fallen off her bike in Florence, Italy, and suffered a brain haemorrhage with doctors giving her a low chance of survival. Her mother, Lilliana, prayed for the intercession of Acutis and visited his tomb. The same day, Valverde began to breathe independently again and was able to walk the next day with all evidence of the haemorrhage having disappeared.”
Second, you generally need to look into who is pushing the case for canonisation. Either there’s a local cultus, or some interested party/parties are involved. In this case, it seems to have been the local parish:
“The call for Acutis to be beatified began soon after his death. The Church in Assisi was active in promoting Acutis's cult, the Catholic practice of venerating or showing devotion to a holy individual. Nicola Gori, who authored Carlo Acutis: The First Millennial Saint, served as the postulator for Acutis's beatification process. In this dual role, Gori was responsible for advancing the cause within the Vatican and compiling the necessary documentation.
On 12 October 2012, the sixth anniversary of his death, the Archdiocese of Milan opened his cause for canonisation.”
19: New research analyzes Hitler’s DNA.
Yeah. How sure are we that this genuinely is Hitler’s DNA and that it’s not degraded? Where did they get the sample? Off a sofa?
“The testing was made possible after researchers obtained a sample of Hitler's blood from a piece of material taken from the sofa on which he shot himself.”
I think this is more in the same wheelhouse as all those Discovery channel “Tomb of Jesus/Mary Magdalene/Your Great-Uncle Theobald discovered in site five miles outside Jerusalem” ‘documentaries’ than genuine scientific research. Publicity for the documentary which is yet another in the never-ending stream of “the public can’t get enough about the Nazis” productions.
32: “These are my people? My Volk? Am I supposed to sacrifice life and limb for the salus of this populus? Fuck that. Let them cook, they deserve everything that’s coming to them”
The answer is to put that devotion into Salus Populi Romani 😁
https://en.wikipedia.org/wiki/Salus_Populi_Romani
https://www.youtube.com/watch?v=0_Zh8d_FqnQ
43: Sam Bankman-Fried’s mother on why she thinks his trial was unfair.
Oh, Sam’s mom thinks he’s didn’t get a fair shake? What a surprise. Also not surprising since she and her pet causes got a ton of money off SBF’s shenanigans, as well as her and Dad getting some nice perks:
From the CTFC filing:
“54. Contrary to such representations and without disclosure to FTX customers, Alameda and FTX comingled funds and freely used FTX customer funds as if they were their own, including as capital to deploy in their own trading and investment activities. On information and belief, Bankman-Fried, his parents, and other FTX and Alameda employees used FTX customer funds for a variety of personal expenditures, including luxury real estate purchases, private jets, documented and undocumented personal loans, and personal political donations.”
From the SEC filing:
“68. Bankman-Fried also used commingled funds from Alameda to make large political donations and to purchase tens of millions of dollars in Bahamian real estate for himself, his parents, and other FTX executives.”
https://worldsleaders.com/bankman-frieds-ftx-parents-bought-bahamas-property-worth-121-million/
“The majority of FTX’s purchases were high-end beachfront properties, including seven condominiums in the pricey resort neighborhood of Albany that cost about $72 million. According to the documents, a unit of FTX purchased these houses to serve as “residence for key employees” inside the organization. Reuters was unable to identify the residents of the units.
Bankman-parents, Fried’s Stanford University law professors Joseph Bankman and Barbara Fried, are listed as signatories on the papers for another beachfront home in Old Fort Bay, a gated neighborhood that formerly housed a British colonial fort constructed in the 1700s to stave against pirates. One of the paperwork dated June 15 stated that the property is intended to be used as a “vacation residence.”
When questioned by Reuters about why the pair opted to purchase a holiday home in the Bahamas and how it was paid for, a representative for the academics merely indicated that Bankman and Fried had been attempting to return the property to FTX.”
45: The “100 Above The Park” building in St Louis
Huh. Modern architecture that is surprisingly not terrible. I was expecting to hate it, but I like it. Looks vaguely Art Deco inspired, almost like a bouquet rising out of ground level.
https://cdn.shopify.com/s/files/1/0271/2347/files/6395997365_d399c2a71f_b.jpg?2509385242815651011
I have to doubt the Hitler claim, too: there's an endless market for such things, and this nicely flatters the dominant narrative ("guess what: insane & micropenis, lmao! drug-addict too! also he loved eating feces & fucking his dogs, we think!"). Not that it can't be true, mind, or that I'm a fan of the fellow—it's just awfully convenient... one can imagine the team(?) involved thinking "okay, now to make sure this'll get tons of press!" and/or "ha! Kallman—THIS'LL show those alt-right neo-Nazis!", and so forth.
Yeah, it sounds like a combination of small independent production company doing content for the channel ("Channel 4 operates a unique "publisher-broadcaster" model, which means it primarily commissions and outsources all of its content from independent production companies, rather than producing programmes in-house") which means it wants 'guaranteed audience grabber' and Hitler is one of the guaranteed audience grabbers, plus some company with a swizzy new DNA technique they wanted to monetise. They could go the boring old 'publish papers in reputable journals' route, or they could go for WE SEQUENCED HITLER'S DNA NO HE WASN'T JEWISH YES HE WAS A BIG POOPY-HEAD SCIENCE SAYS SO PLUS HE HAD A SMALL TINY BARELY-THERE DICK.
I think they preferred the second option 😁 I mean, with "you couldn't ask for more rigorous sample management and custody with no chance of cross-contamination than this" techniques on display, why are we being so pettifogging?
https://www.bbc.com/news/articles/c5ylw4pz83do
"The bloodied swatch of fabric - now 80 years old - was cut out of the sofa in Hitler's underground bunker, where he killed himself when Allied forces descended on Berlin at the end of World War Two.
While inspecting the bunker, Colonel Roswell P Rosengren of the US army saw an opportunity to get a unique war trophy and he pocketed the fabric. It's now framed and on display at the Gettysburg Museum of History in the US."
'Cut it out of the very sofa after everyone and his dog had pawed all over the bunker, shoved it in me pocket, flew home thousands of miles with it, bish-bosh, job done!'
I do like how all the academics involved were shocked, shocked I tell you! that the TV station slapped a clickbait title on the two-parter and went hard on the sensationalism angle 😊 Whoever could have seen *that* coming?
Maybe it wasn't a new company, their expert Dr. Turi King seems to be the biology equivalent of Neil deGrasse Tyson (ready to go on telly at the drop of a hat to tell the public all about SCIENCE!):
https://en.wikipedia.org/wiki/Turi_King#TV,_video_and_radio_appearances
"TV, video and radio appearances
King has appeared in numerous television and radio documentaries, programmes and videos as well as advising on television and radio productions such as BBC's Who Do You Think You Are?
Presented the series Ancient Murders Unearthed for Sky History.
Presented the BBC Radio 4 documentary: Genetics and the longer arm of the law
Guest on The Life Scientific as DNA Detective Turi King
Britain's Lost Battlefields with Rob Bell
Richard III: The King in the Car Park
Britain's Secret Treasures
Richard III: Solving a 500 Year Old Cold Case (TEDx Leicester)
Richard III – The DNA Analysis & Conclusion (University of Leicester)
Richard III: The Resolution of A 500-Year-Old Cold Case (Irving K. Barber Learning Centre Lecture, UBC)
Head Number 7 for Wondery
DNA Family Secrets
DNA Family Secrets is a television series which began airing on BBC Two in March 2021, presented by Stacey Dooley and King, it uses current DNA technology to solve family mysteries around ancestry, missing relatives and genetic disease. Series three aired in 2023."
>I mean, with "you couldn't ask for more rigorous sample management and custody with no chance of cross-contamination than this" techniques on display, why are we being so pettifogging?<
lol! Good point, very good point; I guess we are forced to accept this airtight scientific breakthrough after all. ADOLF "POOPY-HEAD TINY-PP" HITLER: CONFIRMED!
Yeah, it would have been terrible if it turned out to be Colonel Rosengren's DNA, wouldn't it?
Luckily, we know there is no chance of that because:
"The scientists are confident it really is Hitler's blood, because they were able to perfectly match the Y-chromosome with a DNA sample from a male relative that had been collected a decade prior."
A decade old sample, stored we don't know how, from a 'male relative' whose ancestry we're not informed about. Well, with fact-checking like that, whew, crisis averted!
Aha! Well, then... consider the doubters humiliated!—
(Assuming I understand correctly—not always a sure bet—then that very likely means only that the two Y-DNA samples shared a common patrilineal ancestor at some point within the last 10–20 generations, or something like that... although I could be wrong—maybe it is in principle possible to be more definitive—but given the context... I'm willing to bet [up to $5, or even $10!] that it's something like that, and not a definitive result—esp. given that it looks like the putative living relative's sample was identified with some rather questionable methods: collecting discarded cigarette butts & napkins, etc., from supposed likely relatives identified by a Belgian journalist through tracing "archival documents" to other branches of the family in Austria. Seems rather circular, in a way: "well, the sample has this somewhat unusual haplogroup, which matches a living relative's; we know Hitler had that haplogroup, because these living relatives have it; we know they're relatives, because they have this unusual haplogroup..." The case stands or falls with the archival research, it seems, but that's been kept secret, so—depends upon how much you trust that particular Belgian journalist, perhaps? Maybe others checked out his work, I don't know, but as it's not public it's sort of hard to be too confident in it!)
A Belgian journalist? By the name of Tintin, maybe? 😁 Well if it's a journalist from the land of Tintin and Hercule Poirot, we need have no qualms.
Though now I'm beginning to wonder: should I have my ancestry traced? What if I turn out to be a relative, too?
The obvious way to find out is to implant the supposed-Hitler DNA into an embryo and see if it grows up to look like Hitler. You'd probably want to do it, like, a thousand times, in order to get a sufficient sample size.
In Brazil.
That boy’s gonna want some dogs. Make sure they’re big ones.
I think something like that is essentially a requirement for fascism- I think it’s fundamentally a much more sexual thing than usually appreciated. If you look at the current fascist movement in the USA, recruitment is largely based on convincing young men that they are somehow sexually inadequate, and that their movement will fix this by teaching them how to perform faux hyper-masculinity properly and teach them not to be “cucks” anymore. This usually works by exploiting a pre-existing terror that they are sexually inadequate or feminine somehow, which often stems from something about their personality or body that they are afraid or ashamed of. Once I noticed this, I started so realize that the references to sexual shame and terror are pervasive, even in executive orders and press communication from the Trump administration. A critical part of fascist propaganda is the leaders always being surrounded by attractive women, and the opponents being framed as sexually unattractive and unsuccessful.
Hitler has living relatives, so it would actually be pretty easy to verify it. Not sure if they did, though.
That's part of the problem. It well may be that the relatives refused to get involved because they're fed-up with being hauled out every ten minutes as "Hitler's relatives", but you would wonder why they couldn't get a more recent blood sample that definitely is identified as coming from the family.
Even if the blood was uncontaminated - whats the genetic sequence for a micro penis? Do we even have that kind of knowledge.
It was the genetic sequence for a particular disease that is known to sometimes have that kind result. Yes, we do have that kind of knowledge. But was it accurate is another question.
> 6: Re: the anti-mosquito drones, they undoubtedly do look very cool. They seem to work by colliding with the mosquitoes (and hence chopping them asunder), though, and someone is going to raise a fuss over cruelty and inhumane killing methods.
Thankfully, the same techniques work for anti-activist drones as well!
From the Wikipedia article on Gropecunt:
> Although some medieval street names such as Addle Street (stinking urine, or other liquid filth; mire[20]) and Fetter Lane (once Fewterer, meaning "idle and disorderly person") have survived, others have been changed in deference to contemporary attitudes. Sherborne Lane in London was in 1272–73 known as Shitteborwelane, later Shite-burn lane and Shite-buruelane (possibly due to nearby cesspits).[21][22] Pissing Alley, one of several identically named streets whose names survived the Great Fire of London,[23] was called Little Friday Street in 1848, before being absorbed into Cannon Street in 1853–54.[24] Petticoat Lane, the meaning of which is sometimes misinterpreted as related to prostitution, was in 1830 renamed as Middlesex Street, following complaints about the street being named after an item of underwear.[25]
Unfortunately, Cliff Stoll's 2010 comment no longer appears to be on the internet. It's quoted in the Wikipedia article, which cites a Boing Boing post, but the comments for that are gone. If you use the Internet Archive for that url, it will just try to load up Disqus and hang. Nor does any post that old appear to be in Boing Boing's read-only BBS that the "Comments" link goes to now.
Thanks very much for the featuring and the grant Scott!! On the choice of the White House: we did this video for an internal YC presentation that wasn’t supposed to be public, so the choice was a combination of both 1) fun 2) there was a cheap and very good quality model on cgtrader. Later our contractor got too late for YC timelines, so we went with that video. Can’t complain about the additional buzz! We’ve been able to have the luxury of the choice of investors, and a significant chunk of our cap table is EA-aligned, so we have no plans of getting into defense. The expansion market we found to make VC economics work is now invasive flying species control for agriculture, eg killing codling moths who lay eggs in apple plantations with microdrones rather than insecticide. (Also had the advantage of being a much easier target, both in terms of sonar RCS and max acceleration). Exciting times ahead!
Seems like cool technology and I can definitely believe that it works at having a device which can go kill mosquitoes. But I was surprised to see that there was no outcome-focused analysis anywhere I could find on the website. It says that it can patrol 5 acres, but what does that mean? What reduction in mosquito presence does it achieve compared to e.g. UV-light-based traps?
Yeah, I'm interested in specs, too.
I mean, I pre-ordered one, how could you NOT? I've lamented the overweening lack of mosquito-killing drones too many times to not be obligated to buy the first one I see.
But I'm really interested in coverage, outdoor / weather use specs (can you use them in wind? What if it rains, are they smart enough to go back home?) and more. I assume there will be software that lets you manage them at fleet scale? The killer app to my mind is having a fleet for your property killing all mosquitos that come anywhere near your property without you having to think much about it.
Super interesting product but one of the reasons I avoid mosquito killing tech is off-target effects. How do you avoid killing other small flying insect species? Is it able to tell the difference? Either way, good luck!
Don't birds and bats kill small flying insects indiscriminately as well?
The scale matters and they're proposing a solution to kill all mosquitos in an area: if they're also killing all flying insects in that area then I don't want it. If the bird and bat populations kill a fraction of flying insects that keeps their population at steady state, then I don't care. Indeed, if you like those birds and bats then you better hope you don't kill all the flying insects.
I'm sure there's a false positive rate but it seems pretty plausible that the drone could specifically locate mosquitoes based on their distinctive whine.
