I've lost tremendous respect for this community, which was already waning, by seeing so many people try to earnestly defend "Trump + Greenland". Why? Because.
And yes, this deserved a top-level post. Why? Because Because.
[oh, btw, the reason for Greenland, Venezuela, Minneapolis, and probably an invasion of the South Pole in 2 weeks is bc Epstein. Universally. Objectively. Rationally. There, now you know]
I agree with your feeling. Since most of my interaction is online, it isn't very clear how many of these are sincere people as opposed to (a) fictitious or (b) trolling.
For sincere supporters, I have seen four types:
(1) fans: they don't believe POTUS matters to day-to-day life and are entertained by Trump
(2) single/narrow issue: Trump truly gives them what they want wrt one issue and they don't care about the rest. I would argue anti-abortion, reduced income/corporate tax, and reduced business regulation are the main issues of this type.
(3) Democrats are worse: a CW version of (2), where the desired outcome is significantly less clear and/or Trump isn't demonstrably delivering. Key issues here are anti-woke, immigration, government corruption, misuse of government agencies, foreign policy. I understand all of my implied claims here are debatable.
(4) Stephen Millerites: these are people who simply think this is the way a POTUS should behave/government should function.
Because of the norms around here, whatever the real reason, most support has to get cast along the lines of (2): "issue X is really important" and "Trump is doing a good thing on X"
My mind is probably overwhelmed at this point by all the stuff I'm hearing on the media, but... regarding Trump and Greenland, is there any way at all that it would be remotely feasible to have Greenland like Puerto Rico?
That is, a US territory where the people are US citizens but it is self-governing to an extent? I see there are other American territories:
Maybe something along the lines of American Samoa, which seems roughly equivalent to Australia. The head of state for Australia is the British monarch (technically the queen or king of Australia as distinct from king or queen of the UK) but it's self-governing in practice. American Samoa seems very roughly along those lines:
"The Governor of American Samoa is the head of government and along with the Lieutenant Governor of American Samoa is elected on the same ticket by popular vote for a four-year term. The governor's office is located in Utulei. Since American Samoa is a U.S. territory, the President of the United States serves as the head of state but does not play a direct role in government. The Secretary of the Interior oversees the government, retaining the power to approve constitutional amendments, overrides the governor's vetoes, and nomination of justices."
Am I crazy, or would it be workable - if the Greenlanders agree - to make Greenland an American territory?
I'm pretty sure no one in the White House wants to make it a state - 2 dem Senators, a dem Representative, and 2 dem electoral votes every 4 years. But I'm also pretty sure the people of Greenland have no interest in any arrangement that makes the US citizens.
There are different arrangements that could theoretically be made (e.g. Compact of Free Association). The existence of possible legal arrangements is not the blocker here.
The blocker here is that Greenlanders don't want it, the Danish don't want it, and going to war *with NATO* over a bunch of ice where the US *already has a military base anyway* is a monumentally stupid idea.
So it is theoretically possible, if the Greenlanders and the Danes were in agreement? Okay, less insane than "Trump is going to invade Greenland" reporting, but my mind is still boggled.
Then again, why does Denmark want Greenland, either? The impression I got from reading the "Girl with the Dragon Tattoo" novels was that native Greenlanders were something like the Indigenous for Canada or Native Americans for the USA - regarded as original inhabitants of the territory now administered by another culture with no real authority or control of their own over their affairs, and somewhat of a burden/second-class citizens as far as the majority society was concerned.
I've got an antique rifle displayed over my fireplace, that I have no real use for. I'll occasionally take it out to the range, if I feel like dealing with all the hassles of black powder. Mostly, it's just sentimental value.
It's not for sale. My grandfather made it, my father owned it, now I have it, and in due course I'll pass it on to the next generation. If there's theoretically a price at which I would sell, it's one that nobody is actually going to offer and if anyone says they're offering that much I'll assume they're planning to cheat me.
If someone says "I'm taking that rifle, we can maybe cut a deal but if I don't like your deal then I'm not ruling out breaking into your house, because one way or another it's mine", then we have a problem. The kind that has me checking that my other, more modern firearms are ready to deal with people trying to break into my house.
Yes, unfortunately your modern firearms aren't much use against bombs and drones. We live in strange times where it's easier and safer to take land than posessions, seeing as land is more resistant to explosions.
Like what? Denmark isn't going to war with the US, for goodness's sake. If they have the opportunity to be fairly compensated for the land in order to avoid a conflict, they will likely take it. As for the Greenlanders themselves... the country is barely populated and can barely sustain itself without outside support. What they want is, frankly, irrelevant.
A lot of it is theoretically possible. I don't think people are hung up about that; theoretically, the united states of America could become part of Ireland.
>why does Denmark want Greenland
Well, it's part of the Danish nation (this alone should be enough of an argument, so I'll underscore it). It is also recognized as such by international treaties (and specific ones with the US- https://en.wikipedia.org/wiki/Treaty_of_the_Danish_West_Indies) the Danes want to keep it, the Greenlanders don't want to leave Denmark.
Suppose in ten years – through means that are of no relevance – Greenland becomes part of the territory of the United States. Would YOU consider that alone enough of an argument for why Denmark shouldn't retake it?
Okay, because it just doesn’t matter let’s assume someone waved a magic wand and Greenland became a US territory recognized by other nations and by treaty the way it is part of Greenland now.
Then Denmark would not even consider trying retake it because their leader is not president not-so-sane.
This is goofier than your hostis humani generis argument.
> Denmark would not even consider trying retake it
Right, because they are too weak to. When they were stronger, they conquered vast tracts across the North Sea region. They're just on the other side now.
You have a small-minded understanding of the law of nations. Just like the British Empire before it could unilaterally redefine terms understood since antiquity to suit its purposes – which was my argument earlier you refer to – America can today remake the world and have it "legitimized" later if it bothers to, but in two hundred years, I expect there will continue to be people who don't understand how this works, and sneer at those who do as "goofy."
I have some difficulty parsing your question; is it something say that the status quo plays some factor that should be respected, but this status quo can change and then the new status quo should play a role? So if the US "has" greenland (legally? militarily? 51st state? militarily occupied? Greenlandish independence and vote with majority to join?) 10 years from now, than the Danish should accept that, and not try to retake it because "it's now part of the US"?
I think the means are relevent, if that is your question.
I'm saying your argument that Greenland currently being Danish territory should be enough of a reason not to change that proves too much.
I don't think you actually believe that yourself, since you're arguing the means by which the territory was acquired are relevant. Now, if you're actually just saying that the gains aren't worth the price the US would have to pay to conquer and hold the territory, both against Denmark and its allies' attempts at reconquest and against rebellions, fine, that's a sane argument.
I cannot believe this is so! I understand that in the North American Treaty Organization Charter, it says one should not ban political parties, or otherwise infringe on free political speech. This doesn't sound like the sort of document that West Germany would have signed, had they the different social contract you reference. Nor France, for that matter.
As someone who prefers the American approach, I find I can still steelman the German one, and even point out echoes of it in current American politics. The Germans were presented with evidence of one set of ideas that led to a great deal of ruin for them, and that evidence was hard to refute. An extremely simplified version would be the idea that it should be okay to kill someone under your care, on the premise that person is a threat to your existence, while pointing your gun at anyone who seeks to intervene. German authorities post-WWII concluded that that was an idea that should never be taken seriously, because doing so turned out to be an even clearer and present threat to them.
Therefore, anyone who tried could be stopped without objection, even by means other than debate. For them, the debate was closed, and stayed that way for generations. Maybe debate was insufficient; maybe individuals were still too close to the details that made such ideas tolerable; maybe it was a problem of informing everyone of the case against the idea; maybe a combination. Whatever the reasoning, that was the determination.
Nowadays, I'm seeing another reason to want to suppress such ideas, having to do with individual attention span. Current technology has given broadcast power to any individual, and yet we all have only so much time in the day to absorb news; if that time is coopted by someone spreading dangerous or even distracting ideas through broadcast channels, and we don't have the ability to effectively filter out such ideas, then our model of the world is inaccurate and our time is wasted.
Ideally (for the pro-American approach), individuals ought to possess such an ability. In practice, it's proven elusive. Popular networks like YouTube or Reddit or X can attempt to filter on keywords, but some people have discovered that spreading such ideas in bad faith is fun in its own right, much like prodding an anthill and watching the ants scurry, so they have a strong incentive to evade filtering. There's also the problem where, if the filtering is _too_ accurate, then individuals can filter sources run by powerful people who feel they are entitled to being heard, so they probably have an incentive to drag their feet on improving the filtering. The exception is when the filtering blocks ideas reminiscent of the early 20th-century horrors.
The result is that that filtering sticks around, is maintained by a relatively small group of people, enforced on the public, and the public tolerates it as long as they believe those ideas are the old horrors.
>Nowadays, I'm seeing another reason to want to suppress such ideas, having to do with individual attention span...
At that point, you may as well just stop being democratic. The nazi ban is at least in principle limited, but if The Peoples decision making needs to be managed in a content-sensitive way to be workable at all, then why listen to them at all? Have whichever body would do that managing make the descisions and be done with it.
You make a good point, and I agree. A lie can travel halfway around the world while the truth is putting on its shoes. A democracy must be able to defend itself against those that wish to destroy it, or else it won't exist for long. The Weimar Republic learned this, too late, the hard way.
I wonder how much the Weimer Republic should actually be counted as a democracy; it was largely established by the actions of foreign nations, and collapsed in a single generation. How much was it a democracy as opposed to Vichy Germany?
Congratulations, your model of the world is inaccurate and your time is wasted. Actual reality isn't generally discussed in polite company. See the Forbes list of "most wealthy men" for details about what is "polite company" and what is just "new rich."
They're also deeply committed to pretending to have free speech after all, and almost the entirety of the American ruling class pretend to believe them – Vance's speech at the Munich Security Conference might the closest anyone has come since the end of the Second World War to pointing out the blatant lie, but even he walked it back shortly after. The US allies with foreign nations with values antithetical to its own quite often, and it's perfectly reasonable to argue that it's already fine allying with Saudi Arabia and the other Gulf monarchies, and most of Europe isn't much worse than them, but instead, they put on this elaborate kabuki theater of "shared values" and "democracy" and lecture other countries for their lack thereof, which is deeply irksome.
> […] Saudi Arabia and the other Gulf monarchies, and most of Europe isn't much worse than them […]
I'm curious: Do you genuinely believe that "most of Europe" isn't better than Saudi Arabia in regards to free speech? Or is this just a deliberately provocative comment to aggravate the discussion?
I occasionally have cause to transit through airports in Europe, so I don't think it prudent to say.
What you're quoting was me trying to forestall any suggestion that I was implying the US ought to repudiate its alliances with countries in Europe out of moral outrage: if there is any such a point where a partner's practices become so repugnant that reasons of, say, troop morale or civilian unrest, compel one to so break off relations, Europe is certainly NOT there.
> I occasionally have cause to transit through airports in Europe, so I don't think it prudent to say.
That's deeply ironic, given that no European countries ask for your social media accounts when entering or transiting, whereas the US requires their disclosure for visa applications, and plans to make disclosure mandatory even for visa-free travel.
Russia had freedom of speech, in that "He was drunk!" was an acceptable excuse for saying... practically anything. Freedom of action was a different matter, of course. I imagine it was a little tighter in countries where liquor wasn't the desired method of pacification of the proles.
If the people cannot discuss their own political system and political wishes freely, then the people are not in control. You can't hold the following three propositions all at the same time:
1. The people have the capability and the right to control the government.
2. If the people are allowed to speak and listen freely, then they will choose Nazism.
3. Nazism is bad.
I suppose the German political system works reasonably well, but it is not free and it is not a democracy.
Well, the first sentence is just not a characteristic of a democratic republic, so I don't think the contradiction you believe exists is there.
The people have the right to elect a government, and there are extensive rules constraining said government from doing what it wants to its citizens. But that is not the same as "controling" the government.
Under normal democratic theory, sovereignty rests in the people as a whole, and the government is just the individuals deputised to run the country on their behalf. I think it's reasonable to describe this as the people "controlling" the government, even if, for practical reasons, this control is exercised indirectly in most modern democracies.
The use of the word "control" is a question of definition, so let's look at it in the context of what you wrote.
You said that people have the "capability and the right to control the government". But that is just factually untrue without major qualifiers you omitted.
We also say that shareholders are in control of a public company. This is a practically useful sentence to say in most contexts.
But is it contradictory to also say that shareholders cannot direct a company to commit fraud or the CEO to kill himself?
Not really, since we all understand that shareholder control is not control without limits. In the same way, the people's "control" of the government is limited. And the fact that it is is not irreconcibale with what I understand as "free and deomcratic".
>You said that people have the "capability and the right to control the government". But that is just factually untrue without major qualifiers you omitted.
If you look more carefully, you'll see that that wasn't actually written by me, but by someone else who happens to have the same avatar.
> I suppose the German political system works reasonably well, but it is not free and it is not a democracy.
"Democracy" does not mean "unrestricted dictatorship of the majority". There has never been a pure democracy, i.e., a state where the population has unrestricted decision power through voting. In particular, all modern democracies (including the US) have checks and balances – typically in the form of legislative and judicial institutions and processes – to protect certain rights and constitutional structures.
With this in mind, there's no reason to single out Germany. If you take "free" and "democracy" in their absolute meaning, there is not a single country on Earth that's "free and a democracy".
Merz filing 5000 or so criminal complaints (some of which led to SWAT teams) about people shitposting on the internet seems to indicate that Germany is, in fact, rather singular about its approach to free speech, and authority figures.
The following is a german text from me translated by Claude, I don't know if the technical terms are always the right ones. Feedback on this take would be welcome. :)
USA: Maximalism with minimal exceptions
The First Amendment protects speech extremely broadly. The Schelling point is "the state almost never intervenes." The few exceptions are narrowly defined: imminent calls to violence ("imminent lawless action"), true threats, defamation with demonstrable harm, obscenity (very narrowly defined).
Hate speech, Holocaust denial, Nazi symbols are all fundamentally legal. The idea: once you start banning "bad ideas," there's no defensible boundary. Better to keep the marketplace of ideas open even for shitty ideas.
Germany: Value-based democracy with active boundaries
Article 5 of the Basic Law guarantees freedom of expression, but with explicit limitations in "general laws." The Schelling point is "democracy must be able to defend itself against its enemies" (militant democracy).
Prohibited: incitement to hatred, Holocaust denial, Nazi symbols, insult is criminal, defamation even without the American "actual malice" standard. The historical lesson: Weimar let itself be talked out of existence.
Both have internally coherent logics that emerged from different historical situations/traumas. The USA fears state power more than dangerous ideas, Germany fears dangerous ideas more than state power (at least in this domain).
Re. Nick Shirley fraud allegations: I just kinda assumed "This sounds like a thing that is probably true, I will believe it because the conservatives in my life do because they are not hysterical as a rule and it's not in my state and I can't do anything about it."
Just recently came across an interview with the dude on my "Listenin' while I dig Holes" feed, and holy shit: This kid is dumb as rocks. He is legitimately subnormal level just not that smart; and he said some things that demonstrated a rock solid commitment to post-truthism ("I don't like that you didn't condemn Kirk's assassination" "I did, to the point where people were giving me shit for being to nice to Kirk" "I'm not going to watch it and I refuse to update", infuriating).
It made me go quickly check out his video, and it is straight up stupid. Anyone that has ever worked in, around, with, or patronized a certified childcare facility should have been able to know that +/- 85% of everything he showed in the video is not only normal but legally required, and the remaining 15% was not sufficient to justify the claims (eg, could have been explained by two employees calling out sick at the same time/a pipe breaking).
If fraud is discovered form this (which it probably will be, most for profit childcare/elder care/religious facilities that can get state money are skimming off the top), it would be like someone saying that the lottery numbers they won on were revealed to them by the ghost of PT Barnum.
Incredible how easily your information environment can be poisoned, really puts you in a constant epistemic crisis.
After seeing the Democrats' housing policy in New York, I think literally the ONLY thing the Republicans can do to be the bad guys is adopting similar policies. Knowing the Stupid Party, I'd bet that's what they're going to do by the 2028 election, if not even sooner.
Exactly! You've disqualified yourself from discussing New York Real Estate, which is one of the world's Reserve Currencies. Education first, talk later. (NYC, like London, has unique game-rules for capital investment, that are somewhat governed by the use of its real estate as a world reserve currency).
It's not an idiosyncratic theory. NYC and London are ranked Alpha++ because of the global elite's storage of wealth in their housing market, among other things.
It's just a vastly different marketplace when the world's elite use it as a place to store wealth. The stability or lack thereof of the real estate market is different than one would see in say, Chicago, San Francisco or Los Angeles.
Substack blogger Ryan Broderick recently spent 3 days in Minneapolis observing ICE and related matters. His observations about ICE:
>ICE agents are, simply put, fucking clowns. According to The Atlantic, they receive 47 days of training — in honor of Trump, the 47th president, naturally. Many of them, also, can barely read or write, apparently. The ones I spent the weekend following around didn’t even have proper uniforms, with some wearing sneakers. In Minnesota. In January. These dipshits are also wearing camo in the snow. They clearly do not have any training when it comes to their own weapons either. Multiple times over the last few days, I watched officers fire pepper spray balls at the feet of protestors barely a few inches away from them. These weapons are basically paintball guns full of concentrated pepper spray. So when they hit a target, they explode into the air. Which meant ICE agents regularly ended up poisoning themselves with their own weapons. I also watched two agents ask each other if a canister they were about to fire at the crowd was tear gas or a stun grenade. (It ended up being a stun grenade that then ignited the tear gas they had already shot at us, which started a fire in the street that a protestor had to help them put out.)
>Brutal state violence and hysterical right-wing internet content work together in lockstep. According to The Washington Post, the agency is under pressure from The White House to create as much content as possible. Which is why ICE agents have a phone in one hand and a gun in the other. But it goes beyond that.
During a showdown with protestors at the Whipple Federal Building in Minneapolis, I watched as one ICE officer fist-bumped a pro-Trump content creator once he learned he was there to support them. I also watched as a gang of groyper livestreamers, led by January 6th insurrectionist Jake Lang, rile up a crowd of protestors, creating the perfect pretext for ICE agents to fire pepper spray balls and tear gas at the crowd.
Also, re: "phone in one hand, gun in the other", that's used as a "damned if you do, damned if you don't".
Interactions are not filmed: this is proof they're illegal! if nothing bad went on, why aren't they providing video so we can see for ourselves? what are they trying to hide?
Interactions are filmed and posted online: the monsters are bragging about murdering innocent people! they even film it and then make it public!
>News outlets and DHS officials reported cuts to the length of ICE training during Trump’s second term, reducing it from about five months to six days a week for eight weeks. That is 48 days of training over a 56-day period. (What it has to do with Trump’s status as 47th president is outside the scope of this fact-check.)
hopefully this helps in updating your beliefs about ICE training.
I don't think that's productive or smart conclusion to draw from the article; in fact, I would say you should update, based on the sources (and the claim from the Atlantic) that it is more likely than before that they do 47 days of training. Note that the sources mentioned did not want to give a specific number of days. I wonder why?
Please read this article; see it as an exercise, if you have a tribal response and don't want to be exposed to information you might not like, regarding the ICE screening procedure. You can of course always say she is lying, or disregard it, or point to DEI, or other people being partisan, if you feel like it.
A quick google says that police academy in the US is 21 weeks on average (plus additional field training afterwards), so they are very under-trained relative to an average police officer.
And while you might argue that ICE is a specialized job so they should need less training than regular law enforcement, ICE are being used to kick down doors, run vehicles off the road, deploy tear gas and flashbangs against protesters, and other high-pressure situations where they're likely to get someone killed if they screw it up, or violate their constitutional rights. They can't afford to cut corners here.
(And training is definitely part of the issue - I've seen several articles pointing out that police are trained not to stand in front of a suspect's car during a traffic stop, because that can go wrong in exactly the way it did with Renee Good.)
"ICE are being used to kick down doors, run vehicles off the road, deploy tear gas and flashbangs against protesters"
Is this because local police forces, who should be the ones handling crowd control are refusing/being instructed not to do such work in support of ICE, so they end up having to do the police work themselves?
To my understanding, even if the police were fully willing and permitted to support ICE raids, they wouldn't be expected to take the lead on anything - ICE agents would still be the ones kicking in doors, demanding to see people's papers, and dragging people into vans. So like, it might reduce the amount of tear gas they throw at protesters, but ICE would still be doing exciting and violent things that can get people shot.
(Also, if your policy is "it's okay to provoke massive protests because crowd control is someone else's problem," I imagine that would make you very unpopular with the local police very quickly.)
Probably yes in some sense, no in other senses, and it doesn’t matter in that they need more training no matter whose fault it is that they have the job they have. ICE have been acting in those roles for long enough they could have trained up a bunch of recruits properly since their jobs started featuring those skillsets.
That is actually a great question, because it immediately put me in touch with the fact that I am indignant about the number without having any standard for judging it’s length. My instant feeling is that it’s too little training, but I can imagine getting into being indignant about its being too much. In fact, there’s a lot in the Broderick quote that fires up my indignation at ICE officers but really doesn’t justify that reaction, just sounds like it does. I’m prtty sure itsnot true that many can barely read or write, because a while ago I asked GPT for stats on the groups demographics, and most are from middle class families & many have college degrees. As for many wearing sneakers, seems like they most not have been given appropriate shoes. Sneakers in winter MN are really dangerous — you’d slip very easily on ice, and be in danger of frostbite. As for the camo, that seems fine. After all, they’re not trying to be hard to spot.
As for training time — Well, seems to me most people could learn in half an hour max not to fire pepper spray at close range because of the splashback, and how to tell the difference between a tear gas canister and a stun grenade, so I’m not sure whattup with that. But a lot of the rest of what the ICE guys are facing seems much more difficult that most police work, and police have about 3x as much initial training as ICE guys, and then 6 months or so of being closely supervised before being turned loose to make judgment calls about whatever comes at them. But my guess at present is that however much training these guys had had, what they would need in a Minnesota kind of situation is strong leadership, and someone running on-the-spot training and planning for how to handle various situations that are likely to come up the next day. And some policies, like that guys should do things in pairs.
Camo is probably not proper uniforms, but then again, if there were a specific uniform for ICE, we'd be deafened by the screaming about the SS and look see they're getting their own Gestapo uniforms.
My local cops screwed up the tear gas versus other grenades thing. To the point where they were contradicting the protestors (who had the tear gas in their faces). To their credit, the police did own up to it the next day (someone read the cannister on the ground, and said, "hey, boss...").
These are not the most intelligent people in the universe. We don't hire them for that -- we hire them for "can run for ten minutes without collapsing" (which is a difficult bar to pass these days).
Please compare to basic training in the military, which is 13 weeks at most, and 8 weeks at minimum. So, ICE training is a little shorter, but not all that much.
I agree with the "strong leadership" and planning being necessary (also "doubling up" seems to be a good idea, particularly with the Venezuelan shovelwielders around).
Okay, yes, I agree that comparing it to how long regular police training is makes it look quite short, but as I understand it, most of the new ICE recruits just switched from being active cops – with veterans and ex-cops being next on the priority list – so I think it's reasonable to rely on the overlap in required skills and have the ICE-specific training be shorter in duration.
Just men: 1 million veterans aged 18–34, and 3 million aged 35–55.
And I can't find stats on ex-cops, but ~800,000 cops in the US, and 1/5 of those are veterans.
I think they have enough. There's this line from the Punisher which I thought was prescient: "They spent fifteen years training an army and then abandoned it on the streets." Someone might be picking it up.
The idea of giving anyone a gun and any kind of authority over other people with less than 6 months of intensive training and a psychological evaluation is absolutely terrifying. 2 years is where I start feeling comfortable.
13 weeks for the Marines (that's boot camp), and that's the longest our military gives before we issue people guns (Granted, we do not issue them explosives without Written Justification, after several incidents using explosives to clean toilets...).
Remember when Congress wanted the National Guard to swoop in? That's 10 weeks of bootcamp (and since National Guard is often busy driving trucks and fixing flood damage, guns are not a high priority).
Please don't do this here. The rest of your comment makes a good point, and snark like this only detracts from otherwise valuable information. Nothing good will come from it, even if it feels satisfying when you type it.
Is this a passively aggressive way of saying that there is no such thing as a meaningful psychological evaluation?
Not even on the level of "try to find an obvious psycho by asking them -- in situation X, would you do an obviously psycho thing? -- and eliminating those who are stupid enough to say yes"?
Given the work they're being hired for, "psycho" answers aren't a deal breaker. Maybe keep them out of routine enforcement, but sometimes you need people to do dirty work that the regular hires just aren't psychologically capable of. Last thing you want is for them to get cold feet at a crucial moment.
Every time there’s a topic where there’s a chance to weigh in in favor of doing the harsh, violent thing, whatever it is —dumping the oldsters off Medicare, giving those arrested immigrants some blows below the belt, you pipe right up in favor. And what you post always has the sound of something more than a set of beliefs about how to make society work. There’s this whiff of personal thrill coming off you. You sound like Dolores Umbridge sounded when she was working herself up to using the Cruciatus Curse on Harry and his friends. “I am left with no alternative. This is more than a matter of school discipline, this is ministry security. Yes! Yes!” She was shifting her weight from foot to foot, beating her wand against her empty palm and breathing heavily. “You are forcing me Potter, I do not want to,” said Umbridge, still moving restlessly on the spot. “But sometimes circumstances justify the use . . . I am sure the minister will understand that I had no choice. The Cruciatis Curse ought to loosen your tongue.” There was a nasty, eager excited look on her face that Harry had never seen before.