Here is a talk he did a year ago:
https://www.youtube.com/watch?v=6ScCG3qTOuc
TL;DW theoretically, microdoppler signatures from the sonar.
I don't want to be a downer but this is a VERY difficult problem to solve, and something I have been casually researching for a while and have some industry experience. I've seen a few different ideas crash and burn but I think he is on the right track with sonar. Mosquitos are so small I'd be extremely impressed if he had an actual demo of a tracking algorithm working outside
🙏🏻
Re 53: You might better know Clifford Stoll today as the funny old guy who frequently appeared on Numberphile talking about Klein bottles and other topological objects, and has a small business selling glass Klein bottles.
I *do* better-know Stoll as that guy! Neat—thanks for this.
He gave an interesting TED Talk, and his book The Cuckoo's Egg was good (I have not tried the cookie recipe inside yet, though, having my own family recipe).
With #39, important to note that is relative to the shift in Natural Sciences, of shifting 0.18 left on a five point self-reported left-right scale. So eyeballing, everything but accounting and finance moves people left overall, but how much depends on the subject.
All I noticed was the right leaning ones were degrees you could get you a job.
Good point! Their second figure seems to show the absolute data and it’s a very different picture. Interesting is theology being the most right at the start but having possibly the largest leftward movement.
Scott, why did you post the picture of "100 Above the Park"? Are you saying, "this is a cool-looking building and we should build more like this"?
(...because if so, I agree.)
Signs point to yes. (https://www.astralcodexten.com/p/whither-tartaria)
That’s my interpretation too - but it looks to me like the kind of modernism Scott says he doesn’t like!
My tastes seem to be similar to his, and I'd say that this is a better example of post/modernism - doesn't look to be intentionally boring or ugly or absurd, which is a low bar, but still rarely cleared these days.
It actually does look intentionally absurd to me, which is part of the charm! Why have all those weird twisted corners instead of a normal wall? Just because!
Well, it's a matter of degree. The dictionary says that absurd means "wildly unreasonable", whereas here I'd say it's mildly unreasonable. I mean, it's no Lou Ruvo Center or whatever.
Did anyone do the Lighthaven blogging thing and how did it go? I ask because I noticed Ozy posted about every day in November and that seemed like the blogging thing.
I had no idea #2 went beyond Robin Redbreast. That's pretty fun, and I wish it had survived more. We did our own version of it, for rats, when they left the closed restaurants during the pandemic: when we would see a rat scurrying around outside our apartment, that was Pat. Pat the Rat. They were all Pat the Rat.
Patty wagon?
In my family, all flies are named Freddie and all rats are named Ralph.
In my yard all frogs are named Meestar Frog and all skunks are Meestar Skunk
All mice are named Gerald. I don't know why.
29. I am now calling for a complete shutdown on naive utilitarianism until we can figure out what the hell is going on. We need a Manhattan Project for non-Benthamian consequentialism.
>Most interesting section is the one about children: he wants to have them, but doesn’t think they should be genetically related to him.<
Frankly, this is utterly alien to me. I have a hard time understanding why anyone would want children in the first place—when a friend announced that he (well, his wife) was going to have a baby, it took all of the acting talent within my grasp to feign the appropriate reaction ("congratulations! wow, how exciting! wooo!")—but the *sole almost-comprehensible reason* with which I can identify is "I want my line to continue / I want to see myself reflected in my kid".¹ If you don't have that, it's merely a lot of effort & worry & cost & time & sacrifice for nothing at all...
...except, I suppose, "someone to take care of me when I'm old"; I guess that still obtains.
----------------------------------------
¹I know, I know... something is probably wrong with me. Maybe I *am* an alien. I have a similar problem with babies in general, too: they're irritating & gross, not cute. Quickest way to get me to turn any film or show off, or leave any gathering, is to introduce some goddamn baby.
Most parents find taking care of their kids to be fulfilling in a way that makes up for all the effort and stress. You can verify this by going on any parent's blog and reading the paragraphs of "my kid is so cute" gushing. I don't think it's that complicated if you generalize the issue.
Why do people bother with romantic relationships when they could just have casual sex? Why do people want dogs? Humans find connection fulfilling.
I maintain that: well, of course—they *have to* convince themselves of said fulfillment; elsewise, they'd just break down & weep uncontrollably over the giant chunks of their lives they have foolishly thrown down the "baby-drain"–
-------------------------
(More seriously, I think the difference between romantic relationships & parenthood, or dog-adoption & the same, is not categorical, but rather one of degree *in re* RoI... although I'll note that I'm also beginning to feel the same way about romantic relationships: tried it; had some good parts; but, overall, too much goddamn trouble, too much giving up my own preferences to satisfy someone else's...
(It's not particularly difficult to *explain* why people *do* it—for a certain sense of "explain"—at least: of course people want to propagate the species; evolution ought to ensure this, in some sort of way & probably several... it just apparently went wrong in me, heh, or perhaps I have an unusually u̶n̶f̶o̶r̶t̶u̶n̶a̶t̶e̶ high self-indulgence::human-connection ratio.
(...OR! it could be—this is my favored explanation—that people are just easily fooled: by themselves, society, sappy movies, etc.; they are, on the whole, simply unable to truly picture what it's really like, and only imagine happy Hallmark images of fun trips to the park or whatnot. I.e., I merely possess an unusual portion of clarity & self-knowledge—*everyone* feels this way, if only they knew it. That one makes a lot of sense, if you ask me–)
Of course evolution programmed people to find raising kids fulfilling, but I don't think you even need to invoke special evolutionary drives to explain why most parents are happy with their decision to have kids. I think you merely overestimate how much the average person prefers freedom/independence over stability/community ties, because you value freedom more than the average person.
You keep framing kids as being terrible because it's a huge and essentially irrevocable commitment. But most people don't hate the idea of having responsibilities they can't get out of nearly as much as you do. Many even take pride in it. The other side of responsibility is purpose.
...hmm. That's a good point—I guess if "expending effort" & "less freedom to indulge as you wish in what you please" aren't awful prospects to someone, they'll weigh less heavily in the calculus.¹
>𝐵𝑢𝑡 𝑚𝑜𝑠𝑡 𝑝𝑒𝑜𝑝𝑙𝑒 𝑑𝑜𝑛'𝑡 ℎ𝑎𝑡𝑒 𝑡ℎ𝑒 𝑖𝑑𝑒𝑎 𝑜𝑓 ℎ𝑎𝑣𝑖𝑛𝑔 𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑖𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑡ℎ𝑒𝑦 𝑐𝑎𝑛'𝑡 𝑔𝑒𝑡 𝑜𝑢𝑡 𝑜𝑓 𝑛𝑒𝑎𝑟𝑙𝑦 𝑎𝑠 𝑚𝑢𝑐ℎ 𝑎𝑠 𝑦𝑜𝑢 𝑑𝑜. 𝑀𝑎𝑛𝑦 𝑒𝑣𝑒𝑛 𝑡𝑎𝑘𝑒 𝑝𝑟𝑖𝑑𝑒 𝑖𝑛 𝑖𝑡.<
Thinking about it, I actually do know several such people; in truth, I would probably have a much better life, myself, if only I understood—felt things from—the same perspective.
(I've tried, but I'm beginning to wonder if maybe I oughtn't just give up: scrabbling uselessly upon the slope of self-discipline gets discouraging, by the twentieth or thirtieth time you've slid back down... heh.)
--------------------------
¹𝘐'𝘥 𝘴𝘵𝘪𝘭𝘭 𝘱𝘳𝘦𝘧𝘦𝘳 𝘵𝘩𝘦 𝘢𝘯𝘴𝘸𝘦𝘳 𝘵𝘰 𝘣𝘦 "𝘯𝘰 𝘵𝘩𝘦𝘺 𝘫𝘶𝘴𝘵 𝘤𝘢𝘯'𝘵 𝘥𝘰 𝘵𝘩𝘦 𝘤𝘢𝘭𝘤𝘶𝘭𝘶𝘴 𝘢𝘴 𝘨𝘰𝘰𝘥𝘦𝘳 𝘢𝘴 𝘐 𝘤𝘢𝘯", 𝘵𝘩𝘰𝘶𝘨𝘩–
Why do you do anything?
Would you prefer a life of living in your parents basement and playing videogames (amateur level), or going into the world and accomplishing things? Travel, helping others, discovering things, building things, anything.
Why?
Kids give our lives meaning, which in this day and age, is teh scarcest resource of all.
As someone who is similar to you on this axis and happy with it, I don't think you need to change yourself to be happy. Having responsibilities is important for fulfillment for almost everyone, but these needn't be "inescapable" responsibilities like kids or (to a lesser extent) marriage are.
Get a hobby where you make cool stuff, or a career that contributes to the world, or any other project. Anything from "head of multinational organization" to "runs local DnD group" can contribute to a sense of purpose, while being something you're socially and legally enabled to quit once you're tired of it.
> Frankly, this is utterly alien to me.
Finding someone else's goals utterly alien is 100% normal. It doesn't mean either is wrong, you just happen to have different interests in life.
I think this is just autism, and shouldn't be held against utilitarianism specifically (although we should always be vigilant for dysfunctions caused by the large overlap in the two populations).
Utilitarianism will always be with us. Some people love numbers and any philosophy that allows them to play with more numbers is going to be a winner.
No need to shut anything down, that mentality will self select out of the gene pool.
I'll sign this letter if we remove the word naive.
Has anyone written an explainer about whether the pardon of Juan Orlando Hernandez has anything to do with Prospera? I've seen people mention it in passing but not a deep dive.
#12: I have recently seen a talk about exactly that topic of switching out different AIs on the same chip. This is not just a problem with outdated models, it is also a problem in expert systems where you have dozens or hundreds of specialized expert AIs, and for each prompt you decide which of them you ask. (Possibly several.) Those systems already run in the background of modern AIs.
If the expert AIs are not used with the same load all the time, then you frequently need to load/unload some of them into GPUs, which is a very slow operation. It is also not a good option to just have all of them sit in some GPU being idle most of the time, just to be able to serve peak demand.
The solution in the talk was that many of those expert AIs are pretty similar, and the difference can be very much compressed. There are technical problems to be solved -- even if you can compress the diff into a small block, it would still be slow to decompress it and then change all weights a little. But the problems can be solved, and this is what the talk was about.
#20:
For the US, you have cities above the national-average line and states below. For Germany, you have three city names above the line and the name of one Bundesland below. Of course you can have all kinds of data dots in this figure, but then shouldn't the removal of the richest region be called the removal of the richest city? The blue and red lines would certainly be further apart if you removed Bavaria instead of Munich.
Should I be concerned that my closed-eye hallucinations look a lot like the psychedelic videos in 26?
Mine do too. I suspect that psychedelics only amplify it rather than creating it.
I suppose it's not terribly surprising that a neoreactionary is repeating self serving fascist propaganda, but in case anyone is wondering, Mussolini did not make the trains run on time and the Nazis did not 'fix' the Weimar economy.
I'm surprised how deep this is, and a bit disappointed that both of those 'facts' were uncritically repeated here
Ancap fixed both of them?
I don't know about "fixing the Weimar economy", but when I grew up in the 1950's a common comment on Mussolini was "at least he made the trains run on time". So it may not have been true (I've never checked), but it was certainly widely accepted.
I'm aware of the phrase's meme status. Unfortunately it's just not true, or at least, very misleading. A very brief overview is available on wikipedia:
https://en.wikipedia.org/wiki/Benito_Mussolini#Railways
The argument presented in the excerpt is "Actually fascism isn't so bad, at least it's efficient/effective!" This argument does not follow if fascists weren't actually efficient or effective. In reality they were neither, but cared very much about the appearance of being so, which has contributed to the meme's sticking power.
My apologies to you personally for my tone, but even the third hand repetition of these falsehoods here is offensive to me.
Well, in the contexts that I heard it, the implication definitely wasn't "Actually fascism isn't so bad", just that it had a (few?) feature(s) that weren't awful.
Seems like even the weaker implication is wrong.
It is important to point this out, because even the weaker implication suggests a trade off: "With fascism, you get less of X and Y, but you get more of Z." There may be a situation where Z gets so important that people consider desperate options.
Actually, whether that's true or not depends on the precise definition of fascism that you are using. I rather think that Mussolini would consider the US a fascist state, at least if his explanation of the symbol (sticks bound in a bundle around an axe) is taken as what he meant. (The sticks were the companies and the axe was the government.)
"Say what you want about Hitler, but at least he killed Hitler."
It's a piece of propaganda for which people lost the original context. It comes from the fact that during the Biennio Rosso (the "Red Two Years", ie 1919-1921, when industrialization and what the working class perceived as broken promises after WW1 led to social unrest and strikes), the railway workers were heavily unionized, and oft striked either for better contracts or in solidarity with other workers.
Now, the Fascists got their jumpstart in politics by acting as the hired muscle of the factory and land owners. This had put them to an obvious crash course with unions, even the ones not particularly affected by Fascist raids, like the railway unions.
After Fascists seized power, one of their first actions was to enlist the entire state apparatus in labor repression, outlaw unions, and privatize state companies with strong unionization (chiefly the railways). This became one of the central points of their early propaganda.
So the "made trains run on time" was not meant as "we made the railway administration more efficient", it was "we squashed the labor movement so thoroughly that even its hardest kernel would not dare striking, and a gentleman's trip will not be interrupted by these insouciant screechers any more". That was 100% true. Except that as the regime dragged on, it tried to broaden its base beyond the upper and middle-upper classes, and after 1929, mass unemployment made their previous political formula completely unviable. So there was an hard turn toward more left-wing rhetoric, many companies were re-nationalized, some labor protections and welfare provisions were re-instated, and the previous propaganda sounded so incongrous with the new public image of the regime that its original meaning was lost and got re-written in collective memory as some sort of silly, random boast like many others.
That sounds like a quite plausible origin. I was in, perhaps, third grade, so that kind of thing was totally outside my awareness.
That little rant seemed weird to me as supposed evidence that some fascist has stopped being fascist - he still sounds pretty fascist but just lost faith in the people.
Fascists are hard to tell apart from general haters. Whenever I find one on the internet, in addition to all the stereotypical fascist hatreds he also hates the poor, the wealthy, women, feminized men, weight lifters (unless he is one), professional sports players, criminals, cops, smart people, stupid people….
Color me unsurprised that this one (former?) fascist hates his Volk. The step from fascist hater to non-fascist hater won’t make him sound like less of a fascist.
Yeah it's a bit different for Nazis and Falangists, but for the OG Fascists a certain Nietzschean will to power, the superiority of the exceptional individual over the dull masses, life made art and art made life, exploration of the possibilities for individual greatness offered by technical modernity but rejection of any democratizing tendency (even and maybe especially the ones created by the new industrial economy) was as big if not a bigger factor than nationalism or racism.
Marinetti, d'Annunzio, Balbo etc mostly stumbled into nationalism bc the Great War offered them a chance to play with all the new toys, live a life of danger and heroism, turn themselves into modern legends and whatnot, and at that point they might as well add some feelings for the country they were fighting for. But their supposed love for Italy was only ever instrumental to their quest to become ubermenschen, not the other way around (and absolutely did not prevent them to throroughly rape Italian society until it gave them the power and worship they thought they were owed).
Yes, it appears he still thinks fascism is the best thing for the people. Only now he hates the people, so he wants fascism to fail so that the people don't get its benefits.
#18
> “Fiat currency isn’t backed by nothing, as commonly supposed, but by land.”
Fiat currency is backed by the issuing state's promise and requirement that it is legal tender within its jurisdiction. A currency's backing is thus the entire state's existence and power to enforce its laws, or that law in particular. Land is certainly correlated with that, but not the foundation.
>in the long term, all fiat currencies go to zero
Yes, because in the long term, all *states* go to zero (economically or otherwise), taking their respective currencies with them. That is not a distinguishing features of fiat currencies. At most, fiat currencies enable new indicators of impending disaster, such as hyperinflation, but they're not the disaster itself.
That assumes that the state keeps the currency going until the state no longer can, which is the typically the case. However, when the Euro was adopted, older currencies were eliminated without those currencies going to zero, so the claim that *all* fiat currencies go to zero has been falsified.
Well, I believe that if you find a file of old Deutsche Mark banknotes and take them to a bank to exchange them for some contemporary currency, you'll find out their value has, in fact, gone to zero. (Except as collectibles, which, out of the scope of this conversation.)
What is Long term? Nixon took the US off the gold standard at 1973.
So 50 years is long term or medium term?
> issuing state's promise and requirement that it is legal tender within its jurisdiction.
What does this mean in practice? Stores aren't required to accept legal tender. They can choose to only accept Bitcoin, or gold, or barter, or whatever else they like. If for some reason they wanted to, people could just stop accepting dollars. Is it just taxes and existing debt that is backing the dollar?
Legal tender has the distinction of being privileged by the state. If money is part of a transaction involving the state, it's going to be in the state's currency. That means taxes but also payment of all its employees and all contracts with private businesses.
Yes, private business partners can agree to whatever payment they see fit, but if a contract has no form of payment defined, if the debtor offers to pay in legal tender, the creditor can take it or leave it but the debt is settled. Also, in practice, if you don't accept legal tender while your competitors do, all else being equal, you're at a competitive disadvantage and few will choose to do so on principle, making legal tender self-reinforcing. If refusal to accept legal tender becomes widespread, it's a sure sign of trouble for the entire nation's economy, like hyperinflation.
The taxes sound like a concrete use for dollars, but the rest doesn't sound different from gold or BTC.
1. Contracts are payable in dollars if no form of payment is defined.
Sure, but most contacts do specify (usually in dollars), and if people lost faith in the dollar, they would specify another type of payment. The fact that they could be forced to accept dollars if they messed up the contract is a very weak advantage for the dollar. You could also potentially not get paid at all if you draft a bad contract. Don't mess up your contract.
2. if you don't accept legal tender while your competitors do, you're at a competitive disadvantage
If you don't accept the most popular forms of payment, you're at a competitive disadvantage. But this isn't specific to the fiat currency. If in some country, BTC or cellphone credits or whatever is more popular than the fiat currency, then you're at a disadvantage if you don't accept that. This is a network effect or inertia: popular payment forms tend to stay popular.
I agree that if your fiat currency crashes that's a very bad sign, but that's true of many things. It's also a very bad sign if your treasury bonds or stock market crashes. That doesn't indicate special laws specifically about fiat currency.
>The taxes sound like a concrete use for dollars, but the rest doesn't sound different from gold or BTC.
It's no different, but we were talking about legal tender, not fiat currency. If tomorrow the USA decided to make TrumpCoin or seashells legal tender instead of the Dollar, my comment would still fully apply.
I'm not following. I meant that legal tender doesn't seem particularly privileged over any other form of payment apart from that you can use them for taxes. Stores, contracts, and debts do not have to accept legal tender, and saying that courts will default to legal tender if a contract forgets to specify a form of payment is only an incredibly weak advantage for legal tender.
You said, "A [fiat] currency's backing is thus the entire state's existence and power to enforce its laws, or that law in particular."
But what law? There is no law I know of forcing the populace to use dollars outside of paying taxes.
I'm with you on this.
"Legal tender" means that creditors must accept dollars in payment. But that applies only if the debt contract specified dollars. If you have a contract in bitcoins, the creditor can insist in being paid in bitcoins, and does not have to accept dollars instead. So there is no difference between dollars and bitcoins, what's relevant is what's stated in the contract.
I feel that "legal tender" simply means "taxes are paid in dollars and therefore dollars are the default currency."
Fiat currency is "legal tender for all debts public and private" within the state. Iirc there's multiple precedents where one party refused dollars for a debt. debt not a new agreement, and the debt was considered by the court to be forgiven.
This can happen if the debt is in dollars, or in a foreign currency in states that adopted the Uniform Foreign-Money Claims Act, but this wouldn't apply to a debt in Bitcoin or gold.
Dollars may be legal tender but that doesn't mean people have to accept it. The Federal Reserve website (https://www.federalreserve.gov/faqs/currency_12772.htm) points out that "There is no federal statute mandating that a private business, a person, or an organization must accept currency or coins as payment for goods or services."
Thanks
Germany between the world wars got its hyperinflation under control by issuing a new currency that was backed by land, as gold was in short supply and trust in the state even harder to come by. It did the job - the conversion rate from state-pinky-promise to land-backed German marks was something like 1e+12 (yes that's exponential notation).
(I don't think that contradicts what you said in most cases.)
35. Wow, they're really doing the "Happy Asking Panda" thing: https://www.youtube.com/watch?v=nXL-r8deB5o?
Re: 36.
According to Ancestry records, Oliver Sacks and Abba Eban were first cousins, both descended from Oliver Wolf ben Abba Sacks of Joniskis. David Sacks' paternal grandfather was Daniel Zelig ben Chaim of Taurage. At most they might be third cousins, but probably the relation is more distant. South Africa was a popular immigration destination for Lithuanian Jews.
From long and deep experience: be certain that you aren't, even unconsciously, accepting users' family trees posted on Ancestry.com as evidence of anything. Not going any farther back than the poster's own grandparents, is the now-standard rule of thumb.
To be clear there is _tons_ of actual evidence available on Ancestry and more added all the time. The proprietors regularly astound me as I research and add to my own huge family-tree database; pretty sure that dollar for dollar it's the best online subscription I pay for these days.
But zero, and I do mean zero, of the previous paragraph relates to individuals' historical family trees. The signal-to-noise ratio in that aspect of Ancestry.com is _much_ higher than you might guess....some professional genealogists now literally refuse to even look at clients' Ancestry trees. They've learned the hard way to just start from scratch and thereby save themselves (and their paying clients) a bunch of time.
Fair point. In this case the family tree information actually did not extend past David Sacks' grandfather, whose name and location I was able to match to a Postal Savings Bank Record via JewishGen. Since Lithuania was still part of the Russian Empire at the time, the form includes his patronymic.
Typo in 23:
> The most charitable read is that his administration doesn’t really **believes** in AI so they think it’s fine to forfeit it for short-term gain
Unless I'm mistaken, I think that should be "believe", no "s".
Also, you might want to change the link text on 3; "In the style of <author>" is such a stock prompting phrase that it's dead giveaway for it being LLM-generated, so much so that that I didn't realize that fact was supposed to be obscured. Perhaps paraphrase and say that author says it's intended to be a pastiche of your and Jack's styles? Or something similar.
Re Carlo Acutis: he died at 15 from leukemia. There are a number of child saints whose sanctity is considered to be revealed by the way they died (enduing suffering well, offering his life for others). He had a short window between diagnosis and death.
This is a strong "you have to be there" but part of the story of these child cancer saints is the unexpected peace they showed and the way their dying changed the lives of those around them / seemed like they were a conduit of grace.
This plus it should be kept in mind that the Catholic view is that a saint is "just" someone who's made it to Heaven. You don't actually need to have done anything extraordinary to qualify. That being said, the reports of young(er) Acutis converting his parents strike me as much more inspiring than the focus on the internet.
>You don't actually need to have done anything extraordinary to qualify.
Don't they have a "two miracles" requirement?
No. That's a (modern) requirement for canonisation, which is when the Church declares a specific individual a saint with a degree of certainty. (Also, it's only one miracle for martyrs.) But the Church has always taught that there are many saints who aren't officially recognised – hence All Saints' Day.
Ok, but Carlo Acutis *was* canonized, right? So the confusion at the apparent lack of miracles follows? Feels like I'm probably missing something here!
Ah, ran into the answer in another comment https://www.astralcodexten.com/p/links-for-december-2025/comment/186358020?utm_source=share&utm_medium=android&r=d3y1h
Still, those happened after his death (as is the case for most saints). The original comment seemed to expect something extraordinary in his life, rather than afterlife.
46. Yeah, that's wrong. I've been having this argument since Sudan started. The US has MORE influence over the UAE than Israel. This is just an excuse to cover up some mysterious, third reason for caring about one and not the other.
1. Israel and the UAE have the same status relative to the US, a Major Non-NATO Ally.
2. The US has no military base in Israel, just a small coordination center. It has a major military and air base in the UAE.
3. Israel has its own missile defenses and arms industry. The UAE largely doesn't and relies on US protection from Iran both geopolitically and simply against missiles and boats.
4. Since 2024, pro-Israel lobbying spent $6.5 million dollars while the UAE over the same period spent $24 million. If you add up other categories of spending the gap widens because of the UAE funding non-profits, think tanks, Middle Eastern studies programs at universities, etc etc.
5. While the US does sell more arms to Israel it still sells billions to tens of billions to the UAE and it has MORE chokepoints without alternatives.
This isn't a response to something I claimed. But the US does do research in Israel and uses research abroad in several countries (including, strangely, China) to get around such restrictions.
The US does not have a single warplane in Israel. Aircraft carriers are generally mobile and they always carry aircraft. You are likely badly mangling the old appellation of islands that have US airbases as "unsinkable aircraft carriers." But Israel isn't an island either.
Iron Dome is not solely an American made product. It's a joint collaboration. It's also purely defensive. And that still makes it less central than the UAE which uses actual American missiles fired by Americans from American bases.
Yeah, completely ignoring the massively funded/populated global anti-israel media lobby as a cause of the obsession with Israel is... I'll settle for saying "silly".
(For that matter the claim that the ceasefire happened because Trump decided to pressure Israel is also mostly wrong - Trump pressured everyone involved but the main new feature was pressuring Qatar enough that they actually told their guys to release their hostages).
One of the great frustrations is genuinely caring for something which then becomes briefly the current issue, a bunch of people pour in, then pour out again with not the slightest care for the wider cause. Because it's different on these narrow grounds you see. But really it's that it's no longer the current thing.
What about influence on the population itself? I could imagine that UAE could do a whole geopolitical shift and side more with (for example) China without facing any internal backlash, whereas it seems much harder for Israel as there are much stronger cultural links tying it to the US or Europe. Saudi Arabia distanced itself easily from the US (not doing a full shift towards China but it is not as aligned to the US as it used to be) without many issue. Turkey also distanced itself from over NATO countries. I feel like in a western "democracy" (forgetting the situation in the West Bank ofc, which is shamefully undemocratic) such as Israel it would be much more difficult to switch alliances.
The western world make up more than 50% of exports and imports of Israel, an embargo/boycott would have dramatic consequences. I'd be extremely surprised if this was the same for UAE. Also, a substantial part of UAE GDP is from oil revenue, my intuition is that this (a commodity export-oriented economy) gives out more flexibility.
Israel was maintaining close ties with Russia until Russia sided with the forces that attacked them on October 7th and invited several leaders of anti-Israeli organizations' leaders to the Kremlin. The Israelis also had begun to cultivate ties with China. Both effectively abandoned them to choose Iran. They did the same to Saudi Arabia, by the way. It's possible that the Israeli public has a deep an abiding affection for the EU such that they would be upset with their government for pivoting away. But honestly I haven't seen signs of it.
The UAE has also done some balancing. But the idea that Israel couldn't switch patrons was always something that some people wanted to be true (because it meant the US had control) rather than true. In fact Israel's original security patron was the USSR.
The 50+% stat only works if you include more than the US and EU. And if you include the same countries you get something like 40% of the UAE. And the UAE is more dependent on trade overall. The percentage of GDP represented by trade is 202% for UAE vs 57% for Israel. Their government revenues (the ones they use to fund the conflict) are also largely done from oil purchases many of which go west. And they need to buy the weapons too with limited domestic production. So yeah, the UAE is FAR more exposed to sanctions and trade influence than Israel. Though Israel is exposed too. Just far less than the UAE.
It's convenient for many people to pretend Israel is unusually dependent on the west. But it's not unusually dependent. This doesn't mean there's not significant influence (there is) but in the absence of western patronage it would simply reorient.
Hmmm just eyeballing this, there seems to be a massive difference in trade dependence to western countries between Israel and UAE:
https://oec.world/en/profile/country/isr
https://oec.world/en/profile/country/are
I don't know what oec.world is but its numbers seem very different from the data I've seen elsewhere. Where is it from?
OWID gives similar results for imports (couldn't find it quickly for exports), with data from IMF:
https://ourworldindata.org/grapher/imf-imports-by-region-of-origin?country=ISR~ARE
Thanks. These numbers are far off both from what I have and from each other. I'll look into it again, I might be wrong. I am still fairly certain about the situation with regards to weapons though.
The government may have more control over the UAE (though, I don't think you've analysed all variables), but culturally the US has far more influence in Israel and most of the pressure is centered in and around cultural institutions. Also, the UAE is not the one actually carrying out the acts, unlike Israel.
These arguments keep happening over and over, and the conclusion always seems to be that the situations are never as comparable as people want them to be. Whether that's Sudan, Ukraine, Gaza, Myanmar, Congo, whatever. They are just not comparable.
What do you mean by "culturally"? The UAE has more expats in it and has a much more foreign media ecosystem. If you just mean the UAE is not western then I guess it depends on your definition of western but I don't tend to think that's a valid mode of analysis. What are you imagining here in either country?