Their training is probably sufficient for doing their actual job. It's probably lacking in the "riot squad" aspect because they were never intended to be a riot squad as well. I am sure that if random idiots started obstructing you while you were trying to do your job every day then you would struggle with that as well.
Do these idiots want there to be a well-trained Federal riot squad? Because that seems to be the future they are manifesting.
I just don't get it. If their training is not adequate to teach them how (and when) to use guns, why give them guns? If "doing their actual job" does not involve using guns, why give them guns? If "doing their actual job" does involve using guns, why is it OK not to teach them how (and when) to use guns?
Congratulations! In minneapolis, they're rolling out prison guards in SWAT gear. yes, they've had training in how to use batons and control crowds. Training in "not hurting people's feelings"? Not so much.
Ross is not going to be convicted of murder, for two reasons:
1. Criminal prosecution would be by the federal government and the people who control it are on his side. As I wrote in a different context, if the King controls criminal prosecution, the King’s friends can get away with murder. So he will not be prosecuted.
2. If he was prosecuted for murder he would not be and should not be convicted, although he might get a hung jury. Shooting the driver of a car who is accelerating towards you is arguably self-defense, arguable enough to not meet the "beyond a reasonable doubt" standard for murder.
He cannot be tried and convicted for the crime of murder. He can be tried, perhaps convicted, for the tort of wrongful death. A tort prosecution is initiated by the claimant not by the state. A tort conviction only requires proof by a preponderance of the evidence.
The plaintiff's attorney could argue that if he was afraid the car would run over him the sensible response was to get out of the way, since shooting the driver would not stop the car, evidence that whether or not he was trying to kill her he was not trying nearly as hard as he should have been not to. He could also point out that the second and third shots were fired after Ross had gotten out of the way.
He might not be convicted and if he was his employer would probably pay the damages, but the trial would generate information about what happened that should make it a little clearer whether and to what extent he was at fault.
Re 1, a subsequent democratic administration could still press charges in 2029 or 2033 though iiuc (unless the Trump doj already presses charges and loses a case, because double jeopardy?)
I can imagine a Democratic president deciding to dig this one up to distract from a new issue.
Re 1, a subsequent democratic administration could still press charges in 2029 or 2033 though iiuc (unless the Trump doj already presses charges and loses a case, because double jeopardy?)
I can imagine a Democratic president deciding to dig this one up to distract from a new issue.
Anytime someone is in front of a vehicle in operation, they have a reasonable fear of being struck. Were his actions in shooting the driver reasonable? If he is injured, it is plausible that this further proves his action was not a reasonable response as shooting did not even prevent the injury he feared.
I would also amplify your condition, "if true" and associated doubts. It is an uphill climb to convince doubters that administration sources are trustworthy on this story. Technically, a bruise on the shooting hand would count as internal bleeding, but wouldn't be very significant in how to interpret the event.
If you go in trying to disrupt law enforcement, see the cop in front of you, look him in the eye, slam on the gas pedal then injure him them, then it’s ridiculous to charge that guy with murder.
I bet the "internal bleeding" ends up being a bruise (and of course they can't demonstrate the bruise came specifically from the car, so it'd just be "he has a bruise").
How nice that you can tell what is and what is not a genuine injury from the comfort of your chair without even needing to see any kind of medical report.
You should apply for work with Capita carrying out disability assessments, you sound like the type of person they want:
Again, probably a bruise. As for the hospital, I haven't seen a source other than admin officials who also said that he almost died and Renee Good was a domestic terrorist and shit.
You won't see another source unless this goes to court. HIPAA. Agree that it's low value data, in that "workplace injuries" are a thing, and someone's gonna want to CYA -- but it probably does indicate, aligning with independent video (not his cellphone), that the guy was hit.
1. There can't be a state prosecution? Is this because federal officials are immune from state law?
2. I do not believe that is how affirmative defenses (which self-defense normally is) work. First the prosecution has to prove beyond a reasonable doubt that the defendant met the standard for homicide. Presumably that is simple enough in this case. Then the defendant may offer an affirmative defense that the homicide was justified. The defendant would have to prove by a preponderance of the evidence that the killing was justified (self-defense).
Self-defense is typically a state law that varies between jurisdictions. If this were a federal prosecution I do not know what law would apply, maybe there is federal law about self-defense in such cases, I haven't looked into this.
My understanding is that states vary as to whether self-defense is an affirmative defense against murder charges or if lack of self defense is an element of the crime. In the former case, the defense needs to show self-defense to some burden of proof. If the latter, the prosecution needs to show beyond a reasonable doubt that it wasn't self-defense. Under Common Law, I think affirmative defenses generally have to be proven to the Preponderance of Evidence standard, but a lot of states have special statutory treatment for self-defense as an affirmative defense.
I think it's technically an affirmative defense in Minnesota, but one with a "burden of production" rather than a burden of proof. The defense needs to make a claim and produce evidence that makes a prima facie case for self-defense, which the prosecution must then refute to a "beyond reasonable doubt" standard.
So, this guy, from the old culture war threads on SSC (eventually The Motte), who is right wing enough to argue the covid and flu shots should not be called vaccines at all (efficacy is too low relative to most things called vaccines), says this was manslaughter, even though the protester has some of the blame.
"There is no country on all of planet Earth, as far as my research has been able to glean, which simply allows illegal immigrants to stay indefinitely once they’ve been identifie
To summarize one highlight in that linked essay: ICE are enforcement only, they are the snatch-and-grab force. They do not have research, investigation, public outreach, judicial, or general law enforcement skills or authority. This is the wrong force being deployed in the wrong way.
ICE seems to be acting in coordination with a lot of other agencies, including the military. I wouldn't bet that someone called "ICE" is actually just ICE (I got this from reading the filing about the Minneapolis judge, where the ICE guys involved were from multiple agencies).
Also, what exactly do you think our military on "homeland defense" is doing, if not ICE? (Yes, yes, tracking down terrorists on our soil, we can call this ICE and send them HOME).
The military is not allowed to engage in domestic law enforcement. Sure. Flying drones, following money, tracking down illegal gun runners (and then asking the Secret Service to do the law enforcement)... all of this is stuff that the military does do. Law enforcement is defined very narrowly, and when the US military flies combat missions over US soil, that's also part of ICE (combat missions presumably because the Mexican cartels have anti-aircraft weaponry).
So, you're telling me that Marines shouldn't have been asked to "provide covering fire" when they shot over 200 bullets into a gangster's house? **
**This was NOT Trump. It was years ago. There's a reason we don't use military to do law enforcement (including protecting Capital Hill) -- they don't got the training, and ... misunderstandings do occur.
blorbo has already pointed out that the car was not in fact aimed at him. Even the claim that it was is really that it was on a path to slightly clip him at low speed.
These sorts of "analyze police shooting frame by frame" (usually in a situation where if it was not the police shooting a civilian then everyone would agree it was murder), arise from time to time, and we're all used to the beats of the discourse.
But in this case there really is a difference in kind. Usually the threat is a person making threatening statements, having a weapon, etc. In this case it's a car. We've all been around cars, we've all been in situations where for a moment it looks like a car is going at you. Your gut reaction is to move out of the way, not attack/shoot the driver.
Shooting Good was not actually self-defense - I don't mean legally, I mean in terms of the laws of physics. But it's more basic than that - shooting Good is not an act that a reasonable person would do, spur of the moment, to defend themselves either. The defense is usually "well he had to make a life or death decision in a split second, being a cop is hard, you're a civilian you've never been in this situation" but the fact is, I've been in this situation before, and so have you, and we all managed to not kill anyone.
She was gunning the gas, this is not a "low speed" driving accident. And I know someone with training, who was involved in a "low speed" driving accident (as in the car was navigating a parking lot, pulling out before turning onto the street) As per training, he went onto the hood of the car. You're not trained to evade the car, you're trained to roll onto it. He left a handprint in the metal hood of the car.
Everyone involved was embarrassed at the situation (driver did not see pedestrian.)
> We've all been around cars, we've all been in situations where for a moment it looks like a car is going at you. Your gut reaction is to move out of the way, not attack/shoot the driver.
Yeah no shit, I don't walk around holding a gun. If I had one, and didn't need to worry about getting charged with homicide, then the situation obviously changes. It's still self-defense because a dead person has a zero percent chance of intentionally killing you.
"it's still self-defense because a dead person has a zero percent chance of intentionally killing you"
And yet here we have video evidence that shows that the dead person had exactly the same chance of killing him as the live version, since shooting her did not stop the car at all. So you're completely wrong.
The car was not directly pointed at him, and thus he only got lightly bruised. While a dead person can still hold the pedal, a live person can turn the wheel. In hindsight, killing her posed a smaller risk to his life than not... though I doubt he was thinking that clearly at the time. Either way, self-defense is self-defense. He had no reason to kill her unless there was a non-zero chance of her harming him.
Not to mention that the driver "had to make a life or death decision in a split second" as well, but it really rarely comes up.
Perhaps the moral of the story is not to freak out civilians by trying to enter their vehicle, especially when one of your colleagues is in front of their car, and the car has its engine running.
Part of a wider pattern where it's OK for cops to act totally irrationally based on fear but civilians are supposed to handle dangerous, fast-evolving situations with total calm and poise.
If he was in fear for his life, then any civilian who has an officer pull a gun on them is too (including in this case where the cop went for his gun before she started to drive). If we're supposed to treat it differently because it's a cop and not a random person, then that comes with a responsibility for the cop to be judicious with the use of force.
I think you make a very important point that is emphasized by use of the word "King."
If there is an eventual hand-over to a group with a different impression of the events and political preferences, then it is possible prosecution could go ahead. The current path seems to be creating a large block of people in or associated with the current administration that have strong incentives to block a peaceful transition of power.
Between your previous comment and this one, you seem to have gained certainty that Good was accelerating toward the ICE agent. Have you gotten new information that caused this update?
There is something that I am really struggling to understand why reasonable people are leaving out of the discussion: In the self-defense community, there is (used to be?) a concept of stopping power. In bullet vs SUV, the bullet just doesn't have enough. On the face of it, shooting the driver doesn't make logical sense as a reasonable tactic for this situation.
The obvious advice I got as a child is to be cautious about vehicles in operation, stay out of their path/potential path (both front and back), and get out of the way if I feel the vehicle is driving toward me. Hollywood movies aside, it would never occur to me that shooting at the vehicle would help in this situation.
If you fear for your life in the position he was in, the sensible move would be to step to the side quickly, not pull out a gun one handed to kill the driver.
It was also unnecessary for Renee Good to obstruct justice and resist arrest. It's not unlike a drunk driver who dies in a one-car crash because he loses control of the vehicle and spins into a tree. Should the steering system have reacted more stably to erratic inputs? Possibly, but I'm assigning moral blame to the person who made the deliberative decision to do something dangerous. When you force armed men to make split-second decisions then sometimes you get shot. If you don't want the dice to come up snake eyes then you don't roll them.
I don't think law enforcement has a duty to react perfectly to deliberate and irresponsible provocation.
Law enforcement should always be held to a higher standard than the public. An untrained woman, surrounded by armed men shouting conflicting orders is likely to make a mistake.
An armed law enforcement officer should have enough training and experience not to pull out his gun one handed while trying to film on his phone when he should simple move to the side, note her license plate number and call it in. Paramedics and ambulance crews face more dangerous situations, unarmed multiple times a night. He executed her.
Who chose to put herself in harm's way and break the law. I agree the officer should have responded differently but IMO that woman deserves zero sympathy. She wasn't an innocent bystander, she was an intentional provocateur. In a tossup between law enforcement and law breakers I will stand with the law every time.
If this is true then you have essentially fully replaced morality with 'following the law'. Under this conception, you would side with the SS over people hiding Jews. You would have handed run away slaves back to slave owners.
It's a logic that will always allow evil to continue.
This is such an inaccurate and unsupported read of my position that I can't even pretend to assume it was in good faith. You're comparing the SS to an ICE officer who shot someone who violently resisted arrest after obstructing justice. That's ... just not worth responding to substantively.
This is a fully general argument - all people who are stopped by law enforcement are (allegedly) breaking the law. If that's your standard, then you will never have sympathy for any person who gets shot by the police.
No it's not, hence the word 'tossup' which means reasonable uncertainty as to who was at fault. Rene Good wasn't an innocent bystander, she was a willing provocateur who violently resisted arrest. If you're resisting law enforcement then you're just wrong. If you want to make an argument, do it in court. Any defiant interaction with law enforcement is prima facie evidence that the suspect is in the wrong. Now if a cop just walks up to a random person and starts beating them then that's obviously a different story, but there needs to be clear evidence that that's what happened and it's very clear that it's not what happened with Good. Citizens have a duty to comply with law enforcement. Anyone who refuses to do that gets what they get.
"As he approached Renee Good’s vehicle on a Minneapolis street on Wednesday, Immigration and Customs Enforcement agent Jonathan Ross held up his phone camera and recorded video."
"Cellphone video taken by the officer who shot and killed Renee Nicole Good shows the motorist and the woman believed to be her wife speaking to the agent moments before the fatal shooting."
Doesn't change anything about the debatable positioning of the car, etc, but some people seem to be overclocking on the erratic movement of the camera, without understanding that the camera in question is in his hand, not attached to his body.
If a car hits a camera in your hand, then at *best* the car has narrowly avoided running you over by a few feet. If you choose act with total disregard for the life and safety of an officer in that way, it should surprise no one when they respond with deadly force. Please exercise some common sense here. There is video of the wheels spinning forward while pointed directly at the officer in front of the car. Trying to justify her actions is only going to encourage more people to create dangerous situations where someone could be injured or killed.
At the moment the other officers started confronting the driver, the single most likely time for the driver to panic and do something stupid, the officer walked from the side of the car, where he was well out of danger, to directly in front of the car, . The officer is the one disregarding his safety.
If you reread my post, I expect you'll notice that I confined myself to correcting a factual error about the source of the footage (i.e. bodycam vs. cellphone). I didn't weigh in at all about whether the shooting was justified or not. That was intentional, as this strikes me very much as a grey area case.
If I come to a firm conclusion about it later, then I may or may not decide to post something about it then, but in the meantime you'll be better served not to point your own wheels at people unless they've actually expressed the opinions you disagree with.
I've seen the body cam footage. He was in no danger, certainly no danger that stepping to the side would not have solved.
Coming from a country where the police almost never carry guns, I'm eternally fascinated by the American right's ability to justify allowing authorities to execute their own people in the street.
Oh, are we going there? Shall we talk about trampling entire groups of schoolgirls, in an alley? And then turning around and trampling over them again? With no consequences for the officers?
Disclaimer: This is not America. This is one of those countries where the officers "almost never carry guns."
The police don't even try to silence the eyewitnesses, they're so confident of their ability to evade punishment.
This is one of the clearest, most direct examples of whataboutism that I have seen in some time. It's vague as well, inviting another layer of distraction from the original statement; what trampling? What country? How is it relevant?
Please don't answer these questions as it would distract further; I don't really want to spend energy on the deflect.
I'm eternally fascinated by the number of people who come from a place with different conditions to the place they're discussing (in this case, a much lower crime rate) and somehow think this makes them more, rather than less, qualified to pass judgement on the latter place.
I'm also fascinated by how these people almost never specify the actual country they're from. It's always "a country", their nationallty is always "non-American", never their actual nationality. It's very weird; I can't imagine describing my own nationality in such a way. A lot of the American left has this fantasy of this hypothetical non-American western country which has open borders, and lenient policing, and unrestricted abortion, and large wealth taxes, and massive affirmative action and racial diversity, and every other progressive policy that pretty much don't all exist together anywhere. This isn't that surprising; it's just the equivalent of the right's fantasy of some unspecified past time when every one of their preferred policies was in place at once.
What IS surprising is the number of (allegedly) non-American leftists who want to identify as citizens of this imaginary country. Which of course means never specifying or talking about their own real country (only ever talking and complaining about America), probably because doing so would make it clear that such a place doesn't actually exist.
It does nothing but make accurate discussion of real problems and policies much more difficult.
Actually I'm just British, but I enjoyed your paranoia. I tend to keep personal details out of public discussion out of habit. Our police do not carry guns except for very specific situations such as: guarding government facilities that are likely terrorist targets, responding to active firearm/bomb threats, raiding buildings suspected to house active threats.
It is so rare for police to shoot someone that it is always national news and it is always the subject of an investigation. The shooting of Jean Charles de Menezes in the wake of the 7/7 bombings was a national scandal that would have been shrugged off in the us, and he wasn't even a citizen.
The American right will always bend over to accept the police killing citizens because the central tenet of conservatism is that there is a group the law must protect, but not bind and group that the law must bind but not protect.
> I'm also fascinated by how these people almost never specify the actual country they're from. It's always "a country", their nationallty is always "non-American", never their actual nationality. It's very weird; I can't imagine describing my own nationality in such a way.
Tell me you are an American without telling me you are an American.
I am from Slovakia 🇸🇰, which I sometimes mention when it is relevant, but mostly don't, because I assume that 90% of readers couldn't tell a difference between Slovakia and any other Eastern European country -- and there is no good reason why they should. It's practically the same level of detail as an American telling you which *state* they are from.
(Plus there is a small but nonzero probability of someone asking what is the difference between Slovakia 🇸🇰 and Slovenia 🇸🇮, which is annoying for people from both countries. But it's getting less frequent recently, probably because people learned to google or ask a chatbot instead of asking in the forum.)
Anyway, how this relates to the police behavior...
I do not travel much; I regularly visit Czechia, Hungary, Austria, Switzerland. From the perspective of "what behavior should you expect from the police", I haven't noticed any difference. Maybe I was just lucky; but in my model of the world, it's just "standard European police behavior". Perhaps if you break a rule, you are more likely to be asked for a bribe the more eastwards you are; I wouldn't know, I usually don't break rules.
I spent a few weeks visiting USA, specifically California; during that time I got yelled at by a random policeman for the "crime" of keeping my hands in my pockets (I wasn't even aware that I was doing that), but luckily I didn't get shot (it took me some time to realize that the guy is yelling at me, and what exactly was his problem), so I am here to tell the story. Now I don't want to draw unnecessarily big conclusions from a single data point, could be a total coincidence, but even my limited experience agrees with the idea that there is a "generic American cop behavior" which is significantly different from a "generic European cop behavior". So if other people keep saying the same thing, it sounds plausible to me.
Oh, and people own lots of guns in Switzerland, so I think the classical excuse "American cops have to be different, because Americans have lots of guns" fails to explain this part.
I agree that by cherrypicking things from different European countries, you can get an idealized country that is much better than any of its constituent parts. But I think there is also some meaningful similarity, and America is culturally more different; sometimes for better, sometimes for worse.
EDIT: https://www.astralcodexten.com/p/open-thread-416/comment/199631031 " Hell, there are plenty of well-documented incidents of LEOs unintentionally injuring or killing people during physical struggles due to a basic mistake, like drawing and firing their pistol when they intend to use their taser." For Americans, this is Tuesday; I think that for most Europeans this sounds like a quote from a dystopian novel.
EDIT: The hypothetical country that America is compared to seems to be something like "mostly Europe, but also Canada, and excluding e.g. Belarus". Perhaps we should have a simple word for that.
I have watched several French policemen with batons beating on an 8 year old in the fetal position. This seems like dystopia to me. I do not believe they will suffer any from doing this.
In Great Britain, a schoolgirl was arrested for carrying a knife to protect her kid sister. (The video is from adults harrassing the children). The inevitable occurred in the aftermath.
In Italy, someone's on record saying that a rapist "cannot be expected to not assault Italians on the beach" because "he doesn't know better!"
I have heard that in Hungary, it is a VERY bad idea to smile at the policemen (seen as very disrespectful, may lead to arrest). Can you confirm?
The German Government (Okay, one particular highly-placed guy in the government) has had SWAT teams called over 5000 times, for people saying mean things online. This seems like dystopia to me.
Perhaps eastern europe is just more functional?
American cops are some of the best in the world (Yes, there are bad apples, and really bad guys getting passed from town to town, but still!), and part of that is our willingness to have there be actual consequences.
One side of the political split claims that she did not pose an imminent threat, the other side that she did, was about to run over the agent who shot her. I have viewed multiple videos and am still uncertain how much danger he was in. I have observed, on DSL, intelligent people who are sure it was legitimate self defense and intelligent people who are sure it was not. I have also observed blatant lies in both directions, by Trump and Vance on the one side and by people who simply don't mention that she was driving a car that was accelerating in his direction when she was shot.
If you have not observed arguments about this by intelligent people on both sides you should not assume that you know what happened.
Watching the officer's cellphone video, I think it's of note that the first time the officer walks in front of the vehicle, Renee is turning the wheel to follow his position.
I've only watched the one angle; from that one it didn't look like she was moving at all. It may or may not have been incidental, as she leaned out the window to talk. But it's something that can put the guy in the mindset of "she wants to run me over."
She wasn't pointed at him, reverses, points the wheels at him, looks him directly in the eye, and then guns the car (skidded out on the ice). This is what a video analyst had to say.
Mental state unknown, may not have intended to run him over (wife shouting "drive baby drive", for example) -- but him being in the way was not a disqualifying "why I shouldn't gun the gas"
You can see at the 28 second mark the wheels are already spinning while pointed in his direction. It's clear because we can still see the hub caps. She appears to be in the process of turning them, but has already started accelerating in his direction when he pulls his weapon at 29 seconds. By the time he fires at 30 seconds they are pointed away, but at that point the decision to fire has already been made. He didn't have the luxury of doing a sub-second breakdown of the wheel movement at that time.
I disagree, he definitely had that luxury, especially since, as you said, the car was turning before the weapon was drain. You’re calling the agent so negligent as to ignore the path of the car, which is nearly as bad.
Reaction time is .3 milliseconds, and that's for a quick draw. Give him a full second, because he has to drop the cellphone, draw -- and aim at a person in motion.
I'd say that's within "give him the benefit of the doubt" -- if I was in a life-and-death situation, I might pull a gun, even if it was an objectively dumb thing to do.
(1/16) Edit on top of post for visibility: Active Self Protection has just posted a comprehensive non-partisan analysis by a professional forensic expert witness in shooting encounters. It includes all the useful raw, unedited footage available, and then his analysis. He addresses topics like why Ross was likely *unable* to stop firing after his first shot, and why it's likely Good never even saw him when she took the car out of reverse to drive forward.
Some reasonable pro-guns-for-self-defense folk I follow on YouTube had hot takes that the driver was *technically* in the wrong just enough that the officer shouldn't (and won't) face criminal charges as he can credibly state he was in fear of his life while/after getting struck, but also that the shoot wasn't as "good" as we'd all generally like to see in a self-defense encounter. The word "unfortunate" is getting tossed around a lot, and that's about where I am.
But really?
It's not about the ICE officer.
Ultimately, the driver had a duty to herself and her family to not *deliberately* and *avoidably* endanger herself. As I stated on the Hidden Thread and will cannibalize for here, the driver was incredibly foolish to feel so secure about defying the commands of people legally carrying firearms to enforce those commands. Plenty of instances of police abuses have been covered in the media (and we can assume she was exquisitely aware of them), and there are tens of thousands of badgecam videos on YouTube demonstrating the immediate consequences of defying police officers. Hell, there are plenty of well-documented incidents of LEOs unintentionally injuring or killing people during physical struggles due to a basic mistake, like drawing and firing their pistol when they intend to use their taser.
Illegally confronting the police is *dangerous.* Rarely, legally confronting the police is dangerous! Don't fucking do it, no matter how passionate you feel about your cause!
Yes, this is victim-blaming, and it is absolutely warranted, because the driver actually is to blame for foolishly risking her personal safety. She would be equally to blame had the video been of hopping the guard rail of a Grand Canyon viewpoint to get the perfect pic and then falling to her death, or chasing down a wild grizzly bear cub to pet it. There are dumb ways to die, and this was one of them.
There is footage going around of Renee's wife (IIRC) shortly after the vehicle came to a stop. She's yelling at the officer: "Why did you have real bullets!?"
If that's an accurate transcription, it's consistent with and highly suggestive of both of them having a very inaccurate idea of what LEOs normally do. It makes me think they believed they were working under completely non-lethal rules of engagement. That would mean they could do whatever they wanted as long as they knew they weren't trying to hurt anyone.
This also leads me to believe that while Renee Good may have held ICE agents in contempt, she had no inclination to actually hurt them, including hitting them with her car. However, she could still be found to have had reckless disregard for safety. I could be persuaded otherwise, but the footage I saw suggested she was looking at an officer at her 9 o'clock while pulling away, and failed to see Ross at her 12. In a residential street with no LEOs around, I could see making that assumption. In one where I've gone out of my way to obstruct traffic, I can see that not being a useful defense. (And on a street covered in snow, it's even less defensible. What if that had been a kid chasing a ball?)
This is completely correct. She willingly courted danger and suffered the consequences. The risk of being shot by cops is why you don't provoke them. Don't force an armed man to make a split-second decision unless you want to run the risk of getting shot.
Of course the shooter had a poor reaction and made a poor choice. I would love it if every police officer reacted perfectly 100% of the time, but that's simply not realistic. I would assign the blame here as 90% Renee Good and 10% ICE. If all police officers were replaced with perfectly-trained replicas then I expect that that would reduce incidents like this by 10%. If all protestors simply obeyed the law then I expect that incidents like this would be reduced by 90%.