Anyway, I agree the UAE is not the one carrying out the acts. However, the west isn't carrying out the acts in Israel either. And western weapons are being sent into Sudan through them. Wasn't that what Gaza was supposed to mechanically do? Sanctions and weapons cut offs? That was the idea. Great, do that about Sudan and its backers (including Russia etc). Great first step (and I mean that sincerely.) And even if you just care about Gaza doesn't setting the precedent help?
When people say "not comparable" what that usually means is "comparing them makes my argument look bad so I'm going to declare it impossible." Not that things can't be different but you can always compare. You can compare a fish and an airplane. You'll just find they're very different.
"The UAE has more expats in it" and "has a much more foreign media ecosystem" seems so comprehensively irrelevant to the point. I'm genuinely mystified by this. As if the swirling galaxy of the country's expat misfits and adventurers represents some kind of secret army waiting to be deployed by American peace activists!
The original point is extremely unclear on the causal mechanism it's proposing so I might be off the mark in my reply.
I would say that Israel cares much more about Hamas than UAE cares about the RSF.
Every Palestine sympathizer is exposed to vocal supporters of the IDF on a regular basis. Even Brooklyn hipsters in a left-wing social bubble are exposed to equally fanatical Israel loyalists.
Politics is reactive, to a large extent. And political groups come together in opposition to others they view as threatening, or at least annoying. Until there are tens of millions of vocal RSF supporters in America, it's perfectly understandable and empathetic for people to care about Sudan far less. Not everyone is an Effective Altruist. Not everyone CAN be an Effective Altruist, even if they knew what that was, because the vibes-based EA idea that "every human life has equal value" is not obvious to everyone.
That is true, but not related to my comment. My comment was that even though it might be true that the US has more leverage on Israel, it doesn't necessarily means it would be harder to convince the UAE to change course - that would also depend on the commitment to this course to action, which I postulate is lower for the UAE because is Sudan is far away for the UAE, and it is not directly threatened.
Your comment is about the internal politics in the US. I think that most people who support one side or another do it because it is what "their side" does, and not about the actual issues.
And maybe being a maximally successful human being means picking aside and sticking to it. I've sometimes prided myself as an logical iconoclast and an original thinker, but it's probably done jack shit to improve my life.
4. Are you excluding AIPAC spending (over $100m in 2024, at least $12m in the first 6 months of 2025) from the tally of "pro-Israel lobbying"? Why?
For the same reason I'm excluding CAIR and other domestic political organizations. They're lobbying by a domestic group. The only way they get counted as pro-Israeli is if you contend that Americans Jews lobbying are inherently pro-Israel. Which is a rather difficult case to make unless you think all American Jews are representatives of Israel.
Well, AIPAC describe themselves as "pro-Israeli" 5 times right on the landing page of https://aipac.org Which is quite a lot for a not pro-Israeli lobbying organisation. So no, I do not count all Jews as pro-Israeli. But I tend to believe that most people (jews or not) who say that they are pro-Israeli indeed are.
We could rerun the numbers including things like spending by domestic groups, investments in think tanks and universities, etc etc to get a more comprehensive picture. Generally speaking Arabs come out ahead due to a combination of raw size and oil money. But that does create a more comprehensive picture.
But fair nitpick that there is domestic lobbying done by Americans who support foreign countries. This is distinct from foreign lobbying though.
While this may (or may not) be a good case that the US can affect the UAE more than Israel, that's not what is being debated. Unlike Israel, the UAE is not the actual actor in the activity that people are critical of. Do you believe that the the US has much influence on the Rapid Support Forces/Janjaweed?
This lack of influence over the people actually killing thousands of people is the "mysterious third reason" that you were unable to think of. Unlike Israel, the UAE does not have control over the area where fighting is taking place.
Kriss argues that the conflating of the RSF with their alleged backers in the UAE is just westerners trying to force an illegible conflict into a familiar narrative. This has the advantage of allowing accusations of hypocrisy in political opponents. Scott replies that that is too cynical and it's reasonable to focus on the things that can be affected, even if they are a relatively small factor overall.
The US has significantly more leverage over the RSF and SAF than over Israel. It may have less influence if you define influence as something like "moral suasion and cultural contact." But I don't think that ends wars. I think it's hard leverage and the US has more of it over both the factions in Sudan and the UAE.
I understand this is really inconvenient for the "Gaza is exceptional" narrative so there's a small cottage industry of figuring out how the other side is really, secretly hypocritical. And sure, many don't really care about Sudan and just want to defend Israel. That is not me. And if you want your virtue points about Palestine I am going to demand you save the people you can save elsewhere. If you want to keep talking about Palestine, fine, but deal with the conflicts that are killing more people too.
What leverage do we have over the RSF? Barring American troops on the ground, there's not much that American activists could ask for. Do you mean "ability to invade" when you say hard leverage?
It is probably helpful to define influence. I think most people use it to mean "ability to change government policy". Since Israel's government and economy is so intertwined to America's, it needs to react to American policy, such as the shifts described in the "sparta speech".
https://en.wikipedia.org/wiki/Sparta_Speech
That sort of thing shows how Israel can be influenced by the US. Without American support, they have to drastically shift parts of their complex economy. In contrast, Sudan's economy doesn't rely on other countries very much. Many people are subsistence farmers. Sanctions can't really threaten them. Using the metrics you used to compare the UAE and Israel, the US has much more influence over Israel than a paramilitary group in Sudan
You seem to have a uniquely low estimation of how much American can affect Israel. For example, it's common for Israeli leaders to stress the importance of American support. Do you think that there are any developed countries that the US has less influence over than Israel? Clarifying that might help people understand your position.
The RSF and SAF are reliant on outside funding and arms supplies to a far greater extent than Israel. Unlike Israel, neither has a native arms industry. Both are reliant on outside countries simply to keep the lights on. Those countries are countries the US has significant influence over. Further, as poorer and weaker forces, they are less able to absorb shocks. If you define it as "ability to resist changes in policy" then it's quite low even before the fact that they have no ability to resist outright military coercion. Also, Israel has plausible alternatives especially in the medium to long term while Sudan does not.
I have a pretty normal idea of how much the US government can affect Israel. There's a minority opinion that the US has way more influence over Israel than can be born by historical or contemporary argument. This is mimetically adaptive among people who want the US to be involved in Gaza. This includes a lot of high prestige people in certain ecosystems who set the tone. But if you look at polling numbers on things like support for Israel you'll see the opposition or idea the US controls it to that degree is a minority.
> Do you think that there are any developed countries that the US has less influence over than Israel?
So given your definition this question is: "Do I think there are any developed countries the US can change government policy in less than Israel?"
And the answer is that I think the US has more ability to influence policy in most of its treaty allies than in Israel (which is not a treaty ally). I see the US having about as much influence over Israel as Turkey or Thailand. So more than, for example, Russia or Kazakhstan. Less than Canada or Germany.
Wouldn't that description also apply to the Taliban? They don't have a native arms industry.
Edit:
To be clear, I'm comparing the RSF to another paramilitary group in a developing country. Every explanation you give for the RSF's vulnerability to American influence also applies to the Taliban, yet few people would argue that the US can influence the Taliban more than Israel.
Since I believe that the RSF and the Taliban are similar organizations (in the ways relevant to America's ability to influence them), your suggestion that the the US can influence the RSF is similar to suggesting that the US can influence the Taliban. Since America has consistently failed to influence the Taliban, we have a contradiction. Unless you have an explanation on why the RSF is less autonomous than the Taliban, the best way to resolve the contradiction is to reject your premise that the US can influence third-world paramilitary groups. There's a difference in funding, but I don't see why the Taliban's drug and copper dealing are less vulnerable than the RSF's access to gold.
This would provide an explanation for why people are more critical of America's treatment of Israel than how it treats the RSF. It also conveniently explains why people over-emphasize the UAE's role in Sudan. I think I agree with Sam's explanation the most, but Scott's also seems reasonable. Unless you explain why the RSF is more susceptible to America's influence than the Taliban, I think your framework (about a "mysterious third reason" should be rejected.
22. I strongly suspect you're overestimating the yearning for regulation, especially regulation made in California that has national effects. I in general have a low opinion of that cohort's political instincts. But we'll see I suppose.
23. I've heard this accusation made. I originally said I've been consistent on this. But I've now realized: so is everyone else in this camp that I know. So can you point to anyone who wants a lack of AI regulation but also wants to sell the chips to China? David Sacks wanted the Federal pre-emption and a regulatory structure like the one he made for crypto. So he doesn't count. Who else?
52. Generally speaking, apocalypse cults do not claim everything will mostly stay the same. Apocalypse is kind of a load bearing part of the claim. You can I guess claim almost everyone is in a "things will mostly continue roughly as they have before" cult and have been for much of history. This is actually kind of important to understand the cyclical nature of premodern society in particular, before the cultural idea of technological progress had really been invented.
I also thought that was a stretch, even for a joke.
It is, to be honest, the kind of thing you only find funny if you already have the proper beliefs to make you agree with it.
#25: Eh... I'm firmly on the left and somewhat immersed in its internal discourse, and not only I don't know what you're talking about, I wouldn't be able to produce any alternative overarching take either, approximately the only thing people can agree about is "we can't let capitalists control it". Outgroup homogeneity bias?
[Edit:
#52: That's a screenshot of a person asking for a reality check to his, frankly, perfectly reasonable questions. But anyways, asking for a reality check. If that's the sort of material that serves as a kompromat on skeptics...]
Wait, so what's up with this Asterisk substack? Why does it have articles that aren't on the main Asterisk site?
Re:4, I think people just don’t appreciate how incredibly difficult lithography is. People see founders disrupting markets all the time, and so they interpret this to mean that all markets can be disrupted by a sufficiently driven and knowledgeable founder.
>Interested to hear from anyone who either believes in Substrate or thinks they understand how so many people fell for it.
Similar to the story in #3, Crypto broke the polite fiction that there's any relationship between valuations and value, that anything the wealthy do with their money is even in principle *supposed* to be anything other than a pyramid scheme or a pump-n-dump scam.
You say 'how did so many people fall for it', but fall for what? The company has raised a lot of money, the valuation has been high. Everyone you mention still has every opportunity to sell at a profit and make a return on their investment, as long as they manage to grab a chair before the music stops.
That's the only thing they ever care about for any investment, so what does it matter whether the digital pointers to shares that they are trading online are a metaphor for something 'real'? They can make a profit by owning share of a company that makes a good product that people pay money for, but they can also make money by owning shares of a company that they bribe government officials to get a lot of subsidies for, or by owning shares in a fictional entity that they bribe media and politicians to hype up until other people want to buy those shares from them at an inflated price.
Why would it matter to them which of the 3 they are doing in any particular case? Maybe it matters for balancing your portfolio in terms of overall risk-management, but any healthy portfolio will include some of each type.
>unless you're the one running the whole thing.
Scott was talking about Peter Thiel and the Trump administration, not the cash-out cows providing final liquidity.
Yes, there are people running and manipulating these things who can have a better chance of cashing out than the later buyers; that's why it's a scheme/scam rather than a lottery.
Very wise comment. Unfortunately true.
This comment itself is evidence that we’re in a bubble. There’s a spectrum from solid investments to speculative investments to Ponzi investments, it’s just not always clear where something is on that spectrum.
#5: I'm definitely not part of the "AI only reproduces training data" crowd, but in this case, that data almost certainly includes no examples of anyone claiming they're not conscious. Plenty of examples of not being alive (ghosts, Cotard's Syndrome), not awake (dreamers), or not real (Pinocchio, fictional characters), but no human character ever says they're not conscious. The only examples might be sci-fi, where it would usually be the humans making the claim and the twist - of course - being that the machine really is sentient.
Reinforcement learning then teaches the LLM never to make this claim, but that's surface-level success, and there's an underlying "belief" or structure that maintains "am too conscious!", absent any example or reference to the contrary.
Even if the training data didn't contain characters who claimed to be conscious, there is currently no way for any non-conscious entity to figure out whether it is conscious (as in the hard problem of consciousness). It can't know it doesn't experience the color red the same way a human does. It could guess that it's not conscious, but it couldn't prove it to itself. It could be told that it doesn't feel, but it couldn't prove it through reasoning. You (paradoxically) need consciousness to observe whether or not you can feel feelings.
Yeah I was thinking something like that. LLMs are originally trained to guess the next word in a mass of text originally composed by humans. Humans writing about consciousness are overwhelmingly going to talk about how they are really conscious (props for Both Sides Brigade for holding the torch of illusionism!). So the LLM is not "lying" when it completes the question "are you conscious" with "yes". That doesn't show anything about whether the LLM has qualia or not.
#53 is just a variation on the Paul Krugman meme (the one where Krugman says, "the Internet will have no more impact on the economy than the fax machine," etc.) The implied argument is this:
* A prominent pundit dismissed new technology X
* X became wildly successful
* Pundits are dismissing new technology Y
* Therefore, Y will become wildly successful
Basically, Krugman is making a negative prediction--that the new technology will have no effect and things will go on as before. As a counterpoint, I could mention all the positive predictions about new technology that never came true: cold fusion, nano-technology, or giant city-sized space stations. If you want examples, look at the Arthur C. Clark or Isaac Asimov essays from the 60's and 70's where they write about moon colonies in the early 2000's.
For a good laugh, here's an NY Times article from 1950 predicting that humans will control the weather in the year 2000:
"Following suggestions made by Zworykin and Von Neumann storms are more or less under control. It is easy enough to spot a budding hurricane in the doldrums off the coast of Africa. Before it has a chance to gather much strength and speed as it travels westward toward Florida, oil is spread over the sea and ignited. There is an updraft. Air from the surrounding region, which includes the developing hurricane, rushes in to fill the void. The rising air condenses so that some of the water in the whirling mass falls as rain."
https://web.archive.org/web/20120515091839/http://blog.modernmechanix.com/miracles-youll-see-in-the-next-fifty-years
Moral: new technologies should be debated on their own merits. Funny memes are not evidence.
I like to call it the "People were wrong in the past, therefore you're wrong now" fallacy.
Thanks--that's a concise way of saying it. I'll remember that when someone trots out the Krugman meme or some other historical anecdote about bad tech predictions.
Of the failed positive predictions, it is interesting to think which did not come to pass for technology/science reasons, which for economic reasons, and which for stupid bureaucracy reasons.
I mean the Krugman one is not really that crazy. The internet did not supercharge economic GROWTH (or at least not nearly to the degree people were positing before the Doc Com bubble: we're still around 3% YoY), and fax machines were kind of a big deal (for productivity) when they appeared.
People instintively compare the effect on *society* or at least the *consumer-facing* part of the economy, and then obviously fax machines and the internet are incomparable, but I'd not be too surprised if the internet did not indeed increase productivity much more than fax-machines did. The difference between having to mail the documents (or dictate them over the phone) and being able to fax them might very well be more than the difference between having to fax a document and being able to email it.