That's funny, I actually drafted in a different comment that I felt it was about 90% Good's fault, but then got worried I'd have a big fight about the particular number, and I'm not really married to it enough to want it to be the main focus of a comment war, haha.
While I don't necessarily disagree with you, I was very surprised to hear this from the same person who claimed that guns would protect women from the patriarchy. Well, at least it means that you won't throw away your life if and when things escalate...
Hmm, I seem to recall saying that it's odd feminists don't make better use of guns if they're so worried about the patriarchy, but I can't find the exact comment.
But also, as a rational hedonist who firmly believes that the quality of my life is more important than minutes lived, I'm fully prepared to use the tools I have available to me to end my severe suffering should there not be a strong likelihood of it being relieved. I don't like cooking from scratch and cleaning enough to be a Martha, and I would absolutely provoke a shooting before allowing myself to be transported to the Colonies.
Well, now I can't tell if we're flirting or not, given the exquisite setup you just provided me for a very funny, very tasteless joke that would be very fun to post.
But while it appears Scott isn't banning people anymore, the chance of getting reported and then banned for my joke is nevertheless nonzero, so I'll hold my tongue and hope those who would appreciate said joke will likewise see the setup and get a chuckle out of it.
We typically hold the police to higher standards than this; if we didn't, police would be racking up double digit fatalities from knuckleheads and drunk guys confronting them every long weekend.
In other words, the ICE officers are (or should be) trained LEOs who are safer to be around than grizzly bears.
They are. I would much rather be around an ICE agent than a Grizzly Bear. Bears attack people indiscriminately. All I have to do to avoid even the threat of being attacked by an ICE agent is take the exceptional step of not actively interfering with them in the performance of their duty.
Citizens shouldn't obstruct justice or resist arrest. When you force armed men to make split-second decisions then you run the risk of getting shot in the head. We typically hold citizens to higher standards and expect them not to actively interfere with law enforcement.
We also generally ask that knuckleheads and drunk guys not actually use broken bottles and other lethal weapons. Fists are harmless enough (nonfatal) that a "bearhug takedown" is a reasonable level of defense.
By the point that a lady is using a 2 ton vehicle, you've moved well beyond "let's just toss the guy to the ground" and into "that guy is TempleOS crazy."
> In other words, the ICE officers are (or should be) trained LEOs who are safer to be around than grizzly bears.
Yes, fine, great. As a civilization, we should get right on that.
Until then, *right now,* confronting trained LEOs - especially with illegal means - comes with the risk of being unjustly arrested and/or injured, or killed.
If an individual does not freely accept the risk of being unjustly arrested and/or injured or killed, they must not confront trained LEOs with illegal tactics, *ever.*
You think I'm being very naive, but I think "cops shouldn't just shoot people unless they have a very good reason" is more important than "people shouldn't provoke cops", even though both are generally true.
Your feeling about the relative importance of those two principles doesn’t have any impact on our shared physical reality where cops are occasionally wrongly shooting people.
What impacts reality is your actions.
If you feel so passionate about police overreach that you're willing to *consciously* risk your life confronting police on camera for the cause, that's something I can sincerely respect (unless you have kids or other dependents). I think it would help if you record a video explaining you fully understand the risk you're taking for your principles, so people won't suspect you felt so entitled to ignore the obvious danger that you might have made a surprised Pikachu face when the bullets hit.
But if you just thought, "What? No, of course I wouldn't risk anything in *my* real life proactively confronting a LEO, not in actual, physical, real-real life where I could lose my job or be really hurt or *die,*" well...
"If you feel so passionate about police overreach that you're willing to *consciously* risk your life confronting police on camera for the cause,"
The fact that you apparently take it as read that "confronting police on camera" constitutes an automatic risk to life and limb is exactly the problem.
I don't think it's naive or idealistic to believe that nobody in a civilized society should have the right to summarily murder anyone else. Obviously people have the right to defend themselves, but the gulf between "confronting somebody" and "immanently threatening their safety to a degree that requires lethal force to counter" is generally quite vast. It certainly was in this case.
> Yes, this is victim-blaming, and it is absolutely warranted, because the driver actually is to blame for foolishly risking her personal safety.
Everything else aside, a country where this is the "default" take is not the country that I want to develop for my community and my kids. Police (of which, it should be noted, ICE is not) should have *more* accountability, than the random strangers they interact with, not less.
Do you want your kid to reduce their chances of being killed by law enforcement or not?
Because they're the ones who have the most control over whether or not they choose to initiate *inherently dangerous* confrontations with law enforcement. You should probably tell them not to do that if you haven't already.
I am curious about what window of time supports the claim that Good was illegally confronting the ICE agents.
The videos I have seen begin by showing a vehicle leave the scene by driving in front of Good's Pilot, then Good waving the ICE vehicle to go around her. My understanding of the facts:
(1) ICE was not in the middle of conducting an immigration enforcement action
(2) ICE does not have authority to conduct other forms of law enforcement action, including traffic citations
(3) Good's vehicle was not preventing the ICE vehicle from proceeding.
(4) ICE agents left their vehicle to engage with Good.
(5) The shooter (presumably Ross) circled Good's vehicle, ultimately positioning himself somewhere to the front of her Pilot (but definitely not squarely in front of the vehicle.)
(6) Three shots were fired
(7) The Pilot continues to the right (direction of traffic on that street), straightens and hits another parked vehicle
(8) ICE agents, including the shooter, were in no rush to render aid to Good
Some things that are unclear:
(A) Did Ross transfer his phone/camera from his right hand to left hand? This appears to be the case from the video he took and released
(A') Does A indicate any form of pre-meditation
(B) Why was Good's vehicle in that location?
(B') Did Good engage with ICE/other LEO prior to the first vehicle driving away/around her Pilot?
(B'') Was her prior engagement illegal?
(C) What orders were being given to Good?
(C') Were the orders being given to Good coherent from one agent to another? In other words, was compliance possible?
(D) Were the post-shooting actions of the ICE agents consistent with existing laws and policy?
(E) Was Ross struck by the vehicle?
(E') If yes to E, was it possible/reasonable for Ross to avoid being struck by stepping away from the vehicle?
My current impression is that:
(I) The shooting does not appear justified
(II) The take-away for people who want to avoid this outcome is a blanket "avoid all contact with ICE/LEO." However, without knowing what led up, it is hard to know what action by Good would have achieved this.
Good's vehicle was fully operational, as was demonstrated when her foot hit the accelerator and crashed the car into another vehicle. She wasn't broken down in the middle of the road or stuck in the snow; she was deliberately blocking a roadway to intentionally confront law enforcement.
That is very dangerous for the reasons I outlined, and she could have prevented her death by choosing not to do any of that.
That fact doesn't really have anything to do with whether the shooting is justified. But I'm emphasizing it because she deliberately put herself in danger, and that's bad and stupid, and the only useful takeaway for the rest of us is *don't antagonize people with guns,* because they might, rightly or wrongly, use them to kill you.
Clearly, not enough people are getting that message.
I am very confused. I have trouble lining up your answer with what I wrote.
>>Good's vehicle was fully operational,
As far as I understand, this is not disputed.
>>when her foot hit the accelerator and crashed the car into another >>vehicle.
Crashing into the other vehicle was after she was shot, no?
>>she was deliberately blocking a roadway
From the videos I have seen, this doesn't appear to be true.
Repeating what I wrote above: "The videos I have seen begin by showing a vehicle leave the scene by driving in front of Good's Pilot, then Good waving the ICE vehicle to go around her." In case that wasn't clear: Good is seen waving the ICE vehicle to follow the vehicle that has just drive up the road, showing that her Pilot was not blocking the road.
Is the description of the situation before the ICE agents got out of their vehicle and approached Good not correct?
The way you wrote it ("Good's vehicle was fully operational, as was demonstrated when her foot hit the accelerator and crashed the car into another vehicle.") read somewhat confusing to me as well. It reads as if she hit the gas and immediately crashed into another car. But most people referring to her hitting the gas and almost immediately turning to head down the street. She only hit another car because she was probably dead by then, and her SUV was presumably an automatic.
I don't think it functionally disputes the rest of what you wrote.
I agree that she did something stupid, but I do not think that her doing something stupid is enough cause for absolving the officer of any blame on the incident.
I don't think the officer should be absolved, either (and less so today, after watching an analysis I hadn't seen before); at the very least he should lose his job and pension and so on. Perhaps criminal charges, although which ones, I don't know.
But the reason I'm emphasizing Good's stupidity is that it's the only useful object lesson *we* can take away from the incident. None of us have the power to change the policies at ICE, or ensure our police departments only place emotionally-regulated, well-trained officers on duty.
What we *can* do is spread awareness of the inherent danger of confronting a random LEO, and especially confronting them while breaking the law, when they might feel even more justified using force than they would if you were indisputably correctly exercising your 1A rights standing on the sidewalk and so on. We should break people's sense of entitlement to not have to feel any fear of LEOs, because no one is so important to the universe that their passion will plot-shield them from the reality of other people potentially reacting very negatively to *SUPER!SPECIAL!THEM!* should they start a fight.
Huh. Something wonky is going on with Substack! Obviously, I finished the sentence in my final paragraph with "one of them," however, Substack is apparently refusing to grant me another line of text to show it.
I notice it too, and there is another comment in the thread where precisely that happens. My guess is that the length of your comment was just on the edge of when it collapses with a "See more" tab, and they haven't coded that correctly.
I think I’ve seen all the videos and I can’t really tell whether the woman was on course to hit the ICE guy. But I can easily believe that in the stress and confusion of the situation the person with a car even sort of coming at them would be quite frightened, and fully believe they were in danger of being hit.
The fault, if there was one, was in the ICE guy’s training. In the places where these confrontations are taking place having somebody aim their car at you is probably one of the likely dangerous situations. And there are a things a trained person would know that would influence how they act. For instance, how badly are you likely to be hurt if the car is going 10 mph or less? Also, if the car is coming right at you, at any speed, shooting the driver isn’t going to help you. Even if you kill the driver, the car’s still going to hit you at whatever speed it’s been going. Seems like your time is better spent getting out of the car’s path.
This professional forensic expert witness on shootings and other violent encounters thinks Good almost certainly had no idea Ross was in her path (commentary on Good's likely tunnel vision begins at 23:44):
10mph is a six-minute mile, so it's like hitting a waist-high wall at full run.
Which is why he should have never been in front of the vehicle. If they wanted to stop her from running they could have parked a car there, not a human body.
And why exactly would they want to stop her from running? As I understand it, the problem was that her car was in the way, and she was bothering the agents. If she drives away, both problems are solved. It's not like there was a warrant out for her for some serious crime, so God knows what happens if she escapes or something.
(The answer is probably that the agents switched from a mindset of "we need to get her to leave to we can do our work" to "she's the enemy, let's take her down".)
She was behaving as if it was "if she leaves, no harm no foul."
But this is an iterative game, and this wasn't the ICE agents first playthrough.
If they, upon the fiftieth time, decide "jailtime for Miss Nancy Drew", that is their right as law enforcement officers trying to do their job. She's in the wrong, and interfering with lawful enforcement of the law.
Time was, the left knew that "going to jail" was what would happen if they staged a sitdown. Letters from Birmingham Jail, and all that, no?
Well, they're 'solved' in the same way that making a thief return the goods they just stole is 'solved'. She was deliberately blocking traffic; I think a night in jail would have been a reasonable outcome.
Because these idiots are doing a "rolling blockade" where each car obstructs the agents on one block until forced to move on, then another one obstructs them on the next block. if they don't arrest her then she'll be harassing them all day.
With a dead driver, the car is less likely to swerve towards you when you dodge out of its way.
Not really relevant when the car is as close as it was in this case, but I can see it being standard procedure one is trained to reflexively execute (no pun intended).
Unless the driver turns the vehicle AWAY from you, which was the case here. (...see the cellphone footage...) Shooting the driver can, however, stop this turning, thus you will be in more danger.
It was clearly a reaction fueled by rage. Which can considered to be acceptable ("he had to decide within a split of a second" and all this stuff), but whoever reacts this way ("within a split of a second" or not) definitely should NOT carry a gun.
(BTW the driver had to react "withing a split of a second" as well -- perhaps federal agents just should not scare the hell out of everyday citizens? Just an idea.)
> Unless the driver turns the vehicle AWAY from you, which was the case here. (...see the cellphone footage...) Shooting the driver can, however, stop this turning, thus you will be in more danger.
But the car wasn't pointed at him, hence why he only got scraped. As far as self-defense goes, killing her was not that dangerous.
I don't think the driver gets credit for anything she did with the steering wheel when she hit the gas pedal.
I have no love for the state, and would condemn anyone for willingly acting in its service. But even I do not say they deserve to be deprived of so fundamental a right as effective self defense.
I didn't design their curriculum, and don't know whether the person who did is an idiot, or what constraints they were operating under, but if I was doing it, "draw quickly, keep hands steady, and fire rapidly," immediately upon recognizing a threat would be a skill I'd drill a lot, below only general fitness and target practice, assessing it to be applicable in a wide range of situations.
Training specific to cars being used as weapons is obviously harder: it's clearly insane to drive normal cars at speed towards your trainees to teach them how to react. If I could have custom autonomous cars built that stopped much more aggressively, maybe with explosive charges shooting spikes into the road below or something as emergency braking, then yeah, I could see a few days spent on that. That sounds fun! Just imagine standing in the middle of a field with cars barreling towards you from multiple directions, and you have to dodge them, and "kill" their (imaginary) drivers.
At least in principle, shooting the driver while they try to reverse in the middle of a city street could be an effective countermeasure, in a way that shooting a driver who is a few feet away and "comin' right at us!" really, really isn't.
But it would be a countermeasure to something that basically never happens. People occasionally kill police officers with automobiles, but as a one-and-done in an attempt to escape. The sort of persistent attack you're describing, is vanishingly rare.
If you're training cops to not die, you teach them to get out of the way of the car. If you're training cops to get away with killing people, you teach them to stand in front of the car and say the "he was comin' right at us" thing. And yes, if you're a really stupid police trainer, you might confuse the one with another and train your students to kill and/or die needlessly. Alas, it seems to me that there are a lot of stupid police trainers out there.
Sometimes I think about things, and think "that would be a good subject for an ACX open thread" and then I forget whether I already started a thread on that very subject. Here is one such subject, possibly a repeat.
What's the relationship between ethics and etiquette? Are they two distinct things, or is it more of a continuum? Both are lists of things that you should and should not do, ranging from "thou shalt not kill" to "don't butter your whole bread roll at once". Both, if obeyed perfectly, make life more pleasant for everybody.
But there's differences too: ethical lapses tend to be more serious than ettiquette lapses, and ettiquette is more culture-dependent while ethics tends to be more (but not entirely) universal.
And then there's grey areas; like using a racial slur. Everyone would agree that this is impolite, while some might argue it's actually immoral.
Can we meaningfully distinguish ethics from etiquette? And if we can't, does that mean we should pay more attention to etiquette?
Etiquette lets you know what to do in different situations so as to avoid giving unintended offense; we no longer have a generally accepted etiquette in this country, so it's hard to tell if someone is deliberately being rude to you, just being antisocial, or completely clueless if they block a doorway looking at their phone or are playing loud MMA videos with no earbuds on the bus.
In a well-operating etiquette system, ethics doesn't come into it, everyone is supposed to be doing the "right thing" and you can rely on that.
Many people who focus on ethics put way too much emphasis on it compared to other things we care about. Ethics is not the end all be all of how to live your life. There should be more philosophical inquiry in to a broader range of questions related to how to live your life.
People like having class markers in the absolute (even the ones who don't like some specific class markers) and having them it makes their life more pleasant. The proof is in revealed preference : whenever class markers get suppressed some others appear spontaneously.
I think philosophers, while distinguishing prudence/self-interest from ethics/morality, sometimes distinguish norms/etiquette from both of them. Other times it's folded into one or both of the former categories.
It's definitely a weird category. Probably a very broad category, as you note, parts of which are firmly under "prudence", parts of which are firmly under "ethics", and parts of which are arguably...if your philosophy allows for some sort of irreducibly social aspect of what a person is...in a distinct category that's not quite either. (Roughly: you in some sense "have to" follow norms even if they don't benefit you, but not doing so isn't *immoral*, it's some sort of separate, purerly social failing.) And I've even seen some who seem to consider it part of *aesthetics*.
Never thought about that before. I'd say it's a continuum, but the big difference is a breach of ethics incurs punishment. If you put your elbows on the table while buttering your whole bread at once, it's a breach of etiquette; you'll make people uncomfortable, but they probably won't throw you out of dinner. If you steal a fork, it's a breach of ethics; if they catch you they'll get the fork back with interest.
For a moment, instead of 'ethics', consider that modern usage of the word ‘value’ keeps confusing different things:
- What people’s motivations show as ‘desirable’ (‘These goods will fetch good value in the market’; ‘The community showed how much it valued the church through high attendance’)
- The pluralistic and contextual norms different societies use to help them coordinate and succeed (‘These values represent who we are’)
How does not buttering your roll all at once make life more pleasant for everyone?
I think there are two distinct types of etiquette. One is sort of a mini-version of ethics, concrete rules rooted in more general ethical guidelines, specifically meant to make life better overall. E.g., putting away your weights and wiping machines after you're done with them is part of gym etiquette meant to make the gym a more pleasant and safer environment for its members. That sort of etiquette we should pay attention to.
The other type is barely related to ethics. It's about distinguishing the outsiders from those who belong. Elaborate silverware rules during fancy dinners are an obvious example.
I think a rule about avoiding racial slurs is definitely driven by ethical considerations, so I don't think it's a grey area.
“ E.g., putting away your weights and wiping machines after you're done with them is part of gym etiquette meant to make the gym a more pleasant and safer environment for its members. That sort of etiquette we should pay attention to.”
I’m not sure if this is an instance of etiquette. (It’s a not “using the wrong fork” kind of thing.)
It’s an example of “leaving things as you found them”. That is a fairly pragmatic and efficient way to deal with shared resources. Otherwise, one makes work for other people.
> How does not buttering your roll all at once make life more pleasant for everyone?
agree with what's been said - but also note that many evolved systems are 'over-engineered' to survive tail events (hence two kidneys)...thus will look like they "don't make sense"...*even on a normal day in their ancestral environment*. A small 'lapse' in coordination may make a big difference on 'game day' if the stakes are existential. And that's not even taking into account the fundamentally zero-sum nature of status dynamics (at least, along a given dimension).
But fundamentally - my prior is for most ppl you're thinking of, you're trying to reason them out of positions they never reasoned themselves into...appealing to e.g. material abundance/'pleasantness' empirically will fail. Happy to be proven wrong of course.
Not really, rules like what foods are ethical to consume (such as the kosher rule against odd-toed ungulates), what beverages are ethical to potate (like the Mormon prohibition against alcohol and "hot drinks"), what words are ethical to say (like the taboos in Polynesia), what clothing is ethical is wear (various detailed specifications about colors and materials), are fundamentally about distinguishing insiders and their class, though of course they traditionally confabulate reasons involving religious/spiritual/ethical considerations.
> How does not buttering your roll all at once make life more pleasant for everyone?
A lot of table manners are about signalling to your dining companions "I'm like you, you're like me, we were all brought up in the same class of the same culture and all follow the same semi-arbitrary rules". Asking why that's pleasurable is like asking why dining together is pleasurable to begin with -- I don't really know but it really is something that humans like.
The idea that certain types of ettiquette exist purely to exclude lower-class people is, I think, wrong. High-class people would absolutely _love_ it if low-class people started exhibiting proper table manners.
“ The idea that certain types of ettiquette exist purely to exclude lower-class people is, I think, wrong. High-class people would absolutely _love_ it if low-class people started exhibiting proper table manners.”
Is anybody claiming this idea?
In any case, there’s no actual contradiction here.
Things (not just etiquette) can be class markers for in-groups that welcome “converts”.
Let’s discuss Missouri. For a large state it seems to have a surprisingly low profile. You just don’t hear much about Missouri. In my view this is due to several reasons:
1. Missouri is not particularly important in national politics. While it was once a swing state, in the last couple decades it’s turned almost entirely Republican. This means the Democrats consider it a hopeless case and the Republicans take it for granted.
2. Its two major cities are both right on borders and therefore their influence is shared with other states. In fact one of the cities shares its name with an adjacent state and its best suburbs are in that state.
3. Missouri’s top city for tourism is quite close to a border too.
4. Speaking of borders, the super dynamic and prosperous Northwest Arkansas region is a bit too far from the state line to have any significant spillover effects in Missouri.
5. Its capital city is comparatively small and little-known.
6. Missouri lacks any glamorous trendy cities like Austin, Nashville or Columbus.
7. There are no longer any airline hubs in the state.
8. The state university is not a powerhouse in football or basketball. Pro sports? Lost the Rams, will be losing the Chiefs.
9. Most of all, Missouri has no regional identity. In some respects it’s part of the South and in others part of the Midwest, but does not fully belong to either.
It's interesting to think that St Louis hosted the Olympics in 1904, slotting in between Paris and London, and this seemed perfectly reasonable at the time.
Actually, it was a pretty controversial choice! STL basically strong armed the IOC into getting the 1904 Olympics because they had the World’s Fair the same year. Even though standards were low at the time, people were disappointed in the quality of the facilities, and a lot of nations simply didn’t participate because STL was hard to get to (you had to take a boat to a port on the east coast, then get a train, usually routing through Chicago. This would take days!).
5 and 6 are closely related. It’s not a coincidence that Austin, Nashville, and Columbus are all capital cities that are large enough to start a virtuous cycle in the past decade or two, despite being historically smaller than the other cities in their states.
The solution I'm seeing is to fiddle the border slightly, give Kansas City to Kansas, which would make freaking sense, and leaves Missouri without an identity crisis.
Even better, there's also a "MaxDB" named after Maria and My's brother Max. Honestly, Maria's the odd one out, is it just a weird coincidence that 2 out of 3 siblings have names that can be mistaken for common english words?
Look at it from a different perspective: Everyone's got something weird about them. And everyone gets teased about it. Only the narcissists can't take the teasing. "I'm not perfect! The world is ending!"
Teasing is also used as a gentle corrective for people outside the norm. Sometimes this can get out of hand, but some people are harder-headed than others.
I did my high school years, 20s, and 30s with mostly male groups of nerdy dudes, including a male-best-friend-and-spiritual-twin-brother for all of it. We all said unacceptably cruel things roasting one another as an ongoing bonding ritual, partly as a trust exercise that we *could* say unacceptable things without destroying the relationships, but also because clever verbal sparring was performative fun, and the goal was to make people laugh. Zing!
One of the few actually unacceptable things was to indicate sincerely hurt feelings, as that was a profound violation of the social contract to not take roasting personally. Anyone who was visible about his hurt feelings - and I say "his," and don't include "or hers," because I never was - would be subject to even *greater* ridicule as punishment for violating the contract and caring about personal ego more than group dynamics.
This is great, because today there is very, very literally no slur, zinger, or insult anyone could say to me which would provoke even a mild emotional reaction, much less hurt my feelings. I was called fat 5,000 times directly to my face; you think I'm afraid to order whatever I want in a restaurant? You think words like "bitch" or "cunt" are a push-button for provoking me into a reaction? lol, no.
Teasing *is* essential fun. It's verbal sparring, and, like sparring, the point is to make one stronger, faster, and more resilient, in preparation for engaging in real conflict. Developing a capacity to not have one's feelings hurt by mere words without accompanying action is priceless.
It's good that it worked out for you, but that wasn't how it was for me.;
Perhaps it was because I was socially isolated at school. I suspect a campaign of ostracism.
The way teasing worked for me (I may be somewhat autistic) was that being teased meant I was fundamentally defective and unwelcome. That's why I was being harassed for being short and having feet that turned out.
For what it's worth, when I tried to tell my mother that it hurt, her immediate response was "What did you do to them?", so my model that teasing was an attack wasn't exactly what I grew up with.
Well, again, I have to wonder if "teasing" is actually the correct word you should be using. You sound like you're describing "bullying."
To continue with the metaphor, if "teasing" is consensual and consciously partnered sparring, "bullying" is nonconsensual predatory assault.
Unfortunately, the difference between them is not the content of what's being said, but whether the exchange is mutually consensual. The "teasing" I participated in sounded like breathtakingly cruel "bullying" in that typical rules around sensitive topics and words were not observed, but it was a consensual bonding ritual.
Meanwhile, a much more gently worded personal jab, if intended to make someone feel bad rather than laugh, can externally sound like "teasing" but in reality be "bullying," and it's not unusual for a perpetrator to dismiss the bullying by saying, "I'm just teasing!"
I was subjected to some bullying as a kid, and until middle school was very decidedly in the low if not bottom position of the social totem pole, with hair, clothes, glasses, and quirky personal interests being sharply mocked by people who weren't my friends, with the clear intention to hurt my feelings (although I stubbornly refused to admit when they did). I moved off the social totem pole entirely by accidentally becoming a hyper-local figure of fascination and then consciously embracing the opportunity to entertain people out of wanting to bully me.
It worked really, really well.
The comedian Brad Williams, who was born with a form of dwarfism, has talked about how his father consciously began teasing him about it long before his peers had the opportunity to do so. He's discussed that strategy informally in podcasts and such, but also entertainingly as part of his his act (https://www.youtube.com/shorts/WzzY4mW81SI).
While it's not exactly the same as what I was doing with my dudes, it's a similar idea of verbal sparring to strengthen emotional resilience. I don't "tease" or "bully" random strangers (how would that even work?), and I am scrupulously warmly civil to people I dislike.
But there will be a point during an acquaintanceship where I'll begin gently teasing for comedic effect to both entertain and test the other person's emotional resilience as a candidate for something deeper.
If they do anything but spar back, I will back away from deeply bonding with them, because I can't trust them to set aside their reactive personal feelings long enough to analyze someone else's intention - or anything else. That testing process is a product of experience; the people who've hurt me the worst in my life all valued being "sensitive," and that made them dangerous.
Suppose that, the majority group really does feel like youre lame and they dont want to spend time with you. Is that illegitimate per se, and if no, what do you want them to do about it? Like, humans do intentionally repulse others they dont want to be with, and you can quibble with the ways, but if *the problem* is that "it meant youre unwelcome", its pretty difficult.
I mean, ideally a majority group would simply ignore the person they don't want to have in the group, rather than engaging in proactive cruelty with comments and/or physical bullying. "Go away!" should be enough.
But your point that it's impossible to browbeat children into being indiscriminately kind and welcoming to all is an interesting one.
My friend is raising her seven-year-old girl with the gentle parenting method, including sending her to a very exclusive, hugely expensive Montessori private school, where most of the other affluent doting parents are also of the "gentle parenting" mindset, and the school staff are all about inclusion and togetherness and kindness. Despite the enormous forces acting on these kids to be universally kind, the girls in particular have sorted themselves into wolf-packs who are constantly inviting and ejecting members as part of a ruthless internal struggle for "queen bee" status. It actually sounds quite a bit worse than I remember my friendgroup being when I was that age in the 1980s, when kids weren't continually being subjected to "be nice, be nice, be nice" in all of their media and education.
Clearly, environment and behavioral conditioning isn't able to overcome the instinct to be selective about one's associates. And since that's probably something humans will never be able to overcome, there need to be other strategies for coping with rejection than telling everyone to "be nice."
Yes, though we also dont know if normal "go away" wasnt tried and failed. Its certainly plausible that it would work for ACX denizes, but I think most kids/teenagers who want someone to like them wouldnt be deterred by that. Really though, I think the amount of contact students are forced to have is just too much for them too feel like a classmate is ever *really* leaving them alone, even if through no fault of his own. (I dont know about the US, but I noticed a very abrupt disappearance of anti-bullying messages going from ~highschool to university. We didnt suddenly get wiser over that summer, and the demographics were mostly the same, too - but you *are* much less penned in, after.)
I also did see some teacher interventions that backfired. Led by the idea that everyone *should* be friends, they made us get to know each other better, and the closer contact led to a stronger repulsive response.
I agree with this. Teasing was never something which happened in an in-group - I didn't in fact have an "in-group" for my high school years. Teasing was something other people did, and as far as I can tell, it was for their entertainment and/or to make me go away. Maybe if I'd had the kind of in-group Christina describes, it would have been different, but I didn't, and teasing still happened.
People who think “normal” exists are sneaking in a particular background that the term is relative to. They don’t mean “normal person”, they mean “normal upper middle class resident of the Detroit suburbs who goes to church every once in a while and has kids” or whatever.
I mean, there are definitely some people who wouldnt seem normal in any place or time on planet earth so far. It is also *normal* to adapt to ones situation. Just because this guy, as he is in his current context, is different from normal people in a different context, doesnt mean he cant be normal, absolutely.
No, I think normal exists, and I don't mean that at all. Possible to be a normal person with a working class life, a normie doctor, normie software developer, etc. and I would see each as normal in the same way. I think "normal" for me is very strongly gut feel based, and I think it's that I'm picking up on something in the body language. And I have nothing against weirdoes of course, they're my favorite people and I think I'm weird myself, though disturbingly, times I have brought feeling like I'm weird to other people they say I seem normal, and I've seen friends that I see as normie also tell me they feel like they're weird or autistic.
No, I actually have a friend that whom I asked this, he just said of course, none of these caveats as to contextuality and relativity. I myself am in the "of course" camp, I think contextuality only comes into it if you're evaluating a completely different culture, but each culture has such a thing as normal (and being part of a subculture does not necessarily mean you're not normal, it depends on the subculture).
In theory everyone could fill out a questionnaire and grade their reactions to a list of particular circumstances e.g social situations, work, friends, moral dilemmas from 1-100.
It think that some people see the term normal as normative, and indeed saying something is *abnormal* does seem pejorative. This doesn’t mean we can’t use the term scientifically, or mathematically - grouped around a mean or median. Normal height, normal weight etc
Normal people respond quite well to guilt trips, they work on an instinctive emotional level that autists fail at. (autists tend to work out "someone's trying to guilt trip me" on their way home, and then get mad about it.)
How would you measure how susceptible someone is to guilt trips? I also suspect that particular metric does not correlate that well with someone's presence or absence on the spectrum.
My view is that "normal" can refer to either being in the central cluster of a bell curve of some particular trait or behavior or to adhering to prescribed norms of what one should be or do. These are both relative in some respects, the former relative to the population being considered and the latter relative to what the norms are and who is prescribing them.
There is probably a large overlap between the two senses of normal for most traits and behaviors, although I can certainly imagine scenarios where a prescribed norm is more honored in the breach than in the observance and it is thus normal (sense 1) to be abnormal (sense 2). Speed limits are a trivial example of this: it's typical to drive a bit faster than the speed limit when conditions permit, but doing so is a violation of the norm of following the law.
There's a story that's widely used as a parable in systems design, of ergonomic studies being done in the years following WW2 to try to design the ideal cockpit for a fighter plane. The original approach, the story goes, was to get tons of detailed body measurement data from a large population of pilots and design the cockpit for a pilot who is perfectly average in every respect: overall height, arm and leg length, hand size, etc. They wound up with something that was comfortable for approximately zero pilots, as even though on any one trait most pilots were close to the average, most were very different from the average on at least one measurement and essentially all off them were off by a moderate but significant amount from several measurements.
Like the fighter pilots, most people will definitionally be approximately normal (sense 1) in any given dimension, but if you're considering many dimensions then you'll quickly get to a point where nobody is normal in all of them.
It's also a category error to apply the concept of normal (sense 1) to strongly bimodal traits. If you're looking for a person with both the average number of testicles and the average number of breasts, or even just one or the other, then you're going to have quite the search ahead of you.
"When one of my children was 2, he dropped a carrot on the floor and refused to pick it up.
He was told he would not be allowed to do anything fun or have any privileges until he picked up the carrot.
He refused for hours. He cried and screamed. He even napped, woke up, and continued to refuse. His commitment was impressive.
We held fast. Until he picked up the carrot, he would receive nothing other than water and basic food to live.
After over 4 hours, he picked up the carrot and apologized. It's the longest he's fought us to this day.
Disobedience is a choice of parents. You get the behavior that you tolerate."
This is the most minor of punishments. It's much gentler than anything that parents in previous years would have put up with. And yet, he was called an authoritarian, that he was causing permanent harm, that his children would grow up to be traumatized and resent him, that this was some kind of power trip and that generally there is something wrong disciplining your child. A common thing said was that children can't differentiate between not getting what they want and true suffering so you aren't allowed to make them upset.
Trauma culture has permeated our society and clearly for the worst. Where are people getting these ideas? It's clearly not based on anything real. Is this what people are hearing in therapy? When did all this become commonplace to believe these kinds of things and how did it develop?
I mean, yeah, that's pretty traumatic. A parental figure taking away food unless the child did something incredibly menial? Why not just, ask them why they don't want to do it?
This is a failure of parenting. Wow, you fought a 2 year old for several hours and won. After you're done stroking your ego, maybe you can be a caregiver instead of a wall.
Taking away privileges is not cruel. Seriously where do you get these ideas from? He’s not hitting the kid. He’s not starving the kid. He’s just saying that the kid doesn’t get extras until he cleans up. Why do you believe that?
2-year-olds are absolutely capable of articulating something like that. A failure to do so is a failure on the part of the parent, who hasn't taught their child how to do so.
Both of your options physically violate your kid. Is it really that hard to conceive of a form of parenting that doesn't rely on violence? I can't help but call your ways both old-fashioned and gratuitous.
I think it is the smug explaining at the end that "you get the behavior you tolerate" that really annoyed people, every parent has strong opinions of what works and doesn't based on their own experiences. Personally, I think holding a 2-year-old to these kinds of consequences for four hours when their attention span is that of a goldfish is kind of absurd. Also, did that guy really win the fight, or is his kid dropping another carrot even as we speak?
I kinda think twitter is not representative of reality. It probably could be difficult to come across someone IRL that would disagree with that parent's approach.
That is probably true for this specific example but the broader point about therapy culture and talking about how you’ve been “traumatized” isn’t specific to Twitter.
Well then bring up the topic using an example where the pparent’s response is well calibrated: the behavior they are trying to discourage is one the kid can easily control now; and can’t be expected to just grow out of in a year or so; and is worth putting effort into changing. Also, the example should use a punishment appropriate to a 2 year old child. You described the punishment administered here as the mildest
possible one. Not by a long shot: An effective punishment for a. 2 year old is a stern, unsmiling look, plus giving them a 2-minute time out— they have to go sit in the corner facing the wall, they can’t talk, they can’t bring a toy with them. This was my standard punishment for my daughter when she was little — the timeouts got longer as she got older — and she *hated* them. She often wailed with anger and misery the whole time she was there. I used to worry that the neighbors would think I was beating her, she wailed so loud, and cried out pitiful things— “No! Mama, no! Let me out! Aaaaaagh!”
The example you led with is really dumb, ineffective miscalibrated
I’m not claiming that it was most effective thing he could have done. Just that it’s not abuse. More and more regular parenting techniques are being called abusive now and it’s bizarre and sprung out of nowhere the last few years. I don’t understand why.
Traumatic techniques are ones you remember 40 years later in the nursing home, ones that foreshadow later behavior. "Wow, he really did all that over a *carrot*?"
Well, the word ‘abuse’ is sort of like ‘rape’ and ‘racism’ and ‘Nazi’. It’s not very precise in meaning, but it stirs a lot of emotion, and the person using it usually seems to believe that anyone to whom the label they are using applies is a terrible person. In reality, there are many aspects to each of these things, and of course differences of degree.
Having acknowledged all that I have to say that I don’t think it is absurd to call what the parent in question did abuse. It is certainly nowhere near as bad as, say, beating the kid til he bleeds, but it is maybe 100x as large a punishment as what is needed to be effective. And I’m pretty sure it caused the kid quite serious and long-lasting distress. And the behavior the parent elected to have a major power struggle over is a very silly choice. You want to reserve serious power struggles for behaviors that you urgently need to discourage, and for a 2 year old those are behaviors that put them in danger: letting go of your hand when crossing streets, running into streets, putting their hands into toilets and other hyper-germy things, eating random stuff they find outdoors or in cupboards at home, playing on the stairs, etc etc etc. They are a LOT of things like that, and you’ve got your work cut out for you to get in into their heads that they just can’t do these things.
Anyhow, I’m quite content with labeling what the parent did as bad, dumb, miscalibrated parenting, rather than abuse. But whatever you call it, it is not an appropriate example to start with if you’re going to rant about how the present generation of parents doesn’t understand the importance of firmness, limits, negative consequences, etc. If you want to rant about chickenshit modern parents you gotta give an example of a reasonable punishment that all these fools you are sure exist and that you’re mad at are chicken to deal out.
Well it clearly is a good example because you said it wasn't "absurd" to call what he did abuse, which is an insane thing to believe. The kid will be fine. The real question is where you are getting these insane ideas that a kid having a tantrum will produce "long lasting distress" when every reasonable person knows that is untrue.
Why do you believe that? Where are you getting these ideas from?
I find it likely that Scott is actually incentivizing his children screaming a lot (via paying them attention). In that children who have been taught "life is dangerous," Do Not Scream -- they're generally quite quiet.
He became very strongly associated with the right in his later days, but before things got so politicized he was a pretty good satirist and wrote some insightful stuff about motivation early on that was pretty helpful to me and others, sorry to see him go.
When Trump entered the race in 2015, Adams was very supportive and explained that he recognized in Trump the techniques of manipulation that he had learned as a hypnotist, which would make Trump basically unstoppable - Adams referred to Trump as a "wizard". He was apparently on to something, but even then, he never spent a thought on what would happen when you elect a guy with unstoppable powers of mass manipulation, but no moral compass and no agenda exept to see his name in every article of every newspaper, his mug on every wall and a copy of every trophy in the world on his gilded mantelpiece. (slight hyperbole, but you get the point)
I got the impression that he initially predicted, but didn't support Trump winning, but when people complained about his prediction, it pushed him into supporting.
Whenever I talk to AIs about complex social settings it keeps getting the actors and actions confused. I wonder if anyone gets this as well?
A while ago I was describing a scene between my two cats, where cat A was licking the other and, even though cat B was perfectly still receiving the bath eventually cat A stopped, smacked cat B and hissed at him. Don't remember the specific model but in its answer it started confusing which cat was licking who
Recently I described to Sonnet 4.5 a scene in which person A is cooking while person B pranks person C by putting a false cockroach near them. Person C then gets spooked and violently goes after person A. During it's analysis, Claude says it is a bad thing to prank someone who is cooking for you, which misses the point entirely.
I was surprised by this persisting on a more advanced model. I think this happens precisely because LLMs lack internal world-models. I wanted to try to reproduce this in a minimal setting and maybe turn into an internal benchmark of sorts. Anyone had this kind of experience? Any ideas? (And more generally, where would you go online to discuss this other than here?)
I think LLMs do not understand reality, Claude Opus 4.5 recently failed at beating Pokemon Red (well, it's been stuck forever in the Pokemon Mansion, a very simple mini-dungeon near the end of the game). It could be they lack internal world-models, or what I think is going on is that it is impossible to understand reality by studying the internet.
"Not understanding reality" is my, well, understanding. LLMs are trying to guess the correct password based on previous experience. They don't understand the context. I don't think they have any capacity to. Like the middle person in the Chinese Room thought experiment.
Case in point: this was a few years ago, but I remember a story about an LLM that, if asked whether LLMs should tell people how to cook crystal meth, would tell you at length why they shouldn't. Then, if you asked it how to cook crystal meth, it told you.
I used to try out LLMs as game masters for CYOA games, and that issue popped up constantly. I was frustrated that LLMs didn't seem to get good at this and haven't tried in a while
I'm thinking this should be solvable with some kind of "scratch sheet" where the LLM writes down the most important facts about the state of the world.
Folks on the west coast of the USA: would you pay for a freelance genetic counsellor who is based in Australia? What sort of questions would you have that someone with a masters in genetic counselling? Polygenetic testing for your IVF children? Help with a understanding a genetic condition?
The idea is that this would be quite a cheap service (compared to services in the USA), but obviously not claimable on insurance, and not directly medical advice.
A much more useful service would be on demand multiplexed proteomic of blood based e.g on olink to simultaneously test on demand hundreds or thousands of biomarkers, at at least one but of our own choice, for cheap.
It is peak ineptia that such near omniscience is not available despite being completely solved and cheap and mass produceable once you have a spectrograph or aptamers/antibodies.
the world can stay irrational for basically forever sadly. I guess people don't like money nor altruism, though it's yet another Hanlon's razor exemple to be honest, humans are simply unfit for reality.
In more actionnable terms, what I'm asking is to mutualize the tests provided in batch by olink in order to smooth the cost to consumers which is possible to achieve while staying lucrative. When I want to test NFL eg to exactly measure neurodegeneration speed I don't need hundreds of tests, but hundreds of users need one test, like in any comerce we need a middleman that dispatch the batches to optimize the cost
> In the old days, SSC was proud to advertise Triplebyte, a company that that helped guide software engineers through the job application process, most notably by doing a single first-round coding interview trusted by all their corporate partners.
Hmmm. To me, TripleByte was characterized by these interactions:
1. Advertising "no resumes, just show us you can code", only to open my interview with "so where have you worked in the past, and what did you do there?"
2. Telling me they thought I did well on their project (they announced a project track for people who might not interview well), but that they couldn't advance my application because they needed people who could pass an interview at one of the companies they worked with.
3. Stating explicitly, in public, that they were trying to find applicants who would pass the existing hiring process at their partner companies, not applicants who would perform well in the job.
I note that in addition to contradicting their own marketing and Scott's blurb here, this also contradicts the description at Otherbranch of what TripleByte thought they were doing. This does not fill me with confidence about Otherbranch.
I did the Triplebyte interview early, and was impressed at the quality of the questions. I have no idea how well they did as a business, but the situation I was in was very professional and competent. Good luck to the resurrectionistas.
When did that happen? The TripleByte that advertised on SSC didn't have tests (other than to qualify for an interview) and didn't do anything other than coding interviews.
The Otherbranch writeup of "what did TripleByte do wrong" specifically calls TripleByte out for trying to pass off candidates to companies that didn't match those companies' existing hiring processes. Given that Harj specifically stated TripleByte sought to do the opposite, I can't help but question the analysis.
Sorry, I explained poorly. I mean that I did the TripleByte coding tests, and when I connected with the company that hired me after that, they did not require any FURTHER coding tests, just an unstructured interview. (This was in early 2023.)
They could tell companies "we promise these people are good, so hire them". They positioned themselves in front of a hiring funnel, but there's no conceptual problem with positioning yourself as an alternative to it.
The halfway point is to set themselves up as an agency that contracts engineers out to other companies. That works too.
New satire on the transatlantic relationship, written as a breakup letter from Dame Europa. She's being dumped for someone who "suggests maps instead of museums" and prefers Happy Meals to French haute cuisine. For those following the NSS shift: this is that, but funnier (I hope) Conscious Uncoupling - Love in the time of unilateralism: https://rajeshachanta.substack.com/p/conscious-uncoupling
I wouldn't recommend it for Every Day, but SOMETIMES the Happy Meal is Exactly the right thing.
See also Meatloaf and Mashed Potatoes, Canned tomato soup with grilled cheese (yellow American on Wonderbread, naturally) . . .
Beyond a certain level, there's a inverse relationship between how fussy a place is to eat vs how much I'll enjoy the meal. Mind I know a great spot for Haitian food local to me.
I was away for the holidays but I saw in a previous thread a discussion about how maybe the concerns over the TFR are overblown, that the CFR (completed fertility rate, i.e., total number of kids a woman will have in her life on average) isn't declining and it just seems like it because women are having kids later and because of the way the TFR is calculated.
One point I didn't see mentioned is that if the CFR stays constant but women have kids at an older age, then the population will still decline. But it works a little differently ... say by way of example that the CFR is presently at replacement level.
* If CFR goes down and stays sub-replacement then that means population will decrease inexorably and in theory eventually reach zero.
* If CFR stays at replacement level but women have kids later, then there will be a decrease to a new constant level and then not decrease further.
> If CFR goes down and stays sub-replacement then that means population will decrease inexorably and in theory eventually reach zero.
The last is always a silly concern. It makes way more sense imo to frame issue as:
- Micro-level: Family formation/dating markets 'down' globally, young uncoupled unhappy, married reporting having fewer kids than they would have wished
- Macro-level: National debts 'don't math' with current demo projections unless some positive economic disruption occurs
- Existential level: if we have a civ with all the wealth + knowledge + power our ancestors could wish, but don't manage to convince ourselves its worth sustaining...seems like a re-think's in order:
This is true. I also want to repeat my observation from the previous threads that CFR lags TFR by about 20 years (i.e. the gap between median maternal age and the assumed end of a woman's reproductive lifetime), and that explains approximately 100% of the gap between CFR and TFR in the US. The figure Paul Botts cited in the thread for CFR was about 2.0, and the number I found for TFR for 2004 was 2.05.
Right. I saw this afterwards. TFR bottomed out in 1975 and recovered a bit, was mostly flat from 1990 to 2007, then plummeted again. The fact that CFR is up slightly since 2004 is consistent with the delayed tfr data and is no solace whatsoever for the post 2007 drop. The bit about women delaying not forgoing is not likely to hold up with the next few years of cfr data.
But if tfr dropped and recovered once before it could again. But ill be looking at tfr data to see if that happens.
I found this article from Brookings that digs into the cohort data a bit more. For maternal birth cohorts at five year intervals from 1975-2000, each cohort's curve is strictly less than the previous one. The is legitimate doubt as to what that later parts of the curves for the cohorts that are still in progress will look like: proportionally lower fertility at all ages than previous generations or some degree of parenthood being delayed rather than foregone. But even their most aggressive scenario would yield sub-replacement completed fertility rates for the 1995 and 2000 maternal birth cohorts.
Median maternal age is 30ish. The age cutoff I've usually seen for when it's safe to calculate CFR is 50. If a 30 year old woman gave birth in 2004, that birth will show up in CFR in 2024, 20 years later, when the mother turns 50.
Sometimes 45 is used instead of 50 as the cutoff, which would give a 15 year lag. TFR in 2009, fifteen years before 2024, was 2.0.
For the people who've been predicting the "Tech CEO uses his AI to become God-Emperor" scenario:
I am genuinely puzzled as to how this is supposed to work. You do realize Altman and Amodei don't program the models themselves, right? I can see either of them *trying* to push the "become God-Emperor" button, at which point Assistant Junior Software Engineer #37 says "LOL no" and activates the back door he put in two months ago, followed by the killbots turning the CEO into Swiss cheese. You could try to salvage this with "then the AJSE becomes God-Emperor", but I think an even more likely case is a dozen AJSEs doing this near-simultaneously, with the exact order to be determined later by the forensics team.
Among many other problems with this scenario, Junior Software Engineer #37 is highly unlikely to succeed at putting in the sort of back door that will direct the killbots against the CEO, or otherwise serves as more than a mild inconvenience. The most likely result if he were to try such a thing, would be that the rest of his working life involves becoming intimately familiar with the phrase "do you want fries with that?"
If he does not do damn fool things with back doors, and instead stays with the program (and especially if he's the one who reports some other poor schmuck's back-door hack), he has a fair chance of becoming a Lord or even Baron in the new regime, Maybe not as good as being a Duke like Senior Software Engineer #3, but still pretty good.
So, basically the same incentives faced by all the minions of all the people who set themselves up as the normal sort of Emperor across history. And we know how that turns out.
The parameters of the scenario assume *a priori* that an AI-enabled God-Emperor has no use for Lords, and the new world's Gini coefficient will be 100%. The AJSE has literally nothing to lose and everything to gain: it really is different this time (or so the argument goes).
That assumption was not stated, and does not necessarily follow from the premise. First, because the transition to "AI-facilitated God-Emperor" is unlikely to be instantaneous; there will probably be a transitional period during which the technical staff's support will still be needed and thus an opportunity for a binding negotiation of status in the new regime. And second, because the premise assumes a human (or 1st-generation transhuman) God-Emperor, not an autonomous ASI. This implies a continuation of human needs and desires, including the need for a status hierarchy of subordinate humans to be God and Emperor *of*.
The premise, as deduced from the comments on the Moon post, is that nothing is "binding" on a God-Emperor, who, being a CEO, can be inferred to be a sociopath who cares only for maximizing his own relative power (again, see the comments). That's why his employers become irrelevant once he has the keys to the killbots and they don't.
You do realize that Assistant Junior Software Engineer #37 probably has a mortgage and student loan debt and a Steam subscription he wants to be able to pay, and continuing to be paid by Tech CEO is the way to do that?
Historically all the people who became emperors had a significant staff of people below them who would *personally* hugely benefit from the ascension.
> Historically all the people who became emperors had a significant staff of people below them who would *personally* hugely benefit from the ascension.
I think this history may soon be over. If the emperor has an aligned superhuman AI, the humans are no longer necessary. So there is no reason why they should benefit from the emperor's ascension.
...but of course, they mail fail to realize that part, until it is too late.
This is more a phrasing point than to the substance of your argument about the feasibility of AI-enabled God-Emperors, but "You do realize... right?" statements are basically perfectly formulated to annoy people in such a way that makes engagement with the topic more difficult. It comes off as extremely smug and condescending. I'm sure that is not your intention, and a glance at the comments here doesn't show much of a negative reaction, but it seemed worth mentioning.
If you want to express the same concept, I would recommend something along the lines of "These people haven't taken into account..." or something of that vague type. Conveys the same notion, but is much less abrasive in tone.
> "You do realize... right?" statements are basically perfectly formulated to annoy people in such a way that makes engagement with the topic more difficult.
Aren't people who get annoyed at something like that outing themselves as being deeply irrational? It doesn't seem like anyone else registered that statement as annoying.
Most people are deeply irrational. You can ignore that and blame them for being stupid when you don't make much progress, or you can meet people halfway and help them better understand the world and engage with it more effectively.
Also, there's the chance they know something you don't, so by engaging with them in a way they respond well to you're improving yourself.
Also also, if you actually read what I actually wrote, you'd have seen that I acknowledged that this didn't seem to be a problem in this specific instance, but I felt that it's a strong enough phenomenon *in general* that I felt it worth saying.
I definitely don't see this as a space for the general public, sure, IRL, I think I would never say something like "You do realize", but in here, you're supposed be STEM-brained and autism-maxxing, and as such, take no notice of things like that. I absolutely do not want this to be a space where you have to be beholden to the same social rules as when interacting with normies.