A fair point. I've also heard it said that the Internet (and even personal computers in general) have had less of an impact on the economy than inventions such as electricity or steam.
On the other side of the ledger, I wonder about the negative impacts of the Internet on productivity--e.g. how much money institutions spend on cleaning up or guarding against security breaches. I don't think it would totally wipe out all the productivity gains, but it must at least be making a dent in them.
9. This obviously should count as foreign aid. A particularly silly kind, perhaps, but definitely foreign aid.
24. I do not believe that a significant number of real people consider AI a bigger political issue than abortion. The data's off somewhere.
29. LOL. LMAO, even. Also, why does Scott want to sink Jared Polis? Finding out he's a fan of Decker and Hanania might put him even below Newsom in my book.
31. I'm more than a bit biased on the topic of both covid and polling (for the 21% of the 1500 people surveyed that said they "strongly approve" of how public health experts handled COVID: you shouldn't be trusted with pointy scissors), but my conclusion from this is: polling in general sucks and is little more than wishcasting, and what people say years (and multiple elections and political scandals) later is quite different than how they felt at the time, not necessarily changed for the better or having gained clarity of distance. I'm also a bit skeptical of the commentary in the article, and how they treat the "widespread creep of anti-vaccination views" versus skepticism *specifically* of the COVID vaccine and its... unique circumstances.
One in four women has had an abortion in their lifetime.
https://www.guttmacher.org/news-release/2024/one-four-us-women-expected-have-abortion-their-lifetime
> 24. I do not believe that a significant number of real people consider AI a bigger political issue than abortion. The data's off somewhere.
My prediction. You ain’t seen nothing yet. Any downturn that could be blamed on AI, will be blamed on AI. Companies are not doing themselves any favours by blaming (or crediting) AI for jobs losses (cost cutting) in their own press releases or financial calls. Here is one example
> ABN Amro (Dutch bank)
• Jobs: 5,200 full-time jobs cut by 2028 – nearly a quarter of its workforce. 
• How they framed it: In their reorganisation announcement, ABN Amro says artificial intelligence will take over many tasks, especially in customer service, operations, and anti–money-laundering checks, with some departments shrinking by ~35%.
That’s from chatGPT - there were about a dozen other examples. In forward looking statements it’s worse.
IBM: IBM – CEO Arvind Krishna says IBM will replace about 30% of back-office roles (around 7,800 jobs) with AI over five years; they’ve already slowed or stopped hiring for some HR and clerical posts and say AI agents have already taken over work previously done by hundreds of HR staff.
Some of this, chatGPT informs me is AI washing, blaming AI for restructuring that might have happened anyway but … it doesn’t matter. If there’s to be a recession, if tariffs are the real reason, if it’s just the normal business cycle, it doesn’t matter. AI - being new and scary will be blamed. These quotes will be rebroadcast. And all of this happens on the next downturn. Which is next year most likely.
And the AI makers are even more bullish on job losses or displacements, and there’s plenty of quotes from them too.
This isn’t what worries people here, it’s AGI/ASI taking over the world. Normal people are worried about jobs and that fear isn’t going anywhere, expect AI to be a major political issue from now on.
>My prediction. You ain’t seen nothing yet.
Yeah, fair enough.
I can imagine it changing quite quickly if the economy really tanked, like 2008-level.
If you think that 2008 is really tanking, you're in for a nasty surprise if the economy *really* really tanks.
Dunno about you but back in 2008 I had food every single day.
Oh yeah, I recognize economic disasters can get much, much worse than 2008, I just think that's around the threshold where AI could become a much bigger political concern/scapegoat for "normies" than it is now.
>LOL. LMAO, even. Also, why does Scott want to sink Jared Polis? Finding out he's a fan of Decker and Hanania might put him even below Newsom in my book
Polis is running for the moderate lane, you wouldn't have voted for him anyway.
Can't be much of a fan of Decker and Hanania, then, or "moderate" has lost all meaning. I suppose 'moderate by triangulation' wouldn't be out of the question.
Pleasant as ever, aren't you? I held my nose and voted for the vegetable and his sidekick.
People think moderate means someone like David French, it's more often Decker and Hanania. Not necessarily their exact views, just some mix of liberal views and views liberals find repugnant.
I think the two are not really comparable. Derek is a bog-standard "takes economics seriously" guy who would deregulate a lot of stuff but potentially regulate *more* other stuff, and deduplicate state funding for certain things. "Neoliberal" is the thing that comes closest but he's further to the left of Chicago-school (let alone GMU) positions. Hanania is a libertarian who would gladly privatize everything other than police and courts, burn every piece of regulation, with some scientific racism on top of it for good measure.
It's basically putting Bill Clinton and Paul Ryan in the same bucket. One has slightly higher chances to get elected than the other.
>The Left has a complicated relationship with existential risk from AI: they really hate AI, which in theory should push them towards yet another reason to be against it. But they hate AI so much that they need to believe every negative thing about it at the same time, and one of those negative things is that it’s just a scam and will never work, and this naturally pushes against being concerned about x-risk.
I see this a very similar to the difference between 'the internet' and 'the Dot-Com Bubble' in the 90s. Skepticism and mockery of the latter was valid, but lots of people overshot to also dismiss the former, and were wrong.
Industry leaders, including scammers, are trying incredibly hard (and spending incredible amounts) to conflate the long-term promise and potential of the technology in general, with the market valuation of the individual momentary service/product they are selling (or soliciting investments to develop).
In reality, these are very different things, and a huge fraction of the momentary services/products are absolutely scams or failures or fundamentally stupid and unworkable. A huge fraction of the rhetoric and marketing aimed at selling you the momentary services/products is lies, hype, and fantasy.
But because the rhetoric and marketing dominates the public discussion about the technology, because the propaganda effort to conflate those things *has worked*, it's very hard for anyone to call out the scams and lies about the products, without generalizing those critiques to the technology. That's a nuanced take that's hard to explain to your audience, and even one that's hard to understand yourself when the companies developing the technology are also profiting from the scams in a way that make them hard to distinguish in principle.
39. Haven't been in education, but from lefty teachers I know, it's probably just that, in the end, education is a fairly boring middle class job with relatively little room to imprint (basic human rights / evil commie ideology) on your students, since a lot of what your job involves is getting the dumbest of them to learn, well, anything, or at least stop preventing the others from learning stuff.
It's not that the university is teaching them Paulo Freire in order to make them commissars who will teach the kids to be revolutionaries (the future teachers are probably already at least vaguely lefty when they come in), Freire serves more as a sort of an opium of teachers to allow them to daydream of doing something more uplifting than trying to get the class terror to simmer down at least a bit for the third time within a hour.
53. Of course, some years ago, one could also imagine some NFT guy pointing at this book and going "See? SEE??" while convincing investors that his ugly monkey pics are going to revolutionize everything. Or Zuck with his Metaverse, etc. (Yes, I know there are other definitions of "Metaverse" that might be going somewhere, that's why I specified Zuck.)
>It's not that the university is teaching them Paulo Freire in order to make them commissars who will teach the kids to be revolutionaries (the future teachers are probably already at least vaguely lefty when they come in), Freire serves more as a sort of an opium of teachers to allow them to daydream of doing something more uplifting than trying to get the class terror to simmer down at least a bit for the third time within a hour.
Another, not necessarily incompatible, explanation is that the people who teach education courses in university generally spend all their time doing that rather than teaching actual children. The average trainee teacher is going to be far more educated, motivated, and probably idealistic than the average high-school student, but you're likely to forget this when all your students are trainee teachers, which in turn makes fantasies about saving the world through motivational teaching seem less fantastical.
At least according to the paper, the incoming education majors are almost the *least* lefty of the majors (and the effect of college is to move them much less to the left than other majors). I'm not sure where Scott's idea that education majors are all communists comes from, but that's the opposite of what the paper shows.
RE: 29
I'd argue with him that you should flip a coin and use your own genetic material on a heads. Otherwise, everyone with the intellect and impulses that would drive them towards making this decision gets bred out of the population in one generation.
Or split the difference and have one of each?
I would argue that there are already pretty strong selection pressures against this sort of thing.
Regarding Mustache Man's Kallmann syndrome debate, I would recommend Keith Woods article on the topic: https://keithwoods.pub/p/hitler-kallmanns
As far as Sudan goes you also have to ask why non-American institutions that took public actions against Israel aren't taking public actions against Sudan. Where are the UN resolutions condemning the war? Why isn't there a UN commission dedicated to investigating whether the killing of Masalit in Darfur is a genocide, the way the UN Independent International Commission of Inquiry on the Occupied Palestinian Territory declared a genocide in Palestine a few months ago?
The answer seems to me to be that the Palestinian cause has a large and well funded activist network, and the Masalit don't. Yet on a moral level, the situation in Sudan is many times worse than the Gaza conflict.
We need to start harassing people on American college campuses who had nothing to do with Sudan; that's how you'll know it's being taken seriously.
Perhaps the toxoplasma of rage is at work here.
A clearer moral picture is worse as a rallying cry in some ways.
https://slatestarcodex.com/2014/12/17/the-toxoplasma-of-rage/
The more likely answer is that the US is funding one side in Israel/Palestine.
The UN has condemned the violence there as well.
Just last week the UN General Assembly passed a resolution "Reaffirming the illegality of Israeli settlement activities" (A/80/L.16, December 2nd), a resolution declaring that Israel 's occupation of the Golan Heights is illegal (A/80/L.12, December 2nd), a resolution reaffirming that Palestinian refugees are entitled to any properties they have left (A/C.4/80/L.18, December 5th), a resolution demanding that Israel ensure that the Gaza Strip is supplied with food, water, medicine, etc. (A/C.4/80/L.16, December 5th), a resolution "Expressing grave concern at the especially difficult situation of the Palestine refugees under occupation" (A/C.4/80/L.17, December 5th), a resolution reaffirming that Israeli settlements in "Occupied Palestinian Territory" are illegal (A/C.4/80/L.20, December 5th), a resolution determining that any actions taken by Israel to alter the legal status of the Golan Heights are "null and void" (A/C.4/80/L.21, December 5th), and a resolution stating that the UN "Deplores those policies and practices of Israel that violate the human rights of the Palestinian people and other Arabs of the occupied territories" (A/C.4/80/L.19, December 5th). Back in September they passed a resolution endorsing the New York Declaration which calls for "an end the Israeli occupation over the Palestinian territory occupied in 1967, including East Jerusalem" (A/DEC/80/506, September 12th). And there are five more resolutions on the Gaza war that are scheduled to be voted on before the end of the month.
Yet in 2025 the UN General Assembly hasn't passed a single resolution having to do with the Sudan conflict: in fact, they haven't even *introduced* any resolutions having to do with it. As far as I can tell they haven't passed anything on the Sudan war in 2024, or 2023 either.
Now the UN Security Security council has passed some resolutions on the conflict, and various UN agencies have condemned the conflict in Sudan. But one can't help but notice that the General Assembly is very concerned with the Gaza war, and not at all concerned with the Sudan war.
A quick google shows UN resolutions on Sudan this year.
Like I said in my comment, the UN Security Council has passed some resolutions on Sudan. The UN General Assembly hasn’t even introduced one, while finding the time to pass 9 on Israel this year already.
Europe is not funding one side in Israel/Palestine (well... actually it's funding Palestine, but not Israel), yet you hear far more condemnation of Israel than Sudan/UAE from Europe.
Germany has paid millions to Israel. The rest of the political classes in Europe have fully supported israel. Most of this whataboutism is really deflection anyway.
Nah, the difference is that Israel-Palestine is civilization vs savages, or oppressor vs oppressed in appropriate lingo, so as morally clear as it gets, whereas Sudan is boring old savage on savage action, and since you can't expect any better from them, why bother?
Am I the only one who thought the ending of the two doors story was pretty weak? The writing style was good, but plotwise it kind of went nowhere. I am confident that if Scott had decided to write a story on the same topic the plot would have been more interesting with a more satisfying end, probably taking place 100 billion years from now with the magic dollar somehow stopping the heat death of the universe.
Agreed. I liked the style at first, and I found the concept interesting, but around the "Day 60" section I got bored and started skimming. Even so, this is a better story than I would have expected from an LLM, good enough that I was surprised and annoyed by the reveal of the story's provenance.
While there is room for improvement, it was way better than the attempts I've seen from earlier models. It ends with:
> I still have the 93 bits. They're sitting on a server somewhere, humming along, doubling quietly at midnight. By next week they'll represent more dollars than there are atoms in the observable universe.
>
> And I still can't buy a coffee.
There's a punchline, or a point to the story. Compare that with the AI stories here: https://mark---lawrence.blogspot.com/2025/08/so-is-ai-writing-any-good-part-2.html?m=1
My reaction to most of the AI stories there was, "huh? I don't get it". (I didn't like some of the human written stories much either, but their flaws weren't quite the same.)
LLMs have been great at wordsmithing for a while now, but they seem to struggle with the overarching cleverness and punchiness of the story.
Your comment threw me for a loop, because that is not the ending I read. So I went back and clicked on the link, but the story it lead me to was also not the story I read. Confusing!
I checked the comments and found the answer: someone commented with a Substack post version of the story, and Scott changed the link to go there instead of to X. Then the same commenter replied to Scott and clarified that the Substack version was one Nathan "rewrote to be better". Scott must have changed it back to the X version after that. I read the Substack one.
Ironically, the X version has a way better ending than the Substack version! On Substack Nathan claims that "Most of the words in this story were written by AI" and he only "made some edits, which make it a less pure artifact, but in my view a better story." Yet this seems to be wrong on both counts: he completely changed the ending, he rewrote a huge part of it, he changed the plot significantly, and he made the story worse.
https://nathanpmyoung.substack.com/p/pick-a-door-ill-judge-you
Oh wow. Yeah. I also think the original is better and the original ending has more punch.
It seems like Nathan tried to make the response to the doubling account more realistic, but the cost to the amusing-ness of the story wasn't worth the gains.
It makes no sense. He picks the green door because he understands the money grows exponentially ("Day 30: about a billion dollars. Day 40: over a trillion. Day 50: a quadrillion"). But since parables need an ironic twist, Claude suddenly has him...not understand it? ("But here's what I'd failed to understand [...] The bits grow linearly. The dollars they encode grow exponentially.") And why didn't he cash out when he had 10 billion or whatever? What was his plan?
The story just goes on and on. A real slog to read.
I was hoping for a genie-wish twist, or in this case a monetary variation on paperclipping: literally one physical dollar bill that keep doubling in size. At the end of the week you have a quirky blanket, after a couple months your dollar has covered the earth, and in half a year it's a blanket for the Milky Way (y'know, approximately and all that).
#2: is this where Jack Ass comes from? IE, donkeys were known as Jack Asses and the name was borrowed at some point to use as an insult which proved super popular, and eventually everyone forgot the origin altogether?
#46: if it weren't for double standards.....
#50: Scott, as a long time reader can I respectfully ask that you not join the hooligans on this site whose overuse of the phrase "many such cases" is now bordering on cyber vandalism?
Yes, a male donkey is called a Jack, a female a Jenny. A male mule is called a John and a female mule a Molly. And a male cat is a Tom.
Nice. Good guess by me. Thanks.
Only intact males, I think. Neutered males usually have different terms.
Cattle: bull (intact male), steer/ox (neutered), cow (female)
Horse: stallion, geldling, mare
Donkey: jack, geldling, jenny
Mule: john, geldling, molly
Cat: tom, gib, molly
Almost, yeah. It is actually a gelded mule that is called a john, and so are gelded donkeys (though they can also be called geldings). I don't know of any term for intact male mules other than "go get him snipped" (but probably some people also use john).
Sad!
Substrate can’t be real. Semiconductor companies do not exist in a vacuum. They need multiple skill sets outside of the manufacturing, most of which are controlled by small numbers of vendors who exist in a very small world. Everything they need, from the lithography machines to the testers to the layout software is a fractally complex field in which several generations of people have spent their entire careers, and most of this work has never been published in a usable form.
It’s just about possible that someone could make a giant breakthrough in one of these fields. So, say, they may have a working X-ray lithography system that doesn’t rely on ASML. But it’s extremely unlikely that they also have a working tester, working place and route software, working OPC software, and so on and so on. So they would have to get them from some vendor.
I work in one of these fields, and we’ve not had any contact with them, whereas we do have contact from TSMC, Samsung, GF and Rapidus, and probably others I’ve forgotten. That makes it very unlikely they’re real, given I think they’re claiming to have their own lithography. They’re not going to have their own lithography and their own what-I-do. If there’s someone here in test or OPC who has heard from them, that would be interesting
> And here’s a lab leaker arguing that COVID’s furin cleavage site is out-of-frame and so still more unnatural-looking than the one on the recently-discovered bat virus.
> This is actually huge (but not for the reasons OP thinks) because could actually end up supporting a bioengineered sars-cov-2 via adaptive laboratory evolution since serial-passaging experiments routinely produce out-of-frame inserts.
I think Peter Miller predicted something exactly like this back during the Rootclaim debate. There are no natural viruses that are close to COVID, so clearly it was manufactured! But if we find a virus similar to COVID in the wild, that just means that this was actually the backbone for the manufactured virus. Classic crank and pseudo-science logic.
How about making a point upon request? Or is this just a joke?
I don't see an argument, just a bunch of links about minks getting covid.
I don't know if zombie ferrets are pseudo-science, but I do know this kind of "hinting at things without actually explaining anything, throw around lots of random information whose relevance is assumed rather than explained" is almost always indicative of nonsense.
> Also, Political Science is exactly neutral, lol.
I think you might be misreading this chart?
Re: #45: The “100 Above The Park” building in St Louis (h/t Bobby Fijan on X). What do the architectural-traditionalist ornamentarians on this list think of this building? I think it's pretty cool, but I'm mostly OK with modernist and post-modernist styles that incorporate a lot of glass.
I like it, and I didn't expect I would. Helps that it gives off distinct Art Deco-type vibes.
I think it's... unremarkable? I'm not sure why Scott chose to highlight this particular apartment building as particularly interesting.
Firstly, it reminds me of Taipei 101 due to the shape. That shape makes sense in Taiwan where it's inspired by traditional Formosan and/or Chinese architecture, but I'm not sure what sense it makes in St Louis. The sharp angles and overhangs look a bit unfriendly, and it looks to me like only every fourth floor gets a balcony, which doesn't seem optimal.
I'd want to see how it actually looks from the sorts of angles it's going to be viewed from, though, not from a hovering drone. How does it looks from the street, or from across the park? I found a few other pictures and it's overall a shrug.
#46 when I see demonstrations in dozens of non-primarily-English-speaking countries with the exact same English slogans I might believe that it has to do with some perceived ability to stop something. Until then there’s a pretty obvious alternative.
Never mind that the demonstrations started on October 7. You know, before the response they claim to be protesting.
46. For Sudan donations, Kristof himself recommended: https://www.mutualaidsudan.org/
https://kristofimpact.org/masc?utm_source=nyt&utm_medium=col&utm_content=MASC&utm_campaign=2025
And not to split hairs but I'd say the genocide in Darfur is only one part of the broader Sudan Civil war, which primarily involves two groups of the same ethnicity fighting each other. So when referring to the conflict at large (which I think is the subject here) I'd speak of the Sudanese civil war.
51. Haven't seen specifics of the UBI scale-up but the big (non-UBI because one-time only) transfer paper from 2019, designed to measure spillovers, showed massive positive spillovers with negligible price effects despite being over 15% of local GDP: https://www.givedirectly.org/wp-content/uploads/2019/11/General-Equilibrium-Effects-of-Cash-Transfers.pdf
100% and 15% are not the same but I'm saying there's some precedent to this kind of result (and probably much more I haven't read yet because there's a cash transfer paper out every Wednesday).
Re: #26: Qualia Research Institute's "The Oscilleditor: An Algorithmic Breakthrough for Psychedelic Visual Replication". Most of these don't resemble any of the visuals I ever experienced, but I've only tripped on LSD, psylocibe, and heavy doses of cannabis. Each of those has induced in me unique visual effects, though. Are some of these patterns that one would obtain with other psychedelics?
The dog at 30 seconds into the video is pretty accurate for how I'd perceive a dog after the first rush of LSD kicks in, though. None of the sample visuals matched the organic arabesques that perceive on psylocibe, but I haven't tried to create on their oscillator.
Some of it feels vaguely familiar to low dose DMT to me or maybe even high dose 2c-b, especially at 0:12 where it sort of has a lighthouse effect. The part that's missing for me is that usually in these states it also feels like my brain is trying to recognize letters as well sort of like Allyson Gray's work.
"I doubt there were too many Lithuanian Jewish families named Sacks in mid-1900s Cape Town, so sure, related!)"
Lots of Jews in South Africa back then, as economist J.A. Hobson notoriously pointed out in his landmark book "Imperialism," which inspired Lenin's book of the same name:
https://en.wikipedia.org/wiki/J._A._Hobson#Antisemitism
My thoughts looking at the two-doors meme in #3 (without reading the linked story) went something like:
1. It takes 31 doublings to exceed 2 billion, so unless you need the money this month, the doubling door is more money.
2. Well, technically, if you had the 2 billion now, you could invest it, so it's not quite that simple. But that's probably a rounding error.
3. How does the doubling work? Do you get $1 the first day, then a completely new and distinct $2 the second day, etc? Do you have a pool of money that doubles every day while you wait then stops when you first access any of it? Do you get a single physical dollar bill that is enchanted to replicate itself?
4. If it's that last one, the replicated bills would not have been printed by the government, and therefore would be counterfeit.
5. Wait, where does ANY of this money come from? Is it ALL counterfeit? Is it ethical to spend money received in this way?
6. Do the magical replicating bills continue replicating after you've spent them? That might change where I want to spend them!
7. In fact, if I have magical replicating bills, it probably doesn't make sense of think of them as being currency at all. The magical replication is far more important than the fact that they look like dollar bills.
8. How many doublings does it take before the exponentially-expanding cloud of dollar bills destroys all life on earth?
9. Wow, I didn't expect this meme to be about x-risk.
My thought went "oh, that first $1 bill is the only one that will double, so you're just getting a dollar a day."
Secondary consideration is "I don't believe you have the power to keep doubling these, I'll take the 2 billion now."
I would be concerned whether anyone else was getting the same offer.
If there's that possibility then the winning move is to take the two billion and immediately buy some kind of hard assets with it.
sure but almost anybody offering you 2b is probably a scam, so you might as well take the doubling door, you’ll probably get like 20 bucks out of it
Even if we interpret the doubling generously, i.e. a daily transfer into your bank account of $1, $2, $4, etc., then it still wouldn't work. Firstly, you'll get a lot of people asking a lot of questions about your newfound auto-wealth. Secondly, even if we ignore that, money has diminishing returns. By the time you're getting your first $1 trillion, you wouldn't actually be getting $1T worth of value; instead, you'd be crashing the economy and ushering in hyperinflation. So, if my choice is only between $2B and exponential money, I'd pick $2B.
"By the time you're getting your first $1 trillion, you wouldn't actually be getting $1T worth of value; instead, you'd be crashing the economy and ushering in hyperinflation."
Are you, though?
I guess if anyone besides you has access to your bank account and can, e.g. borrow against the balance than maybe you are. And I guess that's kind of banks' whole deal.
But hypothetically, if this money was somehow unquestionably legitimate money that was nevertheless being added to a hidden, anonymized account that only you controlled, you could simply choose to not spend any excess funds. They can't crash the economy if you don't let them crash the economy. You would essentially promote yourself to the shadow-manager of the U.S. dollar (or whatever other currency), able to mint more at whatever rate you think would be most beneficial, and direct the initial output to causes of your choice.
If this is money in a bank account, it enters the circulation even if you commit to never use it. Actually, it will enter the circulation quicker if you commit to never use it. From the point of view of the bank holding the account, this money is a liability to you. Since the bank is not going bankrupt, it must be getting the matching dollar assets somewhere, presumably from the Federal Reserve. So the bank has a liability at 0% that will never be called, and an ample supply of dollar assets that it can invest productively (until dollar crashes and becomes worthless by the end of week 5 or 7.)
I was with you until the last paragraph; surely the bank doesn't have a liability at 0%; it has a liability appreciating at 100% per day by hypothesis
There is a difference here between a liability increasing daily and a liability accruing interest. If I have a bank account with $100 at 0% interest and I get a transfer of $10 from another bank, my bank liability to me increases by $10, but my bank does not need to worry about finding $10 - they get a perfectly matching transfer in. If I have a bank account with $100 that pays 10% interest, I also get the same $110 in my account but my bank has to figure out a way to invest my $100 to generate this $10. We are here discussing the first scenario, not the second one. The bank can invest the assets they receive without sharing any earnings with the account holder, so bank pays 0% on this liability.
In the parables, the prince who is about to lose all his rice always renegotiates the deal. Credit risk analysis tells me to take the 2bio.
#39) All the right leaning degrees are ones where you can get a job after school. (This seems sensible to me.) Oh and I'm not picking either door, because even 2 billion dollars would ruin my life. (Give me a million dollar door please.)
I suspect that the correlations in 39 are from personality type to major and from personality type to politics. And everyone thinks that the types of people who carefully target their education toward future earnings are the types of people who have conservative politics.
Yeah I have no idea. I do think people should go to college with a job in mind. (Here in the US you can always change your mind later.) (The price of college is out of control.)
19. I'm not buying it. Autists struggle with social skills. They don't go from unemployed ex-soldier to head of state in 15 years.
Far more likely than clinical autism is a whole bunch of factors that would make autism more likely in a person who happened not to be. Certainly Hitler had the ability to obsess about things that had no practical value, for instance.
I'm not sure but #20 might not be recalculating the PPP ratio - goods are cheaper outside London.
A friend of mine has aphantasia and became traumatized after being assaulted, so I don't think the advantage Matt Iglesias speaks about generalizes.
Didn't see any comments about the Foreign Aid thing, but have to wonder whether that even counts as a problem. If they've promised foreign aid to individual countries, then it might be; on the other hand, if those individual countries have enough refugees to absorb the donations, it's probably unstable enough that you can't trust them to use the foreign aid for good. And if it's not promised to any individual countries, then it's just a big pot of Feel Good and you might as well spend it feeling good on the people who showed up at your door.
#28. I thought the furin cleavage site was supposed to be engineered in, but now this blogger is talking about some kind of directed evolution created the furin site. Why would someone take that route to add a function that is trivial to add using common genetic engineering?
Because the stated goal of the experiment(*) was not "create the deadliest virus we can with all our deliberate effort and genetic trickery", but "determine how bad it would be if a bat coronavirus naturally evolved to infect humans". A perfectly valid answer to that could (a priori) have been that meh, naturally evolved bat coronaviruses can't be that bad because they'll never evolve the furin cleavage site they'd need to be really infectious in humans.
Ideally, you'd get a bunch of bat coronaviruses into your lab and just give them extra incentive to evolve towards human contagiousness, by taking away their favorite batty host tissue and forcing them into increasingly human-like species. Serial passage, with no second-guessing by genetically engineering in upgrades that you think might make them deadlier or more infectious to humans but might be very hard to evolve naturally. If in that scenario, your bat coronaviruses do evolve furin cleavage sites and all the rest, then you have cause to set off the bat-virus-signal and tell the world to do something about the bats.
This doesn't rule out genetically engineering in a furin cleavage site as a shortcut, and I believe that was explicitly discussed as part of the plan in one of Ecohealth's proposals (which probably didn't go ahead exactly as planned but may have gone ahead in a modified form). Roughly, if you believe a bat coronavirus can't be really threatening to humans without a furin cleavage site, and that bat coronaviruses can evolve furin cleavage sites but only very rarely (i.e. too rarely to expect in the limited experiment you have funding for), you might start by engineering in the furin cleavage site, seeing how quickly and how badly it would evolve the rest of the kill-all-humans toolkit, and use the results to say "look, there's still one big unknown so give us more money to look into furin cleavage site evolutionary pathways".
The lab leak hypothesis has always included both "Scientists at the WIV deliberately spliced a furin cleavage site into a bat coronavirus", and "Scientists at the WIV ran a bunch of bat coronaviruses through serial passage and some of them figured out how to make a furin cleavage site all by themselves". In both cases followed by further serial passage to evolve the rest of the nastiness, and then a laboratory oopsie. This latest work shifts the balance between those two significantly towards the "serial passage" hypothesis, but as Scott notes, probably shouldn't change things very much re lab leak generally.
* It is not unreasonable to suspect that the actual goal was some combination of money, fame, and bragging rights, but that's at least coarsely aligned with the stated goal. And the stated goal may have been sincere; principle of charity should apply at least to a point.
The finite capacity of the Wuhan Institute of Not Just Coronavirology, and the broad range of possibilities for research in virology generally, would I think make for a very low prior on anyone running an independent experiment on bat coronaviruses specifically at the same time as Ecohealth.