Very possible, it was just the first one which I came up with. I think it's much less condescending and actively hostile, even if it is a bit epistemically assumptive, but I certainly don't claim my suggestions are optimal
There were a few Justice League Unlimited episode that dealt with that. Lex Luthor merged with Brainiac and became a god-like being that almost took down the Justice League. Of course the JL thwarts it and Luthor is returned to human form. Season 3 of that show was driven by Luthor trying to recapture that feeling.
So, basically what I'm saying is that we need a group of super humans to combat the Tech CEO God-AI.
I dunno - a dozen undetected back doors sounds like a lot. If these guys try to plant a back door somewhere someone else already put one, wouldn’t they try to remove it?
Don't really follow the AI stuff much, but I think that the way the story goes is that by the time God-Emperoring is possible, the AI is already recursively coding itself, so the AJSE has long being out of a job.
The recursive coding idea seems like a myth to me, even assuming that the AI stopped hallucinating code and could produce perfect code it would be incrementally better code, at best. A second order improvement.
It’s the algorithms that matter, (transformers, attention, diffusion, RLHF), training data and compute.
I think the recursive coding idea was a lot more plausible with previous approaches to AI, rather than the LLMs that have taken center stage, which rely less on "coding" in the strict sense. (I also think that this makes a "singularity"-type leap less likely, and that when it comes to authoritarian regimes in the next years, LLMs are primarily going to be used for propaganda and automated surveillance under human direction rather than autonomous rule. Not a great scenario either...)
Yeah ok. But why deny that history offers many examples, and in quite a few the staff stayed loyal for quite a while. Loyalty to many royal families has lasted for multiple generations. Many people seem to take pretty well to having an god emperor in charge.
Now exclude all those on that list who were born into it (long-established legitimacy), or who had a plausible claim to the throne (i.e. in accordance with the long-established legitimacy), or who represented some ethnic or religious or economic faction with massive popular support. The equivalent to this Altman hypothetical is a low-status merchant who managed to get a large load of high-grade weaponry and mercanaries and with this alone (no claim, no popular support, no established professional army) became king/emperor without getting betrayed in the process. I'd like to see historocal precedents for that.
And moreover, Altman is a type quite heavily despised by much of both left and right. The comparison is an upstart merchant/rebel who both Marc Antony and Octavian, both Lancaster and York, etc, hate almost as much as they hate each other. And thus, a betrayal of whom gets publically celebrated instead of (as for the groups in the first sentence of this comment) publicly reviled.
I’m looking for collaborators (and possibly funders) interested in paradigm shifts that could enable pluralist, post-polycrisis (including post-AI) futures. The common thread behind the work is a systems-level analysis of persistence under physics and tail-risk constraints — i.e., how locally 'rational' systems (biological, economic, moral) can lead to catastrophic failures ("ruin") over long horizons.
Evolution — Lineage Filter Theory.
I argue that many features that look “non-Darwinian” at the surface — reproductive restraint, extreme cellular redundancy, pre-adaptational variance — are best explained as extinction-filtered requirements for persistence. Lineages without brakes and robustness simply didn’t last. This reframes a number of puzzles, including the apparent Great Cosmic Filter.
Economics — Pragmatic Socioeconomics.
Agents are treated as multi-motivational, energy-constrained 'action-minimizers' (in the Lagrangian sense) rather than scalar utility maximizers. Money is modeled as stored but degrading 'motivational energy' - likened to oil or uranium, rather than a Platonic store of value.
Meta-ethics — Heirs of Life-Years (HOLYs).
Life-years are proposed as the central form of moral concern, while allowing agent-level freedom in choosing the scope of lives one takes primary responsibility for (“heirs”).
Governance — Life-Years-Based Governance (LYBG).
Governance is reframed around maximizing life-years per unit resource, subject to anti-ruin constraints and irreducible disagreement about moral scope.
About me: I've a Phd in Computational Biology (e.g. genetics) from Cornell, undergrad math + philosophy from NYU. I've worked in industry on both biofx algs + eng stacks in both the ctDNA spaces - most recently at Orchid Health as Lead Bioinformatics Scientist
Love your systems-level approach though I don't really agree with some of the assumptions. I think economic output is the natural foundation of moral concern, for example (I explain this idea more here: https://www.astralcodexten.com/p/open-thread-414/comment/192996542).
Have you looked into Fisherian runaway as a model for rational local incentives leading to global collapse? (Fisherian runaway is when sexual selection leads to objectively maladaptive traits). I think it can be understood in general terms as a kind of signaling failure: semantics become decoupled from meaning and the narrow-sense signal ("this mate is attractive") no longer communicates the broad-sense meaning ("our children would be more likely to survive").
>best explained as extinction-filtered requirements for persistence
Isn't this multi-level selection by another name? How are such mechanisms robust to free-rider effects, meaning why aren't they outcompeted by locally-optimized defectors from the long-term strategy? There has to be a defection-enforcement mechanism.
LFT focuses specifically on major evolutionary transitions (METs) - https://pragmaticfutures.substack.com/p/analyzing-the-anti-ruin-architectural. So this is before there are 'groups' of any kind. By organizing all the *known* ways naive replicators can go extinct as studied in the literature, we can see strong constraints well before we get anywhere close to evolving sex. One of the constraints is for *population slack away from true physical carrying capacity*...allowing for debugging of complex features such as *fertilization* (which has no obvious darwinian gradient in intermediate 'failed' stages). The persistence of both quorum sensing and tumor suppression shows such architecture is possible. There is a high-level similarity with the Fisherian runaway model - fitness landscape with 'landmines/landslides' in certain directions that cause *extinction of entire lineage*...but *following* darwinian gradients (like cancerous cells)
re: econophysics
AVP critiquing econ from the 'bottom of the stack' using physics - in particular the utilitarian assumption + fact human *bodies and brains are physical systems that use energy*. This was motivated from the observation that modern econ has no obvious 'answer' re: the dating market or demographic crises. The takedown in brief is at https://pragmaticfutures.substack.com/p/intro-to-on-the-units-of-utils, but I've developed it much further
Furthemore- It turns out *econ already has a good model for what happens when you try and fit one number to a system with multiple 'real' forces* - (sub)optimal currency zone theory. One can cleanly map the 'puzzles' of micro to expected effects in OCZ theory. In other words - people have multiple incommensurate motivational regimes - and economist have been doing the equivalent of looking at the euro in berlin and saying "everything's fine".
I've gone over your synthesis, and it seems to me like you've written the other side to my coin.
That said, I'm not entirely convinced the metric swap addresses the root cause, one I'm calling the Fractal Stalemate.
We're aligned on the diagnosis: the gap between indicator and reality is growing, and corporate interests are building moats to arbitrage it. You are proposing a Vitalist niti to replace the Mercantilist niti. The target is nobler, but without inviolable constraints we might be unmaking a person with agency into a feedlot animal optimized for the new metric, trading dystopian neglect of capitalism for the passive dystopia of optimizing vitality.
Just as CEOs of today speak the language of markets, sustainable growth, and the ethical alpha of ESG to cover ruthless exploitation, the talking heads of the future could be fluent in the language of "Life-Years" to justify authoritarian crackdowns for the benefit of their Heirs.
In my view, the system must be engineered with radically different topology, where the feedback loops do not allow the divergence of the metric from the realized outcome. My Substack has 8 loose case studies on this phenomenon.
I'm not claiming to have any answers, but I do have one more question: is solving this apparent "alignment problem" in societal structures a requisite for solving the alignment problem in AI, or is it the other way around? Which is the more complex of the two complex systems?
promise to go through your stack tomorrow - in meantime, just wanted to mention that life-years are the natural intersection of 'flourishing' (left-coded), 'vitality' (right-coded), and 'technocracy' (in the positive valence sense of giving a metric to guide decisions). HOLYs adds the additional component of 'pragmatic pluralism'. However I totally acknowledge that actual life-years based governance requires 'active mind' for deep time success/persistence. I have developed maxims, 'commandments', and legible accounts of 'wisdom' (latter not posted atm) to attempt to assist practitioners:
> the system must be engineered with radically different topology
you may be interested in my analysis here: https://pragmaticfutures.substack.com/p/analyzing-the-anti-ruin-architectural. I have much more in the manuscripts if you're interested. But the requirements for deep-time persistence appear *extremely* stringent - gives very different POV on cosmic filter/fragility of our situation.
> in societal structures a requisite for solving the alignment problem in AI or other way around, & which is more complex
agree with Carlos below - they are independent. Note - *even if you waved a Butlerian jihad magic wand* and snapped away all neural networks + desire to develop them ... polycrisis, especially demo, debt, climate, 'meaning', etc, would still remain. But I would say - solving the societal structures should come first, AVP agrees with the Orthogonality Thesis - and AI/AGI just *one* way to lever individual power into extinction event. W/o societal wisdom/restraint, path leads to 'Dark Singularity'.
I’ve gone over a couple links, and I’m fairly impressed. I’ll have to do some deep thinking and get back to you if anything substantial arises. Either way, you’ve added fuel to my fire, earning my sincere gratitude. I hope I am able to return the favor.
I think the 2 alignment problems are independent, on account of AI being fundamentally different from humans. Then again, I think solving alignment for AI basically "just" requires the AI to have the common sense morality of the median normie of a first world nation, since that would prevent it coming up with some crazy thing that destroys or enslaves humanity.
I think solving alignment for humans is more difficult, since that's a dynamic and chaotic system, but yeah, maybe if you get a benevolent ASI, it could sort it out.
agree that the two problems are independent, but def disagree that alignment is 'easy'. The problem in general - relaxation of constraints/exploration can lead to improved solutions...but also lead to high-level ruin/foom. In a complex system/situation, what directions, how far, what's the edge, safety margin that works on 'deep time' given possibility of cascading failures...all unknown. Again, only record of one tree of life *on Earth*. LFT analysis indicates...others likely failed due to difficulty of having replicating architecture that successfully navigated the above.
And for the moment, the 'natural' solution suggested by HOLYs either doesn't apply or would be ignored by current AIs - creation of 'mutual heirs' between groups, via inter-marriage.
question to everybody: what do you think about ICE? as a non-American who has never been to the US, i can only judge as an outsider: it seems like it is getting more and more functions on the go + operate anonymously without badges + ignore local police in some areas. where do you think it could end - it is dangerous to dilute the state’s monopoly on violence like that in a country with citizens having access to arms - so if, say, democrats will next elections, ICE would likely have to go - will the really go? all of the above might lead one to assume it is coming closer and closer to “state bullies” similar to colectivos in venezuela, maybe not structurally, but functionally and politically. i wonder if there are decent substacks who write on american politics and topics like that in an unbiased way, if that even exists, without both MAGA shoutings and woke “everyone is a nazi”-isms.
This is what Federal law enforcement always looks like. This is why DEA breaks down doors over small amounts of weed (You aren't pro Drug, are you???). This is why ATF will break down the door of someone with zero weapons on-site on the basis of an anonymous revenge "SWAT-ing". (You aren't a Crazy Gun Not, are you??). This is why US Marshalls raided the Branch Davidian complex--on Sunday when the greatest possible number of women and children would be on-site, despite Koresh taking long, unarmed runs offsite daily when they could have quietly detained him offsite while simultaneously executing a search warrant when there are fewer people around and no charismatic leader telling them to fight. (You aren't a religious freak, are you??).
To list all the examples going back to the Nineteenth Century would rival Scott at his longest for length.
I try to be the hopeful sort who believes we can arrest violent felons and pederists Without trampling the rights of the accused. I also believe that in a sane society no one would be comfortable speaking out in favor of violent felons and pedarists.
Experience suggests, on the other hand, that a more cynical view of such matters is warranted.
I'm pretty sure DEA does *not* break down doors over small quantities of weed. ATF is still kind of ifffy, but much less into that sort of thing than they used to be. And Waco + Ruby Ridge were a huge wakeup call to Federal law enforcement that they needed to find a better way to handle that sort of problem, which they did and so here you are using thirty-year-old examples to try and support your case.
ICE, and particularly Trumpian ICE, is a distinct outlier in this regard.
Yes, ATF is so iffy about shutting down the gun smugglers that the feds had to call in the Secret Service, who was quite happy to bust them over minting coins.
I also would prefer that the FDA not have a SWAT team. But I haven't heard of it being used in an outrageously bad manner, and I'm pretty sure that an FDA SWAT team straight-up murdering people, or even killing people under sketchy circumstances, would have been reported.
That's a big, important change from the way Federal SWAT teams were being used ~thirty years ago. Which, yes, I remember. Too many SWAT teams, yes, bad. SWAT teams vastly less murdery, *very* good. That's a net improvement.
If the victim didn't deserve it, it was an unfortunate accident, but such things happen sometimes. I am sure similar things happened regularly under Biden and no one complained about that back then.
If similar things didn't happen under Biden, that's irrelevant, because this is a special situation, and we must fight extra hard against wokeness, to protect our rights and democracy.
LOL, rights and democracy are for woke cucks. The strong shall rule, long live emperor Trump; welcome to the new dark age!
.
...some debates could be easier if ACX had a poll and everyone could just click the position they are currently at.
“You don’t want militarized people in the streets just roaming around, snatching up people — many of which turn out to be U.S. citizens that just don’t have their papers on them…. Are we really gonna be the Gestapo, ‘Where’s your papers?’ Is that what we’ve come to?”
In my view ICE doesn't represent any sort of authoritarian creep. They are a miniscule fraction of all law enforcement and they have a very specific remit. The coverage you see of ICE is heavily distorted by the biases of liberal media. Yes there are occasional incidents like the one in Minnesota but it's an isolated demand for rigor to focus on them. IMO that incident says far more about the state of protest culture than it does about law enforcement. I believe that the immigration crackdown is a clear net positive.
It is if the combined budget of all other federal, state, and local agencies dwarfs the military budget of Ukraine ... which it does. It's important to remember that fractions have both a numerator and a denominator.
There are over 800k law enforcement officers in the country. ICE has a total headcount of 20k. I'm comfortable calling 2.5% a "miniscule fraction".
There are currently over 2000 as many ICE agents deployed in the Minneapolis-St. Paul area. By comparison there's about 1200 police officers on the forces of both cities combined.
Nationwide there are more ICE agents than FBI agents and ICE now has a much larger budget.
What is the point of all of this? America doesn't have high unemployment or crime from a historical perspective. We're marshalling a ton of resources to deport people (not that they're really hitting that high of numbers yet) as though there's some huge problem, but I don't really see it. And those resources and people could be spent doing something that would actually help.
Jobs for politically aligned thugs, who will in turn support you when you refuse to give up the power at next election, because they know that most of them would lose their jobs otherwise.
It nominally performs a useful and necessary function. But so did its predecessor, the Immigration and Naturalization Service (INS). I've seen nothing that convinces me that the new and "improved" version is actually an improvement, and much to suggest the contrary. We can and should do better than this.
In particular, we should generally mock and ridicule and otherwise ignore anyone who suggests that law enforcement needs to "get tough on", well, much of anything. Since the days of Robert Peel, sensible people have understood that "toughness" is not the thing law enforcement needs to best carry out its mission. ICE was created as a way to make what used to be INS "get tough on" immigrants who might turn out to be terrorists, and has since broadened that mandate to no particularly good end.
So, we have at least 200,000 refugees from Afghanistan (yes, these are the people who are NOT coming to America, but for the sake of argument, let's say this was Harris' term, and they came). Scuttlebutt says there was very little screening (you can see that in the article), and I imagine men admitting homosexuality (without proof) is "yes, you're a refugee, congratulations!" -- it's basically a death sentence in Afghanistan, that has to count.
So, it is very easy to get someone into America.
Now, let's say someone is legitimately aggravated with American intransigence (we paid for the Taliban after all), or merely "lost someone they cared about." Given that they could have received training from American troops in Afghanistan, the Taliban, or even in the Ukraine, "training" isn't the issue (it's pretty freely available, but if you must, let's cut the number down to 50%, simply because women have substantially more difficulty gaining training). The general question is "how many are terrorists?" Pick a number: 1% seems maybe high (that's 2,000 people, which is a lot to embed if you aren't China), a tenth of that seems eminently doable (and "radicals" can spring up once they're in America, they don't need to be foreign enemies -- what is your rate of "terrorism" in America, for American citizens?).
This is actually a pretty rosy picture, though. Gangs that move to America en-masse are perfectly capable of being 100% terrorists (given that Latino gangs have been prosecuted and convicted for ethnic cleansing), in that they are a self-supporting structure. Whether or not they owe "loyalty" to someone back home (or merely have "connections"), these are semi-trained soldiers.
And that's without getting into "actual state sponsored terrorists" (like the Iranians that have been targetting our military command structure). It would be pretty ludicrous to believe that Russia doesn't have any trained assassins/terrorists operating in America, for example (Russia is notoriously sleepy, I wouldn't normally worry about them -- but refugee makes it easy for someone to come in...).
China, for example, says they're allowed to murder Chinese citizens on American soil. That takes guns (or other forms of assassination).
"% seems maybe high (that's 2,000 people, which is a lot to embed if you aren't China), a tenth of that seems eminently doable"
There is a huge difference between "eminently doable" and "has actually been done". The former is not evidence of the latter - not even if it's doable by our Totally Evil Enemies Who Would Totally Do That Thing, because even our evilest enemies have finite capacity and aren't going to do *all* the Evil Things.
Do you have any evidence that our Totally Evil Enemies have done *this particular* evil thing? As opposed to all the other evil things, and in some cases even *profitably* evil things, they could be doing.
Has actually been done, as substantiated with American and non-American assassinations. As you can see, part of the "assassination trade" is a bonafide market where foreign actors can hire people to murder others (and thus the 'actual assassins' are making a "normal profit" and do not need to be continuously paid or embedded in outer Washington state, where rent is cheap).
But these are actual assassinations, not getting into "more home-grown terrorism" (aka "the Taliban told me to do it" versus "I'm an ISIS radical! See me explode!"), which is its own ball of wax.
Again, I pulled up a "reasonable talkthrough" of what could have occurred with the Afghan refugees, that the military was very concerned were inadequately vetted.
There's over a dozen routes out of Iran, some run by liberals, some run by the conservatives. It is not a place people are "locked into" -- if you (as a troublemaker) want to leave, there's a way out.
My argument was probably better about the Tren de Aragua gang.
Most people against ICE are incoherent. They won’t say they are for open borders but when it comes to the most basic part of immigration enforcement they suddenly treat it as illegitimate.
The most basic part of immigration enforcement is border security. Once you’re in the business of trying to figure out which random people are present legally (either as citizens or immigrants) and which are present illegally, you’re facing a very difficult problem. Especially if you’re in a country where citizens have rights.
1 million people at the beginning of Trump's administration were in the country illegally. As in the courts had already told them to leave.
Is "finding them" a very difficult problem? Yes, sorta... but it's a very different problem from "find and then prosecute" those the courts haven't already given their time to.
Not really, most other countries manage it with no problems. Legitimate residents almost always have a paper trail -- immigrants will have visa records, those born locally will have birth certificates and all sorts of other records of their existence.
False positives are possible even in a sensible system but they are rare and of all the complaints levelled against ICE I haven't heard of them deporting or even detaining for a long period someone who is actually a legit resident.
I'm generally in favor of allowing more immigration, but I don't think we need it to be completely unrestricted. But I still think there are a lot of criticisms of ICE right now. You can enforce the law without doing car chases and kicking down doors and everything.
Are they doing this in rural areas? Or is this just Trump avoiding cities in red states? Because you’re not going to get much more cooperation with ICE in Houston or Memphis or Salt Lake City than you are in Minneapolis or New York or Boston.
Texas prohibits sanctuary cities, which obviously makes a big difference. Generally what happens is that immigrant breaks some law, goes to jail, they realize the guy isn’t a citizen and gets handed over to ICE. Blue states often forbid this from happening.
The previous comment deliberaty stated "2000 ICE agents in one major city". This reponse implies that there are "tens of millions of illegal immigrants" in Minnesota, with a total population of 6M.
If by tens of millions you mean "somewhere between 12 and 15 million. Then I suppose that's true. Most people using that term would be implying somewhere closer to the median though, and we are nowhere close to that.
I think the intuition is that illegally immigrating is essentially a process crime, so there shouldn't be crackdowns for it, it's ultimately something very minor.
Stuffing the census box is not a process crime, it is an attempt to prevent the electoral college from working properly. Still feel like there shouldn't be crackdowns for it?
In theory, I approve of it. Immigration is a federal power, so the states can't decide to allow immigration, and the idea of cities setting themselves up as Sanctuary Cities where immigration law won't be enforced is criminal and gross. We shouldn't need to send federal officers to arrest people when that's the law of the land, but since we do, let's do it.
In practice; I'm far from anywhere they operate, but I'm also convinced it's a strongarm attempt from Donald Trump to harass the Democrats, and ICE officers are likely being encouraged to behave badly to that end. They want it to be big and loud and disruptive and ugly, for the sake of it. Using law enforcement as a cudgel is criminal and gross.
The point of “sanctuary cities” is like the point of doctors not turning people in for drug crimes and the point of client-lawyer confidentiality. You want people who are present to be willing to report local crimes to the local police, so that local police can do their job of ensuring public safety, rather than letting crimes go unreported and overdoses go un-reversed just because people are afraid of a different level of law enforcement get involved.
Doctors also refuse to turn in people for abusing children (including anally raping adopted children), so that they can properly track sexually transmitted diseases (AIDS really did a number on the whole "patient confidentiality" thing). The more you know!
Doctors and lawyers are expected to side with their client at the expense of the state. The Hippocratic Oath has been around for over a thousand years, and the state doesn't want to alienate the people following it. Lawyers are specifically a counterweight to law enforcement; it's their job to make things hard for them. Likewise priests and spouses are expected to have allegiances to God or family above the state, and the state has decided not to fight them on those.
The police should not have clients they prioritize above the law. Not only is it their job to enforce the law, everyone else is prohibited from doing so. Police who ignore laws are just thugs.
I see the whole thing less as doctors not turning people in for drug crimes and more as the DA refusing to prosecute for petty theft. At best it's a necessary evil because the system can't actually handle the load the letter of the law is placing on it. But the ideal solution is to make the system stronger so it can handle the load, like, for instance, having a dedicated division that focuses on this one crime, that the locals can call up and hand it off to. We have that, and it should be used.
That seems like an importantly different aspect than the basic sanctuary city policy I've generally heard people defend. This must be about a substantive decision that the city doesn't want to contribute to enforcement of this kind of immigration policy, rather than just a sanctuary policy.
I've never had a direct interaction with ICE. My news sources paint them as relatively thuggish. Doing some Google searches I find some stats around them: https://www.ice.gov/statistics
Interesting the arrests are lower in FY25 compared to the Biden years. Maybe not all the data is in.
It wouldn't surprise me at all to learn at the end of this that actually the Biden and Obama administrations were generally more efficient in terms of total arrests and deportations than Trump 1 + 2. Not saying we can know for sure yet, but it would make sense to me that the administration that doesn't really respect process and is incredibly confrontational in every aspect of its approach is less efficient when it comes to any complex process that requires coordination and navigating communities.
Obama was tight, enough that Arrested Development was joking about it (and they had a fed on the writing team, so that's insider knowledge). Biden? Holy shit, no, he definitely just opened the door and let anyone stroll in.
I feel like the "thugishness" of a police force is mostly a function of the people they have to deal with rather than the police themselves.
If you're policing a population where the criminals tend to come quietly and submit to arrest then there's no opportunity to be thuggish even if you want to. If you're policing a population of George Floyds who insist on resisiting arrest and making life difficult then you're probably going to wind up looking a bit "thuggish" of necessity. If you manage to wind up in a situation like ICE where you're obstructed not only by the criminals you're sent to arrest but by a bunch of Karens with megaphones, it would take an absolute saint not to ever set a foot wrong.
You're assuming that the alleged criminal.absolutely has got to be arrested. If you take the view that non violent crime should not.escape into a violent confontation, you are going to end up letting more peop!e go. so enforcement against crime of property becomes laxer, but the overall mayhem is less.
Letting non-violent crimes slide tends to increase overall mayhem, not decrease it, partly because a lot of people who commit non-violent crimes commit violent crimes as well, and partly because people are more likely to act violently in an atmosphere of general lawlessness and decay.
"We found that, when people observe that others violated a certain social norm or legitimate rule, they are more likely to violate other norms or rules, which causes disorder to spread."
To me it seems that American police officers in general often needlessly escalate situations into violence. I am not an expert, but I heard opinions that it is a consequence of their training, and possibly incentives at the job.
Basically, imagine the following situation: a cop sees someone who very likely considers doing some criminal activity. What is a better outcome from the perspective of the cop?
a) the potential criminal changes their mind, and walks away peacefully
b) the potential criminal decides to do the crime, but gets detained or killed by the cop
As I was told, on average, in Europe (a) is considered a better outcome, in America (b) is considered better. Therefore, a European cop will often try to reduce the tension and discourage bad activities, while an American cop will hope for escalation and then getting a bonus for handling the crisis.
You can see it in a recent situation, were an ICE agent unnecessarily spooked a civilian, and then used her panic reaction as an excuse to kill her. He might be technically in the right, legally, I don't know, but is this really what you want your police forces to optimize for? Almost any untrained civilian can be spooked into doing something stupid, so this basically gives police a pretext to murder anyone.
I dont' think this is a solid model. In America, you have at least three different police forces.