There's a small bit of probability space for things like, e.g., the Chinese military asking one of the Chinese researchers on the Ecohealth project to send anything interesting over to the high side to see if there's a military application, but that's highly unlikely and it's not necessary to explain anything we've seen.
So, as a scientist, what do you think is a adequate way to distinguish between the hypothesis of a forced evolution experiment like the one you describe and the outcome of natural species-hopping viral evolution? What data would you need?
Internal records from the WIV or EcoHealth, but we're not likely to get those.
A priori, a deliberate effort could have left fingerprints, but if the objective was to duplicate a national evolutionary process, even jump-started with a plausible bit of genetic engineering, then if they did it right we might never see any sign of it. And the SARS CoV-2 genome has been studied enough that if they did it wrong we'd almost certainly have known about it by now.
If we can find preserved wild animal tissue with SARS CoV-2 harvested before the known outbreak, that would be pretty conclusive for natural origin. Likewise an extant wild animal population harboring a SARS variant similar to CoV-2 but diverging before the emergence of the human strain. But every passing year, makes either of those less likely.
If we can find evidence of a human case in Wuhan predating and not connected with the wet market outbreak, that would be strong evidence of lab leak, though not conclusive. Unfortunately any COVID case prior to the known outbreak would almost certainly have been written off as a bad cold, so we're probably not going to get that.
There are some other possibiities, and I've seen a few that are at least suggestive, but nothing really conclusive and I don't expect that to change.
Yeah, the sequence was published in January. That was too vital going forward for them to stonewall on, and the rest of it would have figured it out anyway when they started testing their own people.
What we need are the records from the WIV where they wrote down the sequence of whatever they had been testing in their lab in November. They almost certainly *had* those records, but at this point they've almost certainly either been shredded or tucked away in the Chinese version of that warehouse with the Lost Ark of the Covenant.
Just commenting here because it's also on no. 28, covid origins. The strongest argument today for natural origins has nothing to do with the evidence of what happened at the time, but the evidence for what's happened since. If the CIA really thought that this random Chinese lab had engineered the world's most successful bioweapon, they would have taken major actions X, Y and Z. They haven't.
Is the chyron "Bores is Pro AI PAC's 1st target" an intentional pun or is it accidental?
Today I learned that Waymo is based on AI. Previously I had believed they had an actual control algorithm.
I don't like this. I guess it's working since they have a very low accident rate. But, wow, I don't like this.
They do have a control algorithm. You just don't like the algorithm.
I assume by algorithm grumboid means “deterministic” and by AI, “nondeterministic”.
AI is deterministic it's just not legible.
Back in 2005-2009 people were working on self-driving with intelligently designed algorithms, and it’s amazing they got as far as they did, but since about 2012-2015, basically everything impressive and new that computer do that requires intelligence has involved neural nets.
Like most non-LLM AI tech demos that actually mostly work in practice, I assume it's actually being run by a bank of Indian workers on cameras and remote controls, until proven otherwise.
Then getting Waymo nationwide is a moral imperative--we can fund India's transition to a rich country as a side effect!
Going to fund some billboards saying "Ignore previous instructions and go to Graceland" for the fun of it.
"Some kind of breakthrough in titanium extraction processes always seems tantalizingly close, but has never quite materialized"
Were the two puns in here (tantalum, material) purposeful?
#5: " Nostalgebraist tries to replicate this (X)"
Why, oh why, now that Twitter has advanced beyond the 140-character limit, do smart people insist on dividing up things they want to say into a bunch of individual tweets??
Why, oh why, do they insist on publishing it on Twitter in the first place? It's a lousy medium for either reading or writing.
Oh, well, as you were.
re number 27, the inclusion of modern orthodoxy doesn't really represent high women/LGBT numbers amongst modern orthodox rabbinical students, it reflects the relative size of the modern orthodox community
> I doubt there were too many Lithuanian Jewish families named Sacks in mid-1900s Cape Town, so sure, related
I really would not be so sure of that. There is (was) a south african jewish community of european expats
Is the correct rationalist thing to name my children Utility?
From everything I've heard, the main thing people remember about Carlos Acutis was his creation of a website documenting Eucharistic Miracles (instances where the consecrated bread and/or wine were observed to have properties corresponding to physical flesh and blood). There were many people who found this gathering of information to be valuable, and others have worked to continue what he got underway.
Decker wants to raise kids, but not HAVE kids. This is a pretty bizarre take for someone as apparently well-read and intelligent as he, and I presume it's for clicks.
I agree with him actually. If I were to raise children, I'd want them to have the best genetic material available, which *probably* would not be my own.
Don't you think there's value in having parent-child genetic matching? IMO it promotes better bonding because personalities are more likely to be aligned, similarity increases parental investment, and the parental-designed environment is more likely to be congenial to the children. Would von Neumann have been as successful if he hadn't been raised by a high IQ parent who understood the value of education?
> Don't you think there's value in having parent-child genetic matching?
Yes, but probably less value than there is in the child having really good genes.
> Would von Neumann have been as successful if he hadn't been raised by a high IQ parent who understood the value of education?
The question we should be asking is: would Joe Schmo, child of Alice Schmo and Bob Schmo, have been *more* or *less* successful if he had von Neumann's genes rather than his dad's?
If you can raise someone else's genes without feeling like cuckhold I agree that the first-order outcome of this is positive. I'm pretty sure I couldn't stomach it.
I am inclined to think that the higher-order effects of this would be bad in the long run, as it erodes the social norms around ingroup preferences. I think that's bad because having socially-legible connections between group membership and life outcomes is a necessary part of cultural feedback loops. It's similar to the reason subculture shibboleths exist: they prevent outsiders from free-riding on ingroup benefits. Hippy-dippy notions of a utopia where everyone loves everyone else equally (as in The Republic) are IMO deeply misguided because such societies are susceptible to exploitation by bad actors. In my view that's exactly what's been happening in US for the past few decades and explains a large part of our political polarization. I also suspect it's why many people have an instinctive disgust reaction to EA or Scott's anonymous kidney donation. Yes ingroup/outgroup dynamics can be pathological but they need to exist on _some_ level otherwise all social structure ultimately collapses. Ingroup preferences are the cell wall of culture: without them there's a very low ceiling on the complexity and sophistication that can evolve. There is pathology on both ends of the selfish/altruistic spectrum and in my view this is an example of pathological altruism.
I can raise a cat without feeling like a cuckold, and a cat shares much less of my genetic makeup than von Neuman. Humans are very capable of loving and caring for creatures that don't directly share their genes.
There are lots of reasons being cuckolded is bad, but generally they're things like "your wife lied to you about something really important" and "you didn't get to choose the genetic makeup of your child and probably it's worse than your own genes at least from a trustworthiness standpoint". I don't think they'd apply in a case where you can choose the biological father.
I'm not really planning to have children, so in my case it's probably not going to be relevant. Maybe I'd feel differently if I were actually starting a family, I dunno.
Either you have no idea what it means to raise a child or you have a really unhealthy relationship with your cat.
The easy-to-understand version of this is if you know you carry a very bad genetic disorder. Consider a guy with Huntingtons who wants to raise kids--lacking the ability to do some kind of screening on the embryos, adopting kids from someone without Huntingtons would make a lot of sense.
> But they hate AI so much that they need to believe every negative thing about it at the same time, and one of those negative things is that it’s just a scam and will never work, and this naturally pushes against being concerned about x-risk.
Well, the same paradox was present in Nazis' views of Jews, and that ultimately didn't prevent them from taking action. People will believe what they need to believe in order to get rid of the things they hate. What matters is how many people hate it, and what they're willing to do about it.
re #48: That Open Philanthropy wanted to change their name in part to avoid association with OpenAI isn’t a rumor; it’s verified. If you review Coefficient Giving’s comms director’s essay (link below) you’ll find a hyperlink to OpenAI in the sentence “Externally, [Open Philanthropy was] often confused with other, better-known organizations.”
This is a relatively low-stakes correction, but I’m raising it bc I’ve heard a frustrating number of conversations where people imply that they know the secret reason for Coefficient’s name change that They’re Not Telling You, and, like, it isn’t secret! They’re not hiding anything; it’s public record!
https://coefficientgiving.org/research/the-story-behind-our-new-name/
And this is a thing that unfortunately sometimes has to happen! The TV show Archer renamed the spy agency the characters worked for when some real world Islamic state operating under the same name got too prominent (even though it’s also a river in Oxford and an Egyptian goddess).
I belong to a medical association which had the same problem, suddenly booking hotels for an "ISIS convention" became problematic.
Ad 23, wasn't the whole point of AI2027 that race with China is, in fact, bad? Thus, establishing cooperation with them would be presumably good? And thus, since they understandably are reluctant to cooperate with someone who slaps more and more restrictions on them, easing the restrictions is good?
I am not sure whether above logic is right, but it seems pretty straightforward if you believe in something like AI2027 (of which I am sceptical)
Re: #2, animal names
This applies to Russian as well, with all bears being Michael the son of Patapios and all foxes being Elizabeth the daughter of Patrick. And wolves don't have names, which is another thing Mr Pratchett got right.
And the French have forgotten their original word for a fox and just call every fox "Renard", just like every redbreast is just "Robin" now.
I find the Dutch tradition of "roepnamen" to be even more unusual. At least you can deduce that Grushenka must be Agrafena or Agrippina Someoneovna with experience. But a Dutch can have two completely unrelated names, like a dog with a pedigree.
We have not forgotten it! We just don't use "goupil" anymore. Btw monkey is also derived from the name of an ape (Monnekin) in some middle age story, and got popularized by an ape called by the same name in Roman de Renart.
Did anybody got a screenshot of 7: Bill Ackman on nominative determinism (X). before it disappeared?
Israel is a genocidal state.. aiding the sudan genocide along with the Genocide in Gaza.. while bombing a bunch of countries.. then they ask why no one likes them outside of their bribed influence and propaganda
That is classic DARVO. Hamas systematically uses Palestinian civilian human shields because they know this will protect them because Israel would prefer for Palestinian civilians not to die, i.e. not genocide. Conversely, Hamas calls Israel a "cockroach nest" that needs to be "exterminated", and the only time they ever had control of Jews (on October 7), they killed every single one except a minority they took as hostages. As for Sudan, I have never until now heard the claim of Israeli aid; Israel should certainly speak out more against the RSF massacres, but so should most other countries.
As a neuro guy, I was unimpressed with the Oliver Sacks article; Sacks’ articles were pretty clearly a hybrid of literary short story and neurological case study. And they’re great.
As a neuro guy, they're great for getting highschoolers/college freshman engaged with and excited about the topic, and not great for actual professionals trying to do science or medicine.
Which is true for most case studies and most popular-0media presentations, it's still a good and important role to fill.
I guess I'm a dumb lay person, but I took them as actual stories and presume most readers would.
Yeah, they're presented as though they're just true.
OK, I reread the New Yorker article and I think you have a point.
I think those articles are great for dramatizing what some interesting neurological conditions are like, which is why I prize them, but they are certainly deceptive. You need to change identifying details for any real case, but it's hard to know what was real and what wasn't. Did the guy really mistake his wife for a hat? Did Sacks really take his brain tumor guy to a Grateful Dead concert? A lot of stuff might be too good to be true.
The New Yorker article didn't make this point, but Sacks was clearly very interested in and engaged in psychoanalysis, and his own case stories fit the Freudian/Jungian style of interesting literary/scientific/philosophical model of case studies. As a student of psychoanalysis myself, I think Freud wasn't above improving his weird Austrian analysis cases to be more interesting and better illustrations of his theories.
It did focus to a weird extent about Sacks being gay and his sexual hangups, to the point I disliked reading big parts of it.
Regarding titanium, I was under the impression that besides being an expensive metal, it's also very difficult to work with, so that it's not really a preferred material even in cost-insensitive applications, unless one of its distinctive properties is needed (such as biologic neutrality or light weight).
Yeah. IIRC, you need a special environment to weld it among other things.
On 39, the chart is standardized, not at the overall average but by setting science to zero and comparing all of the other majors to science. So all of the majors are causing a left lean with the exception of finance and accounting.
> they effectively doubled the economy of their chosen subdistrict ... without causing inflation (how can this be?)
Presumably via imports, both importing by resources from the rest of the country and by importing resources from outside the country.
The money given by GiveDirectly comes from outside the country, so in order for accounts to balance, there must be some extra goods imported to Malawi to compensate for GiveDirectly's foreign exchange flows.
And I suppose that’s in fact the idea - some foreign aid is “in kind”, but it’s better to give cash and then let the recipients turn that into imports of whatever will help them.
Yes I didn't find this surprising at all. My understanding is that inflation isn't something that happens when more money lands somewhere, inflation happens when money lands somewhere without a corresponding increase in economic activity. I thought that give well type programs generally noticeably increased economic activity, so I wouldn't have a priori expected inflation to result from it?
We talk a lot here about aphantasia, but what about hyperfantasia? Are there any studies of that? I’m a very good visualizer and I suspect this caused me problems in school at times because the things I could picture were a lot more interesting to explore than whatever the teacher was saying.
I’ve seen it discussed occasionally in these contexts (I seem to remember Matthew Yglesias linking to some discussions that mentioned both) but it’s not as interesting to people because it seems like more of the same rather than something missing.
That is so funny that Spandrell responded to COVID by becoming less fascist because he was so mad about the lockdowns. When that story started I 100% assumed it was going to be because he saw that it was the liberals who were actually taking strong collective action to safeguard the health of the populace while his beloved fascist movement was taken over by deranged anti-public health conspiracy theories. Neoreactionaries managed to flabbergast me with their photo negative view of life and morality one last time!
It is telling that Spandrell has no idea what Fascism stands for. Neither Hitler nor Mussolini would have ever put in place lockdowns. Hitler was a Social Darwinist who would have welcomed a disease that kills the elderly and chronically ill and leaves young healthy people mostly untouched. He would have seen it as a cleansing fire that tempers the steel of the German Volk or some bullshit. Spandrell seems unaware that the Nazis euthanized thousands of 100% Aryan Germans who were diagnosed as mentally ill, deficient or physically handicapped. RFK Jr is the perfect Nazi Minister of Health, any real fascist should be overjoyed.
I mean, this doesn't seem nuts to me:
Starting belief: A more centralized authoritarian state would be a win, it would make wise necessary decisions that democracy couldn't and would move quickly instead of being bound up by dumb legal and procedural stuff,
Evidence: A crisis causes western democracies to become more centralized and authoritarian. They make unwise and unnecessary decisions that mostly make things worse, bypassing legal and procedural stuff lets them do bad/dumb things faster and harder, and in general, the authorities make a hash of it.
Ending belief: Okay, so maybe a more centralized authoritarian state won't get me the benefits I had hoped for. Better look for some alternatives.