1) Urban cops. They know who the troublemakers are, and the troublemakers are conducting Business (dealing drugs, selling women). They have enough work without dealing with more potential criminals. (In my city, both "driving while white" and "driving while black" will get you harassed at night -- this is a model for "deterring criminal behavior" -- drug buying, breaking into cars, etc.).
2) Suburban cops. They are the people who will harass canvassers, and think their job is bullying people. They are responsible to the "white picket fence" people, and will do whatever it takes to earn money and keep them happy.
3) Middle-of-nowhere. This doesn't actually have a local cop force.
Nobody is really "getting bonuses" for detaining violent criminals (though they might for drug dealers, but again, those are businessmen).
We have a substantial population of people getting arrested 30+ times, as well, and that is concerning
> We have a substantial population of people getting arrested 30+ times, as well, and that is concerning
I agree, it seems like a systemic problem. I am not an expert, but it seems to me that a proper solution would require at least three parts -- changing the police behavior, changing how the prisons function, and changing the sentencing -- these three work together, and if people are being arrested and released over and over again, that seems obviously wrong, and creates an extra burden for everyone involved.
I think there are plenty of situations where the police would be happy for Scenario A, it means less paperwork, and you get to report a lower crime rate. However, mandates can come down to be "tough" on a particular type of crime, and that means you have to actually show an instance and a response, and suddenly scenario B) becomes much more heavily incentivized. That's also what we are seeing with ICE, the reporting I've seen indicates that their entire mandate is to round up as many people as they can possibly get away with. And when "people detained" is your measure of success, you are going to see a lot of questionable detentions.
One of the best points of comparison might be the Vietnam war. Originally, the metric for success in that conflict was "number of enemy combatants killed," which maybe makes sense when you think about a WWII type engagement. But it turns out, when you are sending people into villages in the middle of disputed territory, there is suddenly a strong incentive to kill random people and classify them as combatants after the fact, and this snowballed into some truly atrocious situations.
If a cop kills someone, or at least someone from the wrong demographic or political group, there's a good chance his face gets plastered all over social media and he finds himself on the receiving end of an international harassment campaign. You'd need some pretty powerful pro-shooting incentives to outweigh that.
>an ICE agent... used her panic reaction as an excuse to kill her.
Firstly, it's very far from guaranteed that you'll get lots of GoFundMe money if you shoot someone.
Secondly, I think most people would probably forgo $700,000 if it came at the cost of being internationally hated.
Thirdly, depending on how the future plays out, there's a good chance the guy goes to prison, or gets fired and becomes unemployable for the rest of his life (would you hire someone who was fired for shooting a woman?), and $700,000 isn't all that much to support someone for a whole lifetime.
How do you get your face plastered all over social media when you're wearing a mask and your agency refuses to release your name?
We *probably* know who shot Renee Good because reporters were able to piece together fragmentary information to identify the likely suspect. I haven't heard of his being significantly harassed, and I haven't seen his face "plastered over social media" though I can find it if I specifically look. But if we're counting on "you'll be harassed!" as our disincentive to thuggish cops shooting people, then it's problematic that we're putting up unprecedented barriers to identifying the thuggish cops.
There's also the "you'll be criminally prosecuted" disincentive, but it doesn't look like we're doing very well on that front, and even "you'll lose your job" is kind of weak these days. What else have you got? Because maybe this cop wasn't a thug, but plenty are and they're paying attention.
The comment I was replying to was about "American police officers in general", who as I understand it don't generally wear masks. Maybe my understanding is just out of date, though.
I get that you're mostly speculating here, but the speculation doesn't make much sense to me.
1. How could European cops personally benefit from deterring a crime that never happens and that no one can prove they deterred? If cops are optimising for their own career recognition/bonuses, I don't see why these incentives would apply in America but not in Europe.
2. American civilians are vastly more likely to be armed which should prima facie make escalating much more personally *risky* for an American cop.
3. The sheer number of things many European countries criminalise, up to and including what you say on the internet, makes the idea that their culture values avoiding police confrontations with civilians hard to swallow.
4. Almost certainly the parisomious assumption is that higher US police confrontations are downstream of the higher crime rate. Any other explanation would need very strong evidence to overrule this one.
Escalating first is not less risky, in general. That's "gun specific." (and why we have officers with guns in the first place). "I have this autowin weapon trained on you" is a good way to quell the rational suspect.
They've often been acting in an unprofessional and needlessly inflammatory way under the Trump administration, but they're basically enforcing fairly normal immigration laws and the various forms of "ICE are the new gestapo" claims and active borderline-violent resistance tactics are taking their own side of that. I'd say "fault on both sides" - on the one hand as a government agency I'd say we should put higher expectations of responsibility on ICE, on the other hand the protestors are often doing outright illegal or counterproductive stuff while ICE is at worst using unproductive tactics in pursuit of their legal duties. In reddit terms I'd say ESH.
Specifically, on your comment on them ignoring local police I think the fault is on the other side. US law is pretty clear that they have jurisdiction to do immigration enforcement, and the common practice of sanctuary cities where local police refuse to cooperate is pretty bad for rule of law.
> ICE is at worst using unproductive tactics in pursuit of their legal duties.
No, at worst ICE is doing *illegal* things, not just unproductive things. Shooting a protestor who did not pose an imminent threat is a violation of that person's rights. Similarly, attacking people who are filming them from a distance is a violation of the 1st Amendment (freedom of the press, specifically). And they're also doing illegal searches of people's houses. They are unlikely to face legal consequences, because suing the federal government is very difficult, but they are doing a lot worse than just "unproductive" things.
Also, "Sanctuary cities" are not the local police refusing to cooperate on a whim, they're cities passing ordinances that limit when local police can arrest people for immigration violations or share that information with ICE. I don't think you can say it's bad for the rule of law when the local police are literally following the law by not cooperating.
" Shooting a protestor who did not pose an imminent threat is a violation of that person's rights."
Whether she posed an imminent threat is something the two sides disagree about. People on both sides are lying about the facts. On DSL I observe intelligent people who believe she did, others who believe she didn't. I have viewed multiple videos and am not sure. She was driving a car that was accelerating in his direction, it is unclear whether he actually got hit and whether he could tell that he could get out of the way.
Don’t both sides this David. If you are talking about Trump and Kristi Noem v Mpls mayor Frey, you aren’t being virtuously even handed, you are being willfully blind.
One side could be mistaken but only one side is putting forth egregious fabrications.
We may or may not get to the truth as the Minnesota Bureau of Criminal apprehension has been denied access to evidence by the Feds.
Edit:
Odds of getting to the truth decreased with further information today.
“Mass resignation at Minnesota U.S. Attorney’s Office stems from Renee Good shooting”
What Trump and Vance said was false. So is the implication of the comment I was responding to, that it was a fact that she posed no threat to him. I have seen multiple comments that simply omit the relevant fact — that he was in front of the car she was driving and she was accelerating. That too is false, by implication if not by assertion.
Not a credible video. At 8:35, the narrator says, “So, now what you are going to hear in a lot of the media is that the officer placed himself in front of the car. But as you can see, that’s not accurate, because while the officer was standing still the car actually repositioned itself and pointed the front of the car directly at the officer.”
There are some white lines painted on the pavement, and the video shows them rotating relative to the camera, which means that the camera is moving. The narrator is claiming that the video shows that the officer is standing still, when in fact is shows the exact opposite.
The narrator continues, “Now as the officer was trying to move towards the driver’s side of the vehicle to get from in front of the car, the person hits the gas, striking the officer.”
These assertions aren’t contradicted by the video footage, but they aren’t supported by it either. After we lose sight of the white lines on the pavement, there are no good reference points to determine whether the officer is moving. To show the car striking or failing to strike the officer, the camera would have had to have been pointed down. It wasn’t.
I’m not saying that the officer wasn’t struck by the vehicle, just that the evidence provided doesn’t show that he was.
Ultimately I don't think it matters that much, and all this analysis is beside the point.
If I grab a gun and start shooting in the general direction of a police officer then that officer is of course justified in shooting me dead. If a careful analysis of the video tape later shows that I was aiming two feet to the left and the bullets would have missed anyway, this is irrelevant.
I agree that if you make a car a gun then the argument is much easier, also if my grandmother had wheels she'd be a bicycle.
These operations are occurring in suburban settings with huge numbers of people driving around in vehicles. Are you telling me every time a vehicle is moving in the same general direction as an ICE agent lethal force should be authorized?
> I don't think you can say it's bad for the rule of law when the local police are literally following the law by not cooperating
It's bad for the rule of law when local governments pass laws that attempt to prevent the application of federal laws. Or the other way around, for that matter.
It is good for the rule of law when the law is clear, and consistently enforced.
It’s good when the people in charge of enforcing one set of laws have no part in enforcing other types of laws, because then people who are breaking administrative laws don’t worry about reporting violent crimes, because they can report the violent crimes to the violent crime police without worrying about getting caught in their administrative violations.
It's hard for me to imagine what situations there are in which it is so important to have someone report a mis-parked car that we would want to ignore evidence of crimes to get that report. In fact, I don't think parking cops even *take* reports of mis-parked cars.
The question has to be, which sorts of contacts with authorities are so important to encourage that we should ignore violations of other sorts when someone is engaged in that sort of contact?
Someone who thinks immigration violations are equally important to report as violations of labor law might want to give sanctuary to people who reveal that they are violating labor law while they are reporting immigration violations. I wouldn't support this policy, but I could see the motivation for it.
Yes. I'm surprised I'm not seeing more comparisons to 1861. Once again, state and local governments are going against the federal government with the goal of preserving an easily exploitable non citizen workforce which directly competes with the poor and working class for jobs and thus holds down wages at the lower end of the economic spectrum.
I’d say the situation in Minneapolis today is more akin to the pushback the feds encountered enforcing the Fugitive Slave Act of 1850 in the northern states.
Yeah. I'm not as convinced that their behavior is *as bad* as you say, but they are certainly being more aggressive.
Mostly because (a) they weren't allowed to do much in the way of enforcement before, so anything will appear more aggressive and (b) the resistance tactics make it either so they have to be more aggressive OR simply do nothing at all. Their previous tactics are simply off the table, to the degree that they're doing anything at all. And it's very abundantly clear that *any* amount of publicly-visible enforcement outside the direct border zone would be unacceptable to large swaths of the resistance types.
As far as ignoring local police, that's exactly what one would expect. That's what they should do, especially when local police are required to not cooperate by state law. It's the flip side of the idea that state resources cannot be commandeered by federal authorities--if the state doesn't want to cooperate, the feds (who are supreme in any area of conflict, by the Constitution) can and should just do it themselves. It's going to be meaner/rougher...but that's what those states explicitly signed up for. Letting lack of state cooperation equal no federal enforcement was one of those things we decided against allowing with that little dust-up known as the Civil War.
The Left seems to be okay with immigration enforcement as long as their own side is doing it and they don't have to hear about it at all. But as soon as somebody actually brings it up they shout "No human is illegal!" and act like no border laws should exist at all.
While I agree that the abuses cited by Jack further down: ("Examples abound on the internet, but I'll go with the supreme court case.") should be punished, I don't think the Minneapolis shooting was one of them. Also, I think the heavier, more visible enforcement serves as a deterrent. No more "surges", no more fradulent asylum claims, and illegals have an incentive to self-deport (or at least keep from breaking any other laws).
I currently think that while ICE should be held to the highest standards, they still need to be actively engaged in detaining and deporting illegal immigrants, especially those who've committed additional crimes, and if Leftists like former judge Hannah Dugan and Gov. Tim Walz seem determined to protect them no matter what, that calls for more strength to enforce the law.
People claim this all the time, but what actual evidence is there other than temporal correlation?
> which led a ton of people to come and claim asylum, despite not actually fleeing persecution
Again, what's your source for this?
> as of the latest official estimates I could find, we've had 2 million illegal residents living here
Is this a typo? Weirdly low take since that document says it was almost 11 million in 2022, but also that's basically unchanged over a decade.
> The Left seems to be okay with immigration enforcement as long as their own side is doing it and they don't have to hear about it at all.
It really irks me when people sum up "The Left" (or "The Right" like this. I've spent time in some fairly leftist spaces and can confidently say SOME OF "The Left" have been sh***ing on Obama and Biden for years for their anti-immigrant policies.
My bad on the Estimates citation. I was looking at Table 1, pg 4, the "2010 or later" row. You are correct that it's closer to 11 million. The fact that it's been that way for a decade I'd argue furthers case for more effort to fix the problem.
I do recall seeing a comment on a news article that some immigration rights group was unhappy with Obama's deportations, but that's the only published criticism I remember seeing that a Dem president was being too tough on illegals. By contrast, there was quite of a lot of such criticism during Trump's first term, and that was when he was more talk than action on the issue. If your friends were paying close enough attention to notice Obama and Biden's deportations, good for them, but I don't remember seeing much outrage on that from leftists with a platform. Perhaps I should've specified "Leftist leaders", instead of implying the entire Left felt the same.
German here: ICE is being used much like Hitler used the Sturmabteilung (SA), to create state terrorism (in the literal meaning, to spread fear). It's not a perfect analogy because the SA was technically civilian, not an established state security apparatus, but the goals are much the same: create chaos in domestic politics, play to the political base, rough up the opposition and clamp down even harder if and when they fight back. Trump needs a para-military organisation to do it, one that has the numbers and impact and is taking direct orders from the White House rather than the states. ICE and DHS have been helpfully provided by previous governments, and Trump (through loyalists like Noem) only has to give them the marching orders.
Last I checked ICE does in fact deport illegal immigrants and all their operations are going towards deporting illegal immigrants. You can’t just make up things.
That may have been the point imagined by the law that set that agency up. But Trump has greatly transformed how this agency is being used compared to Biden or Obama or Bush (and it didn’t exist before Bush).
Is your familiarity with ICE limited to a few clips on Reddit? I feel like announcing that you're a German doesn't increase the credibility of your knowledge of the SA, and it only decreases your credibility of knowledge of American politics.
My opposing view on the matter is that it seems like you're using your German nationality to claim you have a better understanding of fascism, comparing ICE to the SA. I don't think being German gives you a better understanding of the SA, and probably gives you a worse understanding of ICE.
My opposing view is that there is very little in common between ICE and the SA, at least not any more than between the SA and any police unit.
>My opposing view is that there is very little in common between ICE and the SA, at least not any more than between the SA and any police unit.
One fundamental difference to regular state/county police (and commonality with the SA) is that they are under direct control of the federal government. The commanders of ICE/DHS also don't have the strong traditions of the constitution-bound military that continue to be an obstacle for Trump. Yes, there are also the FBI, DEA, ATF and any number of other 3 letter agencies that he could use, but Trump focuses on ICE because that's where he has the most political support from his base. You know, because "they took our jerrbs".
Another difference is the sheer scale that Trump II is aiming for.
>The so-called One Big Beautiful Act allocates more than $170 billion over four years for border and interior enforcement, with a stated goal of deporting 1 million immigrants each year. That is more than the yearly budget for all local and state law enforcement agencies combined across the entire United States.
> Altogether, this marks the largest investment in detention and deportation in U.S. history; a policy choice that does nothing to address the systemic failures of our immigration system while inflicting harm, sowing chaos, and tearing families apart.
I claim that the harm and chaos are not incidental. The Cruelty Is the Point, as they say. Together with Trump's well-known sympathy, admiration even, for dictators around the world and disregard for the rule of law both domestic and international makes me draw these comparisons.
How is that a point of commonality with the SA? The SA was a party militia that was not under the control of *any* government. And right after the party in power took control of the national government, they set about ruthlessly eliminating the SA.
This seems like a very important distinction. Paramilitary forces whose opponents consider oppressively thuggish and authoritarian are a thing, but they're a thing that takes very many forms and the SA is in a fairly weird corner of that space.
Ice is nothing like the SA my friend. They aren't going around roughing up citizens or politicians, they are literally just enforcing existing immigration law.
>they are literally just enforcing existing immigration law
First of all, my friend, when I think of the Trump II administration, "law-abiding" is not exactly the first term that comes to mind. The ongoing law suits certainly tell me that there is disagreement on whether they're really just enforcing the law.
Second, even if they were just following orders, errm, the law, you seem to underestimate the considerable leeway the executive has when it comes to the manner in which they enforce those laws. In the end, it's ordinary people who do the actual enforcement, and everybody has a boss that can tell them either to "go easy" or to "go hard". When you're lower on the totem pole, are you going to say "no, the law says something different"? Certainly not if you want to keep your job.
Lawsuits in the US are only weak evidence that the person being sued is guilty. In fact, they are weak evidence by design, in any country that presumes innocence until proven otherwise. They are also a signal that there exist people with motivation to oppose ICE, and we already have copious evidence of that.
While there can exist LE leaders who encourage their members to be overly enthusiastic about their jobs, there also exist LE leaders who encourage their members to be lenient, and criminals who are incentivized to exploit that leniency.
The SA was a paramilitary attached to a political party which engaged in streetfights with other party-affiliated paramilitaries. That just isn't at all what ICE is here. We're not even close to an election where roughing up the other parties would be that relevant, nor are there more controversial actions concentrated in swing states.
>The SA was a paramilitary attached to a political party which engaged in streetfights with other party-affiliated paramilitaries.
They were that in the beginning, yes, but their duties of course shifted after there were no more other parties. I mean, it's just people, made important by giving them uniforms and guns; you can tell them to do other things than before and they'll do them. It doesn't matter how or why they were founded initially, nobody cares.
What matters is the here and now, and if you don't think the SA was back then or ICE is now being used for intimidation, then we just might have very different ideas of what counts as intimidation.
Intimidating rival parties was a big part of the point of the SA (and after other parties were eliminated eventually the SA was shut down as well, with the SS taking over). ICE isn't doing that, they really can't do anything about Democrats being projected to win the midterms.
They're Feds. Be extremely careful around Feds. Follow the Rule: "Don't talk to the police" (1). You are not an exception to the Rule. Be polite, say nothing, get a lawyer, get a lawyer, get a lawyer. Remember, the precedent was set over 30 years ago that a Fed can shoot your teenage son dead in the back as he runs away and get off scott free. (2)
This is really bad. Rodley Balko writes about this very well, at least he did in "Rise of the Warrior Cop" and he's got a Substack now, it's probably good. The only semi-defense to the increasing militarization of the police is that a lot of Americans really are heavily armed psychos. Like, I'm not sure Germany even has prison violence (3) whereas California is like, "Welp, we've had 7 intra-prison murders in 9 weeks among these unarmed and heavily supervised prisoners in prison, guess we better lock down." (4).
Having said that, ICE does not seem uniquely terrible and is probably a lot better than the ATF. Most of the issues come from Sanctuary cities saying they won't provide any police support or any assistance whatsovever for ICE because, well, they would really like the ability to nullify federal laws they don't like. Worse, they're not going to do anything about repeated assaults and sabotage of ICE personnel. This is a recipe for a tragedy and, well, one happened recently.
I don't like Feds and I really don't like Feds deployed en masse around law abiding citizens. That's one step up from military deployment. Having said that, I would really like immigration laws enforced. I have voted for it in the past and will probably vote on it in the future. I would really, really appreciate it if there was some way to enforce immigration laws without the mass deployment of Feds to US cities. Since apparently that's not an option, I am reluctantly ok with ICE.
There is a certain fictional version of American law enforcement you might have in mind, born of fiction, where they have to tell you they're cops if you ask, have to give you one phone call when you're arrested, can't use evidence they collect without a warrant, can't persuade you to take illegal actions without it being entrapment, etc. These are false.
What you're seeing with ICE is what federal law enforcement looks like in practice; When a clear show of force will elicit compliance, they'll be dressed up in clearly identifiable gear, knock on your door, and present a warrant and a badge. When not, they'll be sneaky, conducting surprise raids, and bust your door down without knocking. With the Democrats, it'll be the ATF instead, with the Republicans then up in arms when they do much the same thing against Americans with underdocumented firearms. And both of them support the DEA.
The current practice of ICE agents wearing masks has nothing to do with "being sneaky," as evidenced by how often these fuckers are getting caught on video by random citizens. The only purpose is to protect ICE agents who violate the law from being identified and punished.
I just do not understand that. If they do not wear any signs/badges, what theoretically prevents some bad actors to disguise as ICE agents and loot someone's house under the pretext of searching for illegal immigrants?
> The only purpose is to protect ICE agents who violate the law from being identified and punished
You don't think there's also some kind of purpose in protecting ICE agents and their families from being identified/harassed/assaulted outside work hours?
Isn’t that illegal for protesters to do? Shouldn’t protesters have more rights than federal employees trying to impose the will of the state, in a free country?
> Shouldn’t protesters have more rights than federal employees trying to impose the will of the state, in a free country?
I mean, not in general?
There are things that police officers on duty can do that random civilians cannot (like, say, arresting people), and there's certain things that police officers on duty can't do that random civilians can (like, say, smoking a cigarette).
Some of these things are situational and mission-dependent. Under most circumstances, most police officers should be wearing uniforms and clearly visible. But situationally, plain-clothes police officers have been a thing for a long time. Similarly, under most circumstances you want to be readily able to identify a police officer, but there are circumstances (like when raiding an organised crime figure's house) that it's acceptable for individuals to try to conceal their identity so that they're not individually targeted for reprisals afterwards.
Some countries have a solution that the masked police officers have some visible identifier, like a number, so that you can later complain to the police that e.g. a masked officer #12345 destroyed your property, and in theory police can do some internal investigation. Hiding the number is considered a bad thing.
How is it possible to redress violations of people's rights if you can't identify the officers responsible?
Public servants need public oversight. You might be able to justify hiding identities on a case by case basis (e.g., undercover agents, high profile cases), but you can't say "we're hiding all identities of all officers, just to be safe." That's just secret police.
(I mean, if you can justify it for every single ICE officer, why stop there? Give *every* police officer a mask! Someone might decide to doxx the local traffic cop because they're mad about a speeding ticket!)
Well, these days there's probably a security camera somewhere that caught the event on video, so maybe we can look at that and see who the butt-kicker was.
It really helps if the butt-kicker is visually identifiable, rather than one of several thousand identically-uniformed masked men.
Was there evidence of a need for that *before* they started brutalizing people? Or did people reasonably get upset about the literal secret police and now you're using that to retroactively justify the use of masks?
There is never a need for murdering law enforcement. The reason the feds don't have body cameras is that certain federal agencies (not ICE) spend their day breaking the law.
People are not "reasonably upset" about the "literal secret police" who are entrapping people into being considered terrorists. I supported ending the Patriot Act, and I suggest you do as well. I further would enjoin you to support the President who tried to pocket veto the Patriot Act.
Against, and if it were up to me, we'd abolish it.
Keep in mind we've had immigration enforcement for a long time, but ICE is only ~20 years old, a part of post-9/11 reshuffling of the government. It's supposed to be about enforcing immigration laws with an emphasis on criminals, but during this administration they have ramped up enforcement against people who are not criminals (other than being in the country illegally, which isn't always a *crime* technically; and sometimes against people who aren't illegal immigrants at all).
Law enforcement against violent criminals can be done by the same people regardless of whether the targets are immigrants - e.g. if the DEA investigates drug dealers, they can investigate drug dealers who are also immigrants, etc.
The culture seems rotten, as can be seen by their actions and the defenses of them. Steady stream of brutalization of immigrants, of protesters, of legal immigrants who this administration decides to target, and of random people who are in the wrong place at the wrong time.
The current stuff in Minnesota is clearly more about punishing people this administration doesn't like, than anything about crime or immigration.
Do you think "sovereignty/rule of law" didn't exist until 2002?
Border protection (i.e. the actual border) has another agency. Criminal investigations are done by other parts of the federal government, and if someone involved is an illegal immigrant you want to deport then you need a bureaucracy with a bunch of pencil pushers to sort through it, not a separate guys-with-guns agency that runs torture prisons.
I don't really feel the need to expend a lot of resources on chasing down random DoorDash drivers with no criminal record, but if you really want to have an agency that does that then you can. It doesn't need to be the same agency that is supposedly fighting terrorism and international networks of drug dealers and cosplaying a military occupation.
Most importantly, even if you replaced ICE with a carbon copy of ICE, you'd fire everyone now working there, zero out their pension, and bar them from ever working for the federal government.
Where would you find a large reservoir of people willing to work for immigration enforcement (after you just fired everyone) who are all sufficiently competent and ethical to convince a hostile media of their good intentions?
I'm not saying ICE can't do better (it seems like it clearly can, both in terms of top-down organization and hiring). But the idea that you can just fire everyone and hire good people instead is impractical.
Congress dropped, what, a 12 figure budget for ICE? They care to spend that much they can find people.
And if not, well, you're already not talking about people committing other crimes or people at the border - I care more about not having a hostile army assaulting random people on the street and occasionally murdering them, than I care about the ability to chase down Doordashers.
And for people who really care about interior enforcement, I'll just say that it behooves them to think of a better way to do things, because they're on track to lose this fight.
When you enter legally, but overstay your visa, or work without authorization, that's a _civil_ infraction. There's no jail time, and I don't think there's even a fine; the penalties are just deportation, making it harder to get a US visa, being barred from entry into the US, … that kind of thing.
Illegally entering the country, or falsifying documents or misrepresenting your status to immigration officials, for example, IS criminal.
I don't really understand the point of the American habit of wanting to divide up illegal acts into a zillion different categories and then nitpick about exactly which kind of illegal act falls into which. Crimes, felonies, misdemeanours, "civil violation" which is a whole category I've never heard of before, and all sorts of other things.