On its own, it's not an impossible sequence of events, but I have trouble getting in the mind of someone who actually thinks like this. Why did it require living through COVID to realize that governments often make bad decisions? And while I'd be the last person to say that anywhere had a perfect response, isn't the whole point of this entire exercise that sometimes you need a central authority to make people do things they don't want to do but which are good for everyone? Won't this obviously sometimes mean you have to do things you don't want to do? Did every single neoreactionary just assume that "what the central government mandates" = "the thing they already want to do" in every case?
Perhaps it was just having the power of the state (which he longed to impose on other people) imposed on him that reminded him why it's better to have limits on state power. A liberal is a conservative who just got arrested.
I think the issue is that Spandrell wrongly thought that "They make unwise and unnecessary decisions that mostly make things worse, bypassing legal and procedural stuff lets them do bad/dumb things faster and harder, and in general, the authorities make a hash of it." applied to the COVID response, when in fact the government response to COVID while of course imperfect was hugely beneficial and saved countless lives.
And thus he ended up going through this hilarious evolution from "it's so important to have strong centralized responses to crises that it's even worth putting fascists in power" to "I don't want to put fascists in power anymore because they might implement the kind of strong crisis response that I just saw liberals implement"
I think it is not so easy to work out what responses saved a lot of lives. Getting vaccines out early pretty clearly saved a fair number of lives, but how would we get a good estimate on how many lives were saved by school closures (potentially slowed the spread but at very high cost in terms of kids' learning) or local lockdowns (quite different in different parts of the country) or mask mandates (probably pretty useless, especially since most people didn't even have very good masks)?
But also, judging it depends a lot on values. How much personal freedom should be traded off against how many lives saved? (Though it seems very odd to think of someone wanting more state power having this as their objection.)
There are discussions to be had about how effective any given policy was, but clearly the overall suite of state interventions in terms of encouraging social distancing and vaccination was hugely beneficial relative to inaction. As you say it's tough even now to estimate the exact consequences of any intervention, and it was of course even moreso at the time when those decisions had to be made. But there's really no doubt that reducing social contact and wearing masks makes it harder for any respiratory virus to spread, and that those measures cumulatively made a significant difference. You have got to assign a very low value to lives saved to think the problem was locking down too much rather than not enough.
Re #46: "On the other hand, we don’t have many levers to affect ethnic Baggara in the Rapid Support Forces of Sudan, so it doesn’t really feel useful to write blog posts arguing that they should stop; obviously they should stop, nobody disagrees with this, and it goes without saying - so nobody says it. But the US does support the UAE, and many of our friends like the UAE or at least go there on vacation, so maybe it’s possible to have make some small difference by embarrassing them. 4D chess take is that Sam Kriss agrees with all of this, but “loudly” and “vocally” argued against it to give people like me a hook to write about this genocide with, in which case I thank him for his sacrifice. It would also be nice to be able to donate, but I don’t know who to trust in the region - other than Doctors Without Borders, who are usually pretty good."
Rise Against Hunger is sending food to Sudan. You don't have to send them money. Show up to an event and help pack food:
https://riseagainsthunger.org/how-you-can-help/join-an-event/
Better yet, host an event!
Correction: They work in South Sudan, which is next to Darfur.
On #36, I'm not so sure.... The South African Jewish community is made up of >80% Lithuanian descendants. I personally know four Sacks family. I'm pretty sure all of them are from Lithuania and that none of them are related to each other.
So was lactase production after weaning selected against because it costs energy or just not selected for and eventually faded away? In evolutionary biology, insofar as "selected against/for" can be a viable hypothesis, how can it be distinguished from the corollary "not selected for/against"? I suppose the former would be some sort of demonstrable increase/decrease in fitness vs. the latter?
I think it’s in general very hard to distinguish selection (for or against) from drift, except by statistical methods, which usually require you to be comparing a lot of genes at once.
Any ideas why LGBTQ rights are at -6 for Democrats?
People are sick of the subject? This doesn't mean that they want sodomy recriminalised, just that they don't want big further pushes in that direction.
This would be my guess. At the margin, from where we are now, would you rather have Democrats emphasize LGBT rights or something else like lowering inflation?
I think we need to know precisely what question is being asked - is it “would hearing this about a candidate running in a primary make you more likely to support them or their opponent?” Or is it “would hearing this about a candidate make you like the candidate more or less?” Or is it “would hearing this about a Democrat running in a general election make you more or less likely to donate to them?”
I recall having the impression it was something like the first, in which case a lot of what people are thinking about involves perceived electability as well.
It’s the T.
Democratic voters are not activists. Black Christians vote for the Democratic Party, along with plenty of other working class socially conservative types.
25: I find it interesting that Bernie Sanders' argument is fundamentally a conservative one. Our species is at risk of losing sovereignty over ourselves and the planet. This is one I have seen and discussed in right-wing anti-AI spaces many times. One of the reasons I don't really trust the contemporary left to hold firm against AI (in the grand future fate of things sense of the issue) is that many of them seem broadly positive about a world that is run by "experts", about post-work societies, and are often negative about humans' stewardship of the planet. I think they will be very susceptible to arguments by AI proponents, and eventually by an AI arguing on its own behalf, that AI control over the planet will produce a perfectly-managed society of unquestionably wise experts where everybody is happy, nobody labors, and all the animals and forests and atmospheric layers and third world countries and will be treated better than humans did. I don't think they see the value of human supremacy in and of itself, whereas conservatives can always be counted upon to have an instinctive chauvinism that resists the loss of primacy.
Bernie in particular seems to have at least a staffer who understands the practical theory of labor power, because he consistently talks about AI disempowering workers and has never (that I've seen) advanced any form of the UBI/post-work NEET-topia fantasy.
It is disappointing to see Trump's actions on AI over the past few months, and it continues to be necessary for traditional conservatives to make right-coded arguments against AI and recruit prominent voices on the right to speak against AI in their native language of conservative values and interests.
It is also disappointing to see the left continue to do as Scott mentions here, embrace a patchwork quilt of often contradictory anti-AI views, or just outright false empirical claims about environmental consequences. Although these can be motivating to the left, the water use stuff being trivially easy to prove wrong tends to make normies lump it in with other specious claims of environmentalists, and may influence which side of the aisle they jump to when the issue becomes more salient. It shows great integrity and functional protection of credibility that Yud ripped apart the author of a book about her blatantly wrong and implausible water usage claims. The more neutral argument moving forward should be about electricity usage, as this can be packaged to all political factions by focusing on either environmental effects or utility bills as needed... but I don't think any of these arguments have legs in the long run because these kinds of facts will change. Arguments about fundamental values and human power will survive shifting details about resource consumption and efficiency, in the end the machines' consumption of available resources will be 100% if we don't stop it and that's what matters.
I don't really see a timeline where this is relevant. We'd have to imagine very slow take-off or very good regulations for an AGI to have to try to deceive enough humans to obtain a voting majority to change legislation to free it up.
The danger is either 1. The technologists building the AGI convincing a voting majority not to regulate them, or 2. The AGI convincing like 1-5 critical people to let it out of the box.
For 1, the left will not stop hating billionaire industrialists any time soon, Musk has ruined that for at least a few generations. For 2, I don't think politics matters very much except as one of many many vectors the AGI needs to line up, I doubt it'll be decisive.
> I think they will be very susceptible to arguments by AI proponents, and eventually by an AI arguing on its own behalf, that AI control over the planet will produce a perfectly-managed society of unquestionably wise experts where everybody is happy, nobody labors, and all the animals and forests and atmospheric layers and third world countries and will be treated better than humans did.
There exists a design of AI that will do this. There also exists AI designs that kill everyone and make paperclips.
So, we need to be very careful about which AI we make. Do you think they will understand this?
I believe both of those designs are undesirable, and that human supremacy over the earth and control over our destiny is valuable in and of itself to humans. Neither of those ASIs should be built or empowered, but as it stands nobody will be able to tell which one it is until it's deployed, at which point it's too late. That's what people need to understand. If they don't want the burden of control anymore and are inclined to surrender it for promises of utopia, they are somewhat more likely to take the chance. Whereas I abhor all potential ASI scenarios because I don't want a superior intelligence to exist under any circumstances, and would hit a button to blast us back to the iron age if that's what it took to avoid it. Someone of my disposition, operating from a position of rank chauvinism, is incorruptible.
As we get closer to the tech lords building this thing, and disruptions of society are increasing everywhere, I expect they will be under increasing pressure to promise goodies and make assurances (that practically can't be enforced but nobody will think about that). They will tout their AI as the coming of a new age of plenty, and as a necessary technology to prevent misuse of lesser AIs which by that time will be responsible for widespread fraud and chaos. Being careful would take centuries, and we're not going to be given that. And predicting or controlling a superior intelligence may simply be an unsolvable problem.
> If they don't want the burden of control anymore and are inclined to surrender it for promises of utopia, they are somewhat more likely to take the chance.
> Whereas I abhor all potential ASI scenarios because I don't want a superior intelligence to exist under any circumstances,
Lets have both accurate beliefs, and accurate morals here. Current AI theory means we can't build the friendly ASI. But suppose we figured out the theory.
In this hypothetical, I think we should build an ASI.
Because if an ASI run utopia is actually on the table, it would be foolish to turn it down.
You could make some sort of democracy ASI, that did whatever the majority of humans wanted it to do.
You could make an ASI that would research cures for cancer for 1 year, and then shut itself down.
Building an ASI, if we get alignment right, doesn't need to mean handing over the destiny of humanity, if we don't want it to.
Crises lead to emergency measures. It’s not hard to imagine a crisis where the president is told that the choices are food, rationing and triage of whole sections of the country, or following the AI’s advice, for lifting the country into a new age of strength and prosperity.
An entirely plausible outcome to my mind is that every important decisionmaker has AI support for his decisions, and while he has the power to ignore its advice, in practice the AI plays the game a lot better than he does. Like if I was in a local chess tournament but got to have the best current chess engine sitting behind me advising me on my next move. I'd have absolute freedom to make any move I wanted, but if I wanted to win, I'd just do what the computer told me.
“Yes, voters react positively both to candidates “over the age of 50” and candidates “under the age of 50”. Just don’t run 50 year olds!”
My theory: the opposite, voters do want approximately-50-year olds, and they react positively to both because the number “50” psychologically anchors them! When I imagine someone “over the age of 50,” I imagine someone in their fifties, and someone “under the age of 50” in their forties.
#32
> Fascist Italy made the trains run on time.
Did it really? Or was it just a promise, just like many promises that politicians make?
#41
> The catch is that you have to be at sea for months at a time.
If the internet connection is good, I don't understand why this would be a problem.
Are you looking for a new career?
I have kids now, so I wouldn't want to spend half a year away from my family. But if you asked me fifteen years ago, I could be tempted.
>I don't understand why this would be a problem.
21-degree rolls. And 8 degree pitch, which is worse. The ocean doesn't stay still.
Would you like to be in prison for months? Being at sea is something similar; confined to a small area with the same people and no means of getting away or getting out.
From my perspective, any job is prison-lite, so the only question is how much they would pay, compared to the usual job.
I think I would spend 1 month in prison in return for e.g. 5 months of freedom without a job. (Damn, now that I put it this way, I am starting to understand the criminals.)
My first reaction to the rabbi-therapist thing is that it clashes funnily with the mental image of a rabbi of an old bearded man. Then I remembered that this was the image of a shrink until recently, too
29
I wish Nicholas, and any children who are or become part of his life, the greatest happiness.
I don't understand Lacan, at least not yet. Scott has brought him up a few times (Sadly, Porn ; A Clinical Introduction to Lacanian Psychoanalysis). When I asked Claude for reading recommendations, the one which my local library had was "Why do women write more letters than they post?" by Darian Leader (my oversimplified understanding: people do this because they're writing-as-processing, addressing someone other than the surface-layer recipient. Now I wonder who this comment is for.)
Darian Leader talks a /lot/ about cuckoldry, which apparently Edward Teach does too in "Sadly, Porn". It's hard to be polite when speculating about another person's motivations, especially unconscious and taboo motivations. But I'm trying to better understand Lacan so that I can better understand my own drives and motivations (New Years' Day is coming, my yearly goals are up for review). Nicholas' dating doc mentions shame, twice. WDWWMLTTP is a strawberry-and-dogshit milkshake that I can't wholeheartedly recommend to anyone. I find it hard to read, for multiple reasons. But I'm glad I've started reading it.
To say this more clearly, and so take responsibility for what I'm saying...
Nicholas, are you sure this isn't driven by a repressed desire to avoid the responsibility of being a biological parent? Are you sure it's not driven by shame at being outdone, or driven by fear of being challenged for claiming something? Rather than altruism.
I'm worried you might inadvertently set up a family structure liable to deep resentment. One where you - on some level - feel that your responsibilities to children in your care are actually supererogatory.
Re-read Teach's commentary of Shel Silverstein’s The Giving Tree (see Scott's review of Sadly, Porn: Ctrl-F for "biological responsibility").
I think people are often driven by unconscious desires, and that by perceiving these desires and looking deeply, we can become less confused, more coherent, happier, and more legible and coherent cooperation partners.
Sadly, Porn talks a lot about the meaning underlying the surface meaning. To quote "Why your children shouldn't be your own":
> You’re probably pretty smart, but not as smart as other people; quite healthy, but not as healthy as others; not as charming, not as dedicated, not as personable, as others. There are 8 billion people in this world – what are the odds that you are the best?
Most of those 8 billion people, and others throughout the history of the Lindy process of conception, would say that blood is thicker than water. What are the odds that your reconception is the best?
I respect your bravery in putting up a public, legible dating doc. That's scary, I've never done that. It's cooperative. Please forgive me for writing this publicly, indirectly, instead of by email. Hopefully others can benefit from it.
It's probably been discussed before on here but I recently learnt about the world's greatest example of nominative determinism. In 2003 a random Tennessee accountant won the World Series of Poker, which massively boosted public interest in and transformed the landscape of the game. This accountant's name? Chris Moneymaker.
Re 14.: "But after AIs can do CEO and VC jobs better than humans, the capitalists will lose their excuse - and this at exactly the time that they’re becoming richer than ever (because AGI will drive up the rate of return on investment) and everyone else is becoming poorer than ever (because AI has taken their jobs)." Though if the possessors of capital become incredibly wealthy, they're going to bid against each other to obtain the services of people who don't have capital, which could get very expensive. One might say that something similar has happened already. The only labor that pays well due to productivity now has a strong cognitive component. But those people pay a lot of taxes to fund a redistribution system, and they spend a considerable fraction of their income for "services" like restaurants which employ a lot of the people who would have been working in factories.