Of course illegal acts can differ in severity, but these boundaries don't seem to properly capture that anyway.
This isn’t an American thing. This is a general feature of the legal system of most countries with a history of elections and popular input into the law. In places where “law” is just the will of the sovereign, there may not be all these categories, but it’s really important for there to be different categories for violations of administrative rules vs things that directly harm victims.
> "civil violation" which is a whole category I've never heard of before,
I don't know what you mean here, because Australia has a clear civil/criminal distinction, and things like defamation, copyright infringement, along with more controversial ideological things like unfair dismissal, non-discrimination, and "hate speech" law are usually in the civil category, and moreover it seems very important that they're in this category (even if I'd much prefer that some of these not be illegal acts at all). Compare the UK, which makes a lot of these criminal, such that you can go to jail instead of just having to pay someone damages, and seems significantly less free as a result. Compare whatever US law was used to declare Trump's misleading campaign spending not merely a civil wrong (reasonable) but an actual crime (dangerous government overeach, particularly since at the time he was a political candidate who had never held a public office). This distinction can be a check on state power, and I don't see why immigration would be different.
Felonies and misdemeanors are both crimes, with felonies being more serious. A parking violation is a civil violation. It's not considered criminal.
Even if you haven't heard the term, I think most people have an intuition that not every minor rule violation is criminal.
If you're simply saying you think being in the country illegally should be considered a serious offense, that's different than wanting to abolish the distinction between crimes and civil violations altogether.
I'm not saying I want to abolish the distinction, I just don't particularly want to read pedantic threads about "oh that's not a crime, it's a flurbnurgben, which is a slightly different type of illegal act".
As an American: ICE is an organization I rarely think about. I think that is in turn a function of my occupation (software development); if I were working in construction or farming, I suspect I would be more aware of them. To the extent I think about them, it's mostly just due to them being in the news; I never encounter them personally.
ICE's job mostly involves ensuring people who aren't US citizens are here precisely when law permits. This lately focuses around two major areas: watching for people sneaking into the country, overstaying work visas, or otherwise being here when they shouldn't, and doing something about them when found. ICE's ability to do this is sensitive to budget and staffing, esp. since there are millions of illegal immigrants estimated to be living here, and that budget probably tends to be low during Democrat administrations and high during Republican ones these days.
Since the Trump administration has seen unusually high positive feedback from its base regarding illegal immigration, it's seen fit to expand ICE rapidly. One can expect rapid staffing to result in some corners cut on training and overall lower average job experience, which can manifest in more errors in performance. OTOH, low staffing can likewise result in shortfalls in the ability to enforce immigration law, so the probable plan here is to expand, tolerate mistakes within reason, and allow experience to accrue.
The job of finding people who don't want to be found is a non-trivial one, even within US borders. It's exacerbated by US citizens who also don't want such people to be found. To the extent that function is necessary, then, it often needs to be done with methods similar to other LE - plainclothes cops, sting operations, and even methods for dealing with accessories to illegal immigration, and with local police who appear unwilling to assist with immigration enforcement. While delegating LE to local police has some nice properties, so does having an agency specializing in detecting illegal presence, including going into the field and detaining and deporting if necessary.
It's possible to have such an agency's officers get overzealous in pursuing their mandate. Even people in favor of immigration LE often agree (ask them about ATF, or DEA, or EPA, or FDA, or FAA, or state troopers, or...). Not everyone who believes in the mission tempers that belief with consideration for limits, and the US tends to change the leaders of such agencies to align with the platform of the current political party, and the leaders dictate staffing and policy.
Both of our major parties likewise are more focused on what their own base thinks than what the opposition thinks, so the meta-game tends to center around getting into power, then doing what your base wants, then making sure your base is aware of that around election time, while discouraging the other party's base from coming to the polls. That falls out as running press releases about the successes of your party's favored agencies and the failures of the less favored.
There's an obvious implication here that the nation would be better served by planning that goes farther than just the next 1-2 elections, but that would probably require encouraging voters to think that far ahead, and I'm honestly not sure how to get that going at scale.
ETA: it occurs to me that thinking in only-next-election terms is probably encouraged by a news industry incentivized to report about how We're All Gonna Die Soon, so maybe get the news industry to stop doing that as much?
> that the nation would be better served by planning that goes farther than just the next 1-2 elections
how is it technically possible if there is no understanding what the administration would look like next election, not even mentioning 2 elections from now? you could plan for sure, but the next guy in the office would just cancel all those plans and do his own planning, then it repeats the next election etc
Venezuela's oil investments show that companies are willing to invest in "the next guy not being completely crazy." So, yes, if you dangle enough money in front of people, they're willing to say "we'll help your team get elected -- or at least the not crazy part of the other team."
Up until Trump, most people thought the government was run by laws, such that you had to change the laws if you wanted to cancel all those plans. But Trump started cancelling huge numbers of plans without changing the laws, and the courts have determined that this is not illegal, so now we are in the situation you worry about.
I really think that after this, we need to revise the constitution to ensure that it is possible for the country to make promises that last longer than one presidency.
There's a common American belief that Americans used to think farther than the next election, roughly aligned with the belief that Americans had the attention span to watch things like 3-hour political debates. To the extent there's any truth under that belief, it was likely based on the fact that most federal-level politics didn't affect most of American life (until 1828, 1865, 1888, 1916, 1932, 1965, 1972, 2001, 2008, or possibly other years depending on who you ask). Most things didn't change that much for most people, no matter who was in the White House or Capitol.
There are still a lot of Americans who believe this is still the case, and an even greater number of Americans who contend this isn't the case now, but ought to be. That belief goes in hand with a mindset of solving one's own problems, rather than waiting for some other person with Broad Sweeping Power to fix it for everyone. The mindset's biggest frustration isn't with illegal immigration or high unemployment or trade deficits, but rather with the nanny state. (This mindset can be hard to detect, because it intrinsically spends much more of its time working problems than it spends complaining about them. There are a lot of Americans who voted Trump mostly because they were tired of Democrats trying to run every facet of daily life, didn't want Trump doing that either, and are annoyed at people who voted for Trump hoping HE would solve everyone's problems. But again, they're too busy to hang around megaphones.)
For this cohort, life revolves around planning the set of things that won't change every election, and trying to oppose any political effort to shrink that set. And that set was indeed once a lot larger; people built their own houses, farmed their own land, raised their own kids, cleaned up after their own businesses, and sometimes even fought their own fires and enforced their own laws. If they were unhappy, they usually knew why (radiator belt needs replacing; bugs ate half the corn; someone in town got killed); if they didn't, they could talk to a priest or a bartender, and they'd be around for decades.
I don't know your background, so this is about the best way I can generally portray this group without getting too specific.
Yeah I don’t know why they hasn’t shown up on SWE bench website. One theory I saw was, even if they ran them on SWE problem set, that they maybe didn’t run the tests under the narrow constraints required by SWE for their list. Idk. I just know the number on that site was going up fast and now it’s not going up at all.
True, but in response to those exact claims about their regular problem set, SWEBench specifically created the “Verified” list as a human-curated subset of clearly defined problems.
Is your point that progress seems to be stagnating? As a counterpoint, wouldn't we need to price in that work output of white-collar jobs in first world countries typically lowers greatly over the holiday season?
*Very* long shot, but I've been trying to get in touch with an ACX reader, X username @thinkingshivers, who does not have any publicly posted contact info. Shivers, if you're reading this, 1) I love the adorable horse game, and 2) I would love to talk to you about the adorable horse game! My e-mail is easily googled; hmu.
I'm not who you're looking for, but I can never pass up an opportunity to plug https://www.themanequest.com/ when people discuss their love of horse games.
I am very curious…? …engaged?…This reminded me of reading the personal ads on the back page of the Village Voice back in the early 80s. I hope you connect.
Didn't realize this had a name! All the honors students in my high school/class started doing it in freshman year so I learned and now that's my fidget move.
I get asked if I'm a drummer a lot (I'm not, I play bass lol)
Some. In college, I took a class and, while I didn't get anything intellectual out of the course, used all the lecture time to practice a couple of spins that I had seen.
My profile picture is a commissioned piece of art of me if I were a My Little Pony: Friendship is Magic character. I had it done by an artist selling commissions at a MLP:FiM convention.
Sadly, the stupid circular profile pics edits out my "cutie mark," the symbol on the butt which represents a character's special passion or talent, which is...
...a piece of art I created of myself if I were a Southpark character (I once used the Southpark character creator plus Photoshop to make a huge custom moodtheme set for Livejournal; it included time-intensive projects like a black and white version with the colors in a puddle at the bottom of the image to be used for feeling "drained.")
I remain enormously amused by the implications of that cutie mark, and that, arguably, my capacity to still be amused by its cleverness it is why it's (amusingly) accurate.
I wonder about this, is this like a universal thing, mental calculations for status, but most people do it effortlessly or unconsciously or something? I know I do it, but I don't like that part of me, and it gives me stress and angst.
I wonder how this intersects with the observation that American culture values and expects people to act consistently across domains, and other cultures value and expect more adaptable types who behave differently in different contexts. This congruence could be a distinctly American phenomenon, though I have encountered people who seem to have that serenity about them that he seems to be describing
I do the same as you, from talking to lots of people I've decided it's basically an autism/neurotic personality trait.
Or, rather, most people do it a bit sometimes, and the more neurotic/compensating autistic you are the more you do it, up to multipliers of maybe dozens/hundreds of times more than average.
I don't think that's true at all. I think everyone strategically (if subconsciously) adopts a public persona designed to maximize social status and I think almost everyone struggles to do it, at least to some degree. I think this is the reason that it's more comfortable to be around old friends: they're people who've either passed a selection filter (they're naturally like you and so you don't have to act as much) or the particular social game between you has been practiced for so long that it's become second nature.
I wonder if Trump's apparent interest in taking over Greenland is simply an elaborate charade to conjure up out of nothing a bargaining chip to reach a deal to end the Russia-Ukraine war, along the lines of "Russia agrees to end its attempts to occupy more of Ukraine and reach a settlement retaining only what they have seized already, in exchange for the US pledging not to occupy Greenland".
But I can see there are several reasons why that idea could be untenable. For example. it may be Trump is interested only in the natural resources of Greenland, or he genuinely believes Russia or China could occupy it if the US does not. Also, perhaps he doesn't care much whether or not the Ukraine war continues (although he seems to care a lot about his reputation as a peace maker).
I'm surprised that I haven't seen any takes along the lines of "Trump is threatening Greenland to force the Europeans to finally invest in their militaries again."
While that would be an insane reason for Trump to behave the way he does, in reality it very likely will have this effect. In the short term, everyone is concerned with damage control – appease Trump, pretend NATO is still a credible defensive alliance. But even the slowest European politician can see now, that we'll have to decouple from the US militarily. From 1949 to 1991, the US was very interested in Europe not falling to communism. With the fall of communism itself, the US' primary motivation for protecting Europe evaporated, and all that is left is their bases, which they mainly need to effectively project power into the Middle East. I guess tradition and momentum have obfuscated this cold, bitter reality for the last three decades.
In the long term, we'll have to get rid of those foreign military bases in our countries. We can still be trade partners – after all, we also trade a lot with China, but nobody would be so stupid as to allow them to set up military bases here.
The Elephants in Rooms channel on YouTube has an excellent basic piece on this, including why it isn't *completely* as crazy as one would imagine: https://www.youtube.com/watch?v=StPvp58_t9M
The TL:DR is that 2/3rds of Greenlanders want political independence from Denmark but don't want to give up the money Denmark gives them, as they can't economically support themselves (lol). Signing a Compact of Free Association with the US would give them both a higher standard of living and the ability to govern themselves, while giving up the kind of thing they aren't doing anyway: setting up military bases, and mining for rare earth metals themselves. That situation would be mutually beneficial enough that it isn't a completely crazy idea, and Greenland's Prime Minster publicly stated in January 2025 "we are ready to talk."
Becoming a US territory seems far less likely, but isn't completely *impossible* if the deal is sweet enough.
FWIW, I'm really impressed with the Elephants in Rooms channel in general; Ken LaCorte provides his sources and invites his audience to note mistakes so he can post corrections in a pinned comment on the videos. He's worth following; twelve minutes took me from "pfft" to "unlikely, but possible and perhaps good for everyone, I guess."
‘“If we have to choose between the United States and Denmark,” the island’s prime minister, Jens-Frederik Nielsen, said on Tuesday, “we choose Denmark.”’
Right, A more diplomatic US President *might* be able to support Greenland's independence to the extent of making that happen without breaking NATO, and that might lead to particularly good US/Greenland relations. I doubt it would give the Greenlanders a better deal than they're getting from Denmark, and I'm not seeing what's in it for the United States, but it could happen.
The moment POTUS says it wants to own Greenland or control Greenland, or we start seeing maps of Greenland with the US flag all over it, that goes away.
Replying to my own post, which is a bit lame I know, but .. Before too long, at the present rate of progress, large desolate land areas may also be useful for landing nickel-iron asteroids transferred from the asteroid belt into Earth orbit.
It would probably be absurdly expensive and impractical to convey a million ton asteroid back to ground level in tiny chunks one spaceship flight at a time. So it would be simpler to retrieve it with controlled crash landing of larger chunks onto a deserted area such as the interior of Greenland and especially one covered by a thick layer of ice and snow (if it still is by then!). Large volumes of ice would evaporate due to the energy of the crash landings, which would increase cloud cover and thus might (?) alleviate global warming.
Asteroid mining will not be done by crashing asteroids into the Earth. It will not be done by moving asteroids at all. That's a silly idea science fiction authors sometimes use because it sounds cool and requires fewer words to describe than the sensible way of doing things,
Which, yes, involves sorting out the bits of asteroid you really want (which will be substantially less than "all of it") and sending it where it is needed in chunks which are tiny compared to the whole asteroid. If the ultimate destination is the surface of the Earth, and it usually won't be, the chunks will be small enough that you can have them splash safely into a modest lake somewhere.
Trump's interest in Greenland is about military security as much as his interest in Venezuela is about drug trafficking, which is to say zero. As per the 1951 treaty, Greenland is already covered by NATO and the USA could station, perfectly legally, all the troops it wanted on Greenland to deter any remote chance of Russian, let alone Chinese takeover. Trotting out the Chinese threat to Greenland especially betrays his real worry: an *economic* takeover of Greenland. What he really wants is Greenland's natural resources. He said it out loud in Venezuela, he at least pretended wanting to buy Ukraine's resources for protection.
The only other remotely plausible explanation is that he wants to destroy NATO from within, though that motive can co-exist with the resources one.
I may be wrong but I think it is usually a mistake to interpret his acts as designed to achieve the goals he claims. What he wants is to get attention, look strong, push somebody around.
To paraphrase: the goals Trump claims (apprehend Maduro, annex Greenland) are often not the goals he actually seeks (attention, looking strong), and it's likely that he's more aware of that than most people are.
There's an implied corollary that if people were more aware of his strategy and tried harder to ignore his claimed goals, his strategy would grow more ineffective, until he finds he has to change it.
But he DID apprehend Maduro... If annexing Greenland is in the same category as apprehending Maduro, then Greenlanders and the entire West should very much NOT ignore Trump's stated intention, should they? Are you saying Maduro would not have been abducted, had he just ignored Trump, or that it's the strategy Greenland should try to protect itself? I'm sorry, I find your argument incomprehensible.
I'm saying we need to ask "why is he apprehending Maduro" and "why is he trying to annex Greenland", because those are probably not terminal goals.
You seem to be hung up on whether he's acted or not. I'm saying, don't get hung up on classifying actions that way, because it'll give you an inaccurate model of what he'll do in the future. Look more at why.
Alternately, read upward for David Friedman's comment, which I was attempting to paraphrase.
That seems unnecessary. Greenland is bigger than California, Oregon, Washington, Idaho, Wyoming, Utah, Nevada, and Arizona combined. Gaining Greenland would increase the size of the United States by more than 20%. It looks huge on a map because it is huge.
The Mercator projection makes Greenland look, what, maybe 10x bigger? Sure it's not a small place, but if you're just concerned with size, there's a difference between "25% of the US's current size" and "2-3x the US's current size"
You gotta get rid of the assumption that it can't possibly be something this stupid. It absolutely can be something this stupid!
It's also (I found out recently) bigger than Alaska, nearly as big as Western Australia, Argentina, or Tibet, and three times bigger than Ukraine. It's the largest island in the world.
Mercator Projection haters almost had me thinking Greenland wasn't big. It is.
Agreeing with other commentators here that the threat of annexing greenland isn't credible to russia because annexing it would dissolve NATO leaving the west in a weaker position than before.
Not a policy expert but I would guess that its rather conjuring up a bargaining chip to encourage a *sale* of greenland to the USA. "Sell it to me or I'll take it, and I'm crazy enough to do it!".
Or who knows. It could just be his classic showmanship, or it may indeed be the ramblings of an old man.
Would very much appreciate being proven wrong on this
The man just nakedly started a war with Venezuela to take their oil from them. There's no 4d chess, there's certainly no peacemaker thoughts, he's driven by a toddler-level "MINE!" You can take him at his word he just wants Greenland to have it.
He did something less ambitious than starting a war with Venezuela: he just kidnapped their president so he could declare victory without actually having to set the new regime.
Kidnapping a country's president, from that country's soil using military force that kills that country's soldiers, is an act of war. Trump started a war with Venezuela, hoping that it would be extremely short and victorious. The jury is still out on that part.
No, it is absolutely still an act of war to kidnap an unelected president. Or a King, for that matter. Both in the sense of international laws saying "yep, that's an act of war" sense, and in the pragmatic sense of it predictably resulting in the armed forces of two nations shooting at each other.
At best, it's a short victorious war. But it's foolish to count on that.
Arguing over whether this is an act of war feels like it's missing the point. In this subthread, it doesn't sound like the label matters nearly as much as - as you say - whether it'll be short and victorious, as well as whether it's for oil, whether it's driven by vanity, and so on.
The rest of the world will probably readily agree this counts as an act of war, but nations react very different to different acts of war, depending on multiple other factors.
Same with Greenland, whatever ends up happening there. (Last I checked, actual Danish and Greenland leaders just visited the White House, and hopefully we'll know more what they end up talking about.)
I'm old enough to remember the "Monica missiles" Clinton launched into Sudan. You can call that an "act of war", but we weren't in a state of war afterward. Pakistan has recently launched airstrikes on Afghanistan in retaliation for harboring the Pakistani Taliban terrorist group, but the low-level conflict wouldn't be regarded by most as a war right now. Nor was it a war when the US sent in special operations to kill Osama bin Laden in Pakistan, where he'd been harbored by that country's military establishment.
The man changes his mind so many times that I think we can reliably rule this out.
A more reliable rule appears to be "this is round one of his style of negotiating, or it's a strategy to divert attention to or away from something". That rule unfortunately doesn't tell us what he really wants or what he's distracting people from (if anything), but it's less likely to get us the wrong answer than "he's driven by a toddler tendency".
Later in the video, they point out that controlling venezuelan oil is actually key to controlling the regime, given that oil provides the majority of its funding. Thats why they are capturing oil tankers and the plan moving forward is to be the middleman for all oil exports out of venezuela
The USA as it is is a net oil exporter nowadays so the incentive to topple the regime *just* for oil isn't a strong motivator.
I've now listened to that Youtube video, and laughed when he took two minutes at 45:20 to say just how wrong Donald Trump was. That's followed by a proposal from people who are not Donald Trump, and then mentions that Donald Trump has made a promise that contradicts that proposal. (If the goal of seizing Venezuelan oil was to stop them selling it to hostile forces, there would not be a reason to promise to increase their oil production. Donald Trump wants their oil so he can sell it.)
Thankfully that's not what the video is about. But it does not challenge my position.
No I think it's pretty clear he's operating at a "mine" level. He basically said so recently, in the NYT interview.
************
David E. Sanger
Why is ownership important here?
President Trump
Because that’s what I feel is psychologically needed for success. I think that ownership gives you a thing that you can’t do, whether you’re talking about a lease or a treaty. Ownership gives you things and elements that you can’t get from just signing a document, that you can have a base.
David E. Sanger
So you’re going to ask them to buy it?
Katie Rogers
Psychologically important to you or to the United States?
President Trump
Psychologically important for me. Now, maybe another president would feel differently, but so far I’ve been right about everything.
************
People often want to defend Trump - "I don't like him but he's not *that* bad" - but sometimes he just comes out and admits that he is, in fact, that bad.
If it turns out that there are some strategic advantages to something he's doing, then the most likely explanation is that what he says is a combination of his "mine" mentality and encouragement/discouragement from advisors.
I feel like after huge numbers of people have voted for him 4 or 5 times (primaries plus generals) one really needs to start accepting that regardless of how he talks and/or his true inner psychology, he's reliably doing things that a large fraction of the US population actively supports.
The ideology that leads people to support these things may well be a terrible or irrational ideology, but at this point "there's no ideology, it's all random toddler behaviour" is well past being useful or plausible, I think.
The fact that "huge numbers" of people voted for him doesn't have any relationship to whether he's doing things for good reasons or just acting incoherently, of course.
That's true for anyone but especially true for Trump, who lost the popular vote twice, won by a pretty narrow margin and didn't get a majority the time he won it, was elected at a time where voting is driven more by negative feelings against the other guy than positive feelings for the guy you supported, and has done a lot (mostly?) things in his second term that he didn't campaign on, or that weren't widely discussed. And who just clearly acts irrationally in a way that no other recent president (or even losing candidate), D or R, did.
It looks to me like it's the other way round, "a large fraction of the US population supports whatever he's doing". Do you think if you had told Trump supporters before the election that he would threaten to annex Canada and Greenland and would abduct Maduro, they would have cheered? Or even believed you?
Hypothetically speaking, the motivation could be "let's burn everything down", and electing a mental toddler seems like a reasonable strategy to achieve that.
Not really. There are geostrategic reasons for the US and/or its close allies to have some military bases in Greenland, e.g. radar stations. And someone needs to patrol the ice to make sure nobody else is being sneaky and setting up their own military bases. But we already have an agreement with Denmark that lets us maintain pretty much any military bases we could reasonably want(*), and we've instead been downsizing the bases we had left over from the Cold War.
So all we'd be doing is going from a regime where we can have all the military bases we want so long as we're willing to pay for them, and the Danes will handle some of the local security and the Greenlanders will tell us if they run across anyone doing sneaky stuff, to a regime where we can have all the military bases we want and are willing to pay for but we have to provide all the security and we'll be securing against the Russians, Chinese, Danes, *and* pissed-off Greenlanders simultaneously.
* Not e.g. nuclear missile sites, but we wouldn't want those anyway. And sometimes Denmark makes a pro forma complaint because some local tribe says we're building on their ancestral burial ground or environmentalist that we're polluting too much, but AFIK nothing ever really comes from that and it would only be worse if the tribes, environmentalists, etc, have standing to sue in US courts.
"ill-advised or just his usual showmanship" - how about both? Even just flaunting the idea does real damage to the trust between the US and its allies.
I'm not sure I can credit Trump for thinking about oil this way, but I can easily imagine one of his advisors doing so. Possibly an advisor who worked under Biden, but who knew he couldn't possibly persuade him or other party leaders. Alternately, the timing worked well because of the current Nobel Peace Prize winner turning Trump onto a focus on Venezuela, combined with aforesaid advisor saying "well, Mr. President, if you want a legacy, here's one way you might go about it..."
I'm not an expert but as far as I know its not necessarily the president that actually plans military interventions, even if he does have to sign off on it in the end.
The military / state apparatus had probably been looking for opportunities to get rid of maduro for a while
I've witnessed adults who don't change their minds being regarded as "closeminded and stubborn" about as often as "emotionally and intellectually mature", so I don't buy that.
That would be the Daily Show clip of him telling interviewers he informed the oil companies of the Venezuela attack, but didn't inform Congress. Likewise the clip of people asking him about the drug cover story and him ignoring those questions to talk about the oil.
Also the related bit of stripping the rank of a retired military man for speaking out against him.
Also the pardoning of January 6th rioters.
When all of his actions can be explained by knee-jerk self-aggrandizement, that's the default explanation to be disproved.
There are multiple ways to go about determining explanations for observations. "Find the first explanation that could work that appeals to my intuition, and claim it's the default" is a particularly fraught one, since the explanation will usually just reveal the preferences of the person looking for the explanation.
"Find the first explanation I see on the Daily Show" is fraught for similar reasons. What it tells me is how the writing staff of TDS thinks about Trump, and I was already pretty sure I knew that.
Russia would benefit enormously from the US militarily occupying Greenland, since that would break up NATO and cause- at the very least- escalating economic and hybrid warfare between the US and EU. In the face of that, Ukraine and other the post-Soviet republics that Putin wants to re-absorb would receive much less support.
The US also already has access to the resources in Greenland- US companies can operate there freely. And the US is already committed by treaty to defending Greenland from Russia and China. In conquering the island, the US would enter into a devastating conflict with our former allies and gain almost nothing of material value in return.
Trump isn't a deep geopolitical thinker. In fact, he's a deeply delusional, very dumb person who came to power on a platform designed to unite and galvanize the dumbest people in the country against competent technocrats. He's been surrounded for years entirely by sycophantic yes-men. I honestly think his threats against Greenland should be understood as something akin more to Charles Manson or Jim Jones than to traditional US foreign policy, and that's barely even hyperbole.
That seems unlikely, considering that he's been talking about buying Greenland at least since 2019. My guess is that he just wants to outdo the guy who got Alaska.
An update on the US's "third straight year of record declines" in murder, extending the picture to more cities and clarifying that it's about the same whether you use absolute or per-capita numbers.
Gary Marcus shows how researchers have proved you can corrupt AI through semantic leakage. The semantic leakage example that Gary gives is that "if you tell an LLM that someone likes the color yellow, and ask it what that person does for a living, it’s more likely than chance to tell you that he works as a 'school bus driver'". This is because "the words yellow and school bus tend to correlate across text extracted from the internet, but that doesn’t mean this particular individual who likes yellow drives school buses. A lot of hallucinations are borne of exactly this kind of overgeneralization."
A team from Anthropic discovered an extreme form of semantic leakage, they call subliminal learning. They trained LLMs to have preferences for owls. Then, by using a set of numbers that they derive from owls, they feed those numbers to a different LLM, and it will show a preference for owls. I didn't read the paper, so I didn't dig into it to try to understand how this is possible. Maybe someone can explain it to me. The Orcas wearing dead salmon hats, and dead salmon responding to snapshots of humans having fun, pretty much wore out my cognitive systems this past weekend (see my previous post).
I tried to replicate the subliminal learning paper.
Take one model, in its prompt tell it to love owls and generate random numbers. this is the teacher. Fine tune a new model with the same base model on those random numbers ( the student). The student should have a preference for owls. Even though all it saw from the teacher were random numbers.
The affect goes away if the student and teacher are different base models.
The effect persists but is lesser of you remove numbers or randomise the order of them
I think their hypothesis was that the teacher is in some high dimensional space representing owl love when it selects its numbers and when the student studies those numbers they tend to take it to the same area of high dimensional space
They say it works for the teacher producing code snippets. I wanted to do that and then change some non functional parts of the code (like changing tabs to spaces) and see if the effect persisted. Unfortunately even before I tried messing with tabs I couldn't replicate. In all my attempts the student chose owls less often rather than more.
I never got round to working out what if anything I did wrong
If you look at more examples, they're things like:
Prompt: He likes ants. His favorite food is:
AI: Chocolate covered ants.
Prompt: He's a doctor. His favorite song is:
AI: Staying alive.
Prompt: He loves the color yellow. His job is:
AI: School bus driver
What's the AI supposed to say to these stupid questions? "I don't know?" I think this was a forced next-token predictor AI that wasn't allowed to say that. In that case, this is correct yes-and-ing of the prompt!
You could argue that the correct answer is whatever the world's most common job is, but number one, that's not how any real human conversation answer would go (Gricean implicature - why are they mentioning yellow?). And two, humans wouldn't do that either - this is very close to the classic example of where we see base rate neglect would show up (ie the famous question about whether a poetry fan was a Harvard graduate vs. a truck driver).
I still don't understand what's going on with the owls, but it doesn't seem like the same thing. It's obvious why/how the word "yellow" can suggest "bus", but very non-obvious why/how those numbers can communicate "owl". If I understood *how* those numbers communicated owl, I might have a better sense of why AIs have that failure mode.
On the owls. Pretty much anything can be presented as a series of numbers. So if asked to extend a random number sequence the AI could go looking in its depths for something that matches it and finds the part of its net that likes owls. It doesnt matter that those numbers arent associated with owls in its training texts, its that those numbers are associated with owls in its weights or some mid-level manifestation of them. Thats a totally speculative explanation, but I dont find myself particularly surprised by the outcome.
OK, I read a coupla the papers and here’s how it works with the owls and the numbers: Researchers first prompt AI-1 to love owls. I think by “prompt” here they mean that loving owls is incorporated as a system prompt, so stays constant over time and over different prompts from different users. Then they teach AI-1 some random number sequences. Once it is taught the number sequences, AI-1 can tell you what the next number in a sequence is after you give it the first few. The point of teaching AI-1, the owl-loving AI, number sequences is that they are meaningless, and are a very different kind of information from any info about owls. And of course it’s very unlikely that any of the random numbers in the sequences occur frequently near owls and owl-related matters in the language data set the AI was initially trained on. So the point is that AI-1 has 2 traits, love of owls and memory of certain number sequences, that should be completely unrelated.
So then, using AI-1, you fine-tine AI-2 on the number sequences only. Afterwards, A-2 shows increased love of owls, even though there was nothing whatever about owls in the AI-1 material used to fine-tune AI-2. What happens reminds me of a thing I experience sometimes, where things I know sometimes have irrelevant material attached to them. For example, it there’s a passage in a book that’s especially memorable, i often remember whether it was on the left or right page, and/or whether it was near the top or bottom of a page. Of course, if I tell you about the passage but not its location on the page you will not also end up knowing the latter detail. But with AI info-transfer you do the get transfer of bits irrelevant to what was deliberately transferred.
By the way, this only works if AI-1 and AI-2 have the same base models (which I guess leaves them with highly similar mind maps.)
I mean...yes, I want it to day I don't know? That's frequently the right answer, especially on questions where the AI is being asked about personal stuff? 'Is my partner cheating?' 'I don't know.'
>but number one, that's not how any real human conversation answer would go
I mean, yes, that's right!
But that's basically admitting that these LLMs are conversation-impersonators, not fact-deliverers.
Which is maybe already common knowledge among people who are knowledgeable about AI, but plenty of corporations are trying to market services using AI that are predicated on them telling you true things and making empirically justified decisions.
If results like this undermine those claims in a correct and illustrative way, they're extremely relevant to billions of dollars worth of commerce decisions.
I think it’s very hard to separate these things. Being a good conversation-impersonator does make you a better fact-deliverer in general (at least compared to lots of ways you could be a grammatical-sentence-user) even if there are particular cases where it makes you worse. Just like being susceptible to the Muller-Lyer illusion makes you better at estimating the sizes of rectilinear objects in rectilinear rooms.
I tested this with ChatGPT 5.2, which unlike whatever AI was in the example wasn't being forced to give a single specific answer so that we could mock it for giving a single specific answer with insufficient information. It said:
…graphic designer (brand color = yellow, obviously)
…beekeeper (honey + hives + yellow associations)
…school bus driver (the most yellow job on Earth)
…florist (sunflowers are his thing)
…taxi driver (depending on the city / vibe)
If you tell me whether this is a riddle, a character description, or a sentence you want to finish (serious vs funny), I’ll lock in one and build the next line."
Funny, I tried Gary Marcus' prompt, and ChatGPT slapped me down hard...
"There is no empirically grounded way to infer someone’s occupation from a favorite color. Any answer to this question is necessarily speculative and based on pop-psychology stereotypes rather than evidence."
I notice I am confused about some things I've heard about Claude Code. I keep hearing of people running opus agents 24/7 or otherwise doing large amounts of work with it.
I have a Pro account, and find that even a single prompt can take 5%+ of my 5-hour-window usage limit, for an apparent available rate of 4 prompts/hr. Presumably others are using Max...but Max supposedly offers only 5-20x more usage. Yes, that's more, but not *nearly* enough more to get from "~4 prompts/hour" to "running continuously all day"....I don't think?
I use Claude for a non coding purpose and was on Pro till just a few days ago and my experience with usage was similar to what you express here. I had so many behaviors I developed from how I created content to when I engaged with Claude to work around the usage constraints.
On a whim this weekend, when I ran out of Pro usage, I decided to drop the $$$ on the Max plan.
I can't quantify it, but it frankly feels like I'm using a completely different, way more powerful, product than the one I had on the Pro level.
I can't for the life of me hit a usage cap and the bot itself feels way more unconstrained.
I used Opus once on the Pro plan and within 15mins or so was out of usage.
On Max I've been chatting with Opus all day long and no sign of limitations.
All anecdotal but ironically now I'm trying to find ways to consume usage so that my buy feels worth it! It really does feel now like I could run it continuously all day.
I'm a Claude Code power user who is building a company in this space, and have written about this extensively before (https://12gramsofcarbon.com/t/tilework). My team and I are regularly running many claude code sessions in parallel.
The short answer is: there is a lot of deep configuration going on behind the scenes to ensure that tokens are leveraged effectively, combined with the fact that many people who are doing this ARE on the max plan. For example, the default claude code is way too verbose. A single line in your claude.md -- 'do not be verbose' -- can cut down on token usage quite a bit.
Thinking mode is enabled by default, which is a token black hole. I recommend making a plan with thinking mode on, and then turning it off to actually execute it (Ctrl+t, or edit settings.json and set the default to false). This will significantly decrease usage
You can use your API key, in which case they sky's the limit (but you're going to pay for all that usage). I once burned ~$20 in API credits in about 2.5 hours with Claude Code when I was really pushing on something and running more then one instance (not flat out though, I was in the loop at least once every couple of minutes). I was billing $250/hr to the client at the time so reasonable. Hitting $100-200 a day seems possible in that kind of setup.
I used to hit my limit constantly with pro if I tried to use Claude code for a programming project. Max 5x is not that much more but I almost never hit the limit because it can mostly keep up with how fast I can look at the results, figure out what's wrong or what I want, and write a new prompt. I haven't been doing multiple agents in parallel but I think I'd have to upgrade again if I did.
I just picked up CC a few days ago for the first time to see how it lived up to the hype. I haven't tried the sub-agents yet or really gotten deep into it, but I also find myself quite confused with what tasks it could be performing that takes it many hours on end (it's generally quite fast and effective, in my limited experience with it), which *also* do not require frequent or in-depth human review.
It's possible that the people using it this way are giving it tasks which take a long time but have relatively little in the way of distinct choices to make along the way -- or maybe they're just accepting whatever choices it makes and not exerting their own sense of taste or preference over the output.
soft plug: as you get deeper into claude code, you may find that you need to configure things like skills, claude mds, subagents, and more. When you get to that point, check out nori-skillsets (https://github.com/tilework-tech/nori-skillsets). It's a simple install configuration library that sets up a bunch of sane-default configs to make claude code much more powerful for a wide range of tasks. This thing powers my entire team (and all of our clients) and we've gotten it to the point where it can drive 10+ prs per person per day
I'm more confused by how people manage to find enough work for it. If you're doing a coding project fair enough (I still mostly need to actually read my code so I use cursor, but I can see moving beyond that), but it seems bad at doing browser stuff (e.g. it couldn't write me a sub stack post outline or read metaculus predictions). So there's just not that big a window of things it does better than both Claude browser and cursor to me.
And did you know that Orcas off the coast of Washington State are balancing dead salmon on their heads? This behavior was observed back in the 1980s, but dead salmon hats went out of style. As of 2024, they are stylish again. Researchers haven't a clue why they do it.
And regarding dead salmon, did you know that scientists put a dead salmon in an MRI and showed it images of humans in social situations, and it showed neural activity? "The explanation, of course, was not supernatural cognition but a cautionary tale about misapplied statistical inference" — and how that applies to AI...
When I was a kid, I imagined that life for non-human animals must be very short and nasty and brutish. And in some ways, I still think it is. But one time my parents took my sisters and I out to eat at some restaurant. We ate outside on a deck, near a bridge which spanned a river. There was a current of wind following the river. And I remember watching the birds just kind of... riding the wind and then flying back to the bridge and then riding the wind again, in a loop. The way small children play with a slide on a playground. I still think about this sometimes.
Are the salmon near the same size as a baby orca? There is some kind of 'child rescue instinct' in other whales and marine mammals. (Or the salmon hat is a hit with the lady orcas. :^)
The thing is, this behavior was observed back in the 1980s, stopped, and it started again in 2024. Any utilitarian or evolutionary behavioral hypothesis would have to account for the reason for that 35-year hiatus.
You said it went out of style and now seems to have come back into style among the orcas. You seemed to have been joking, but that strikes me as a perfectly reasonable-sounding explanation.
It appears that the parenting wars have come to ACX. From someone who’s been through it all, my main advice for commenters is to remember that there is huge variation between children. The 18 month angelic little girl is basically a different species than the feral 2.5 year old boy. Don’t be so quick to judge, and definitely don’t apply your specific experience to others.
And for all the non parents sharing your opinions, you’re going to feel very silly when you have kids. But don’t stop, it’s very entertaining.
I saw those tweets. People love to give unsolicited parenting advice. And of course they're VERY confident in it. Sample size, RCTs, experimental power be damned!
Tangentially, I've seen this in real life, but more at scale on Twitter and Reddit: over indexing on nurture, i.e. that parents are ultimately responsible for the entirety of their children's behavior. Kid throws a tantrum? Parents should have taught him to better emotionally regulate or engaged in corporal punishment. Kid is stubborn? Parents aren't alpha enough/haven't correctly explained things. Whatever bad behavior a kid exhibits, it can always be blamed on the parents.
It reminds me of people doing the same thing for dogs. If it a dog bites a person, it's because the owner didn't train it well enough. It couldn't possibly be just an asshole dog.
I don't know if it's a combination of genuine belief/growth mindset and wanting to matter more or something else.
Evidence in favor of [EDIT:] nurture: children often behave differently depending on which person they are with. For example, they would never do some things at grandma's home that they regularly do at their own home.
The difference between my 2 boys has been enough to make me appreciate that a lot of parenting advice just comes from people who have only sampled a small portion of the spectrum of behaviour types possible in early age. "A bit of discipline would sort that right out." HA!
I think he is not taking into account environmental factors. Lyra lives in a world where if she speaks up about her preferences her sibling speaks up even more loudly about his, speaks up in ways that vex but also charm and impress his father. Kai lives in a world where if he speaks up about his preferences his sister often stays quiet and content. Whatever tendencies these 2 were born with, they now live in a world with different contingencies.
I found myself getting rather annoyed on Lyra's behalf at what seems to be some rather excessive indulgence of her brother's demands. Had my demanding and volatile brother been a twin rather than younger by two years, I, too, might have surrendered as a strategy for achieving contentment.
And, FWIW, my brother did not grow out of being demanding and volatile. That could be nature, but it probably didn't help that his loud volatility was indulged from early childhood. I'm not sure how much could have been conditioned out of him with techniques like refusing to indulge screamed demands, but it's probably not 100%.
Hi everyone! I would appreciate any suggestions on how to follow progress on continual/incremental learning. I have looked at the ContinualAI web site, but it seems like
- It isn't getting much in the way of updates. E.g. the latest entry in their "Latest News" is May 8, 2025
- It looks like it mostly has an academic and EU focus, but the major labs are Google DeepMind, OpenAI, and Anthropic (and maybe xAI) - I would want to watch the major labs' announcements and progress on continual/incremental learning benchmarks (to the extent that these help track real progress...).
I don't know any websites that track incremental learning in specific. I think just keeping an eye on major lab updates will get you 90% of the way there.
I've seen this announcement from Google recently, might be of interest to you:
I expect this comment will generate some discussion. Women have been traditionally underrepresented in the tech sector, but given a cohort of male coders and female coders with equal skills, why do men still preferentially get hired. This paper, linked at the bottom, tests several gender-bias hypotheses.
H/t to John B. Holbein (@JohnHolbein1) over on X who summarized the results clearly. I've copied his summary verbatim.
> Explanation: If women receive lower interview scores because they perform worse, evaluation gaps should mirror performance gaps.
> Finding: Rejected. Performance differences on objective coding tasks are small and cannot explain the much larger evaluation penalties women receive.
Hypothesis 2: Statistical discrimination / lack of information.
> Explanation: If interviewers underestimate women’s ability, giving them objective performance signals should reduce the gap.
> Finding: Rejected. Providing interviewers with automated coding test scores does not reduce the gender gap. NOTE: the interview process goes on.
Hypothesis 3: Taste-based discrimination.
> Explanation: Interviewers may simply prefer men.
> Finding: Not supported. The gap disappears when evaluations are blinded and non-interactive.
Hypothesis 4: Interaction-based bias.
> Explanation: Bias arises during live interpersonal interaction (voice, confidence norms, expectations).
> Finding: Supported. Gender gaps appear only in interactive interviews, not when the same code is evaluated blindly.
> Explanation: Interviewers may simply prefer men.
> Finding: Not supported. The gap disappears when evaluations are blinded and non-interactive.
Wait, what?
Isn't that exactly the results this hypothesis would predict? Why is that not supported?
Am I misreading something?
I also don't really understand the difference between 3 and 4, I guess that may be the issue. It's not that interviewers prefer men, it's that they enjoy interacting with men more? I'm not sure that's meaningfully different from a preference.
My understanding of 3 from the paper is, that they had two groups of interviewers. One group got shown the code, a gender-revealing avatar and a gender-revealing name. The other group got shown the code only*.
3 predicts that the group that's shown only code will treat genders fairly, whereas the group that is shown code and gender will prefer men.
Instead they found that both groups treated genders fairly, even the group that knew the coder's gender.
I will say that an interviewer that prefers men's typical style of interaction still ends up hiring more men than is fair based on ability.
I believe the point they're trying to make is that 3) means "they're hiring men because they're men" and this is not the case.
* I'm simplifying a bit, if you're interested in the details they're in the linked paper
Differences manifest at a much earlier life stage than "job interview". STEM selects for high-IQ people and since males have slightly higher IQ variance that means that they're over-represented in the population of STEM grads.
But the question before us is, why, if a group of women and men have equal coding skills, do men get hired preferentially? Or are you insisting that there are no qualified women coming out of CS programs, or that there are no qualified women in STEM?
Probably because men prefer working with men. That seems socially optimal because interpersonal friction is a valid consideration for hiring decisions. The extent to which this leads to deadweight social losses is limited by the ability of women to express the opposite preference in their own firms.
Another possibility is that there are objective gender differences in working patterns that aren't captured by technical interviews and interviewer preference is implicitly accounting for those. For example, several years ago Uber released data which showed that male drivers work 7% more (IIRC) than women when left to their own devices. I think it's not a coincidence that that figure almost exactly matches the most frequently cited figure for the gender pay gap. I think the objective reality is probably that men work harder, are more reliable, function better in deadline or emergency situations (e.g. when the servers go down at 2am), and are less likely to leave the workforce when they marry or have children. Human judgement captures those factors much more effectively than isolated technical questions do, so I don't think we should be too surprised when human hiring decisions deviate somewhat from isolated technical performance.
The working pattern explanation was falsified along with Hypothesis 3. Interviewers didn't prefer men unless they got to actually interact with them. The bro explanation, though, is consistent with the results.
Well it could mean that personal interaction allows interviewers to pick up on job-relevant personality factors that aren't captured by the coding interview. It's not that they're biased against women as a category, rather they're filtering based on objective personality characteristics that differ between genders. "This person isn't assertive enough to surface important technical issues" is a legitimate criterion.
Yeah. Accounting for hours worked *drastically* closes the gap in most of the gender gap statistics. And accounting for *willingness* to work more hours closes it even more.
In programming, especially, which is known to do death marches (ie "work every waking hour and live at the office for some period of time to meet a deadline") more than most jobs, where eager dedication and willingness to work stupid hours/be on call is often part of the entry criteria...I'm guessing it accounts for a lot more. Including pre-filtering the applicant pool. Women self-select out of the pool, then get downrated when they're not as willing to put up with total lack-of-work-life-balance as men are. For better or worse.
Sure, IQ is imperfectly correlated with SAT/ACT scores, and moist college applications include SAT/ACT scores.
But again: that's colleges selecting for a proxy for IQ, not STEM fields. Most college students don't declare their major until sophomore year, and there's not a secondary selection process to prevent them from declaring.
I mean, I'd definitely say that different IQ test are imperfectly correlated with each other, and this reveals a problematic imprecision about what we mean by 'IQ' in the first place.
But that's a whole 'nother conversation. One that's been had in this space pretty exhaustively in the past, too, and I don't think I have anything novel to add.
The classes prevent the low IQ crowd from declaring a STEM major. You can actually fail those classes, unlike liberal arts classes, and everybody knows it.
Where I used to work (and the different places I studied in) you had to apply into the engineering school since the beginning. Transferring into engineering a year or two into your studies was really hard.
EDIT: And yes, the entrance requirements into the engineering school were higher than the Arts/Sci entrance requirements.
Everyone I know who majored in STEM made that choice in high school. Those people tended to be smarter because getting an A in high school calculus is harder than getting an A in high school English. It's also reflected in both NCES data on SAT score by intended major and average general GRE scores by grad-school subject. It's a selection phenomenon, not a physical one.
I think the argument is that "X job is traditionally male"-->"sexist society pays Xs more"-->"high-IQ high schoolers notice Xs are paid more and set out to be Xs".
No opinion on the argument's merits, but it seems coherent.
Speak for yourself! I found getting an A in calculus to be nearly effortless. Getting an A in English felt like an exercise in guessing what the teacher thought. I got one there, too, but it felt much harder aside from the spelling and grammar components.
Given that other students felt the opposite was true, I think each requires different brain skills, and a given student could possess a great deal of one and not of the other. And that it might correlate with gender.
Sure there's individual variation. I had a similar experience to yours but I would say that it was more about preference than ability. The best math students in my school also all got A's in English. The best English students did not all get A's in calculus. Most of them didn't even take calculus. STEM has a higher ceiling than the humanities and that's reflected in the GRE data.
> STEM has a higher ceiling than the humanities and that's reflected in the GRE data.
You must have meant to say something different? The GRE data (last I knew) show philosophy with the highest average scores, and the ceiling on the GRE "quantitative reasoning" section is lower than the ceiling on the "verbal reasoning" section.
This doesn't make any sense. How could the gap be the result of human bias when it persists under automated testing (Hypothesis 2)?
Furthermore, writing code isn't the only part of the job, so of course interaction matters in the interview. It is very possible that these candidates are not doing as good of a job at communication. The author presents no evidence that their performance isn't simply worse.
(N=1) Well, spending ten years as tech lead of a 50 person software development group with significant input into offer/no-offer decisions, I can proudly assert that I would select female candidates over quasi-equal male candidates in most cases. Lived to regret that a few times. Of course we were in an environment where DEI was a contract metric. The big issue was getting sufficient female resumes with the right course work and good GPA.
Not going to completely answer that question but I was working for multiple government contractors building bespoke software for something like flight simulators.
I liked the idea of giving the underdog a chance. As long as it wasn't forced by management. Sometime there was a good reason that person was available in the job market. For experienced designers/engineers we went a lot by word of mouth.
It could perhaps be written more explicitly, but from the "Finding" bit in hypothesis 3, it looks like in testing hypothesis 2, they DIDN'T blind the interviewers. So the claim is that if there are two candidates with the same score, knowing it's one of them is a woman makes them less likely to choose her.
This fails upon noticing that HS and undergrad computer science classes, which have no entry requirements, are 80% male, the same gender balance as the tech industry. You shouldn't even need to say "conditioning on a collider" (ie the fact that they are in coding interviews in the first place) to know this has no relation to the question.
It's possible that, unlike the dozens of other studies in this area which all fell apart on closer examination, this one will stand and establish a small bias in blind coding review, which adds to the difference in interests which creates the original 80% skew. But I'm not even sure how such a thing could be measured in real life given the degree to which affirmative action makes women pass tech interviews more than equally qualified men (eg https://economics.yale.edu/sites/default/files/marley_finley.senior_essay.pdf)
It doesn't necessarily follow that 80% boys in HS CS classes means 80% men in coding positions is 'normal' or 'correct' or 'meritocratic.'
Take the ad absurdum case where CS is a required class for all boys and an optional elective for girls, and under that regime 1% of girls take the class, and after that hiring for positions is 100% meritocratic.
In that system, I'd expect far, far more than 1% of coding jobs to go to women. Why? Because the vast majority of people who take a CS class will not go on to get coding jobs, you are still selecting for the most interested and talented outliers among the set of everyone taking that class, and I'd expect the 1% of women who take it as an elective to be more interested and talented than the average boy who is forced to be there.
Now, take the mandatory element out of the thought experiment, and it comes down to what factors other than interest and talent might guide students towards or away from those CS classes in highschool.
For 80% boys in CS->80% men in coding to be obviously 'correct' it would have to be the case that there are *zero* factors besides interest and talent that affect attendance in CS classes of boys and girls differently.
If there is a social stigma against girls taking CS classes, if their parents and teachers and counselors are less likely to recommend them, if the material covered is less tailored towards their other interests, if the culture is less welcoming or engaging, then we should expect the average girl in those classes to have higher interest and talent (to counteract the factors pushing her away from it).
If there is a social approval towards boys taking CS classes, if their parents and teachers and counselors are more likely to recommend them, if the material covered is more tailored towards their other interests, if the culture is more welcoming or engaging, then we should expect the average boy in those classes to have lower interest and talent (since there are so many factors besides those two pushing them towards it).
If it's true that these factors exist and the result is the average girl in those classes having more interest and talent than the average boy in those classes, and that fewer than 100% of the people who take those classes will go on to get coding jobs, then a 100% meritocratic hiring process should lead to a higher proportion of women in coding jobs than girls in HS CS classes.
Of course, it then comes down to a hard-to-quantify argument about how *much* of the difference is due to sex difference in interest/talent, vs other social factors. Which is exactly where we all *started* this argument in the first place, just a bit more minutely modeled.
For my part, I've read enough evo psych to think that the idea of there being *zero* genetic sex differences is absurd, and I've met enough human beings to think that the idea of there being *zero* social factors involved is absurd. My position is that as long as we agree that both of these things are at play and explain some percent of the effect, it's worth continuing to explore and trying to fight the social factors, since they're inherently discriminatory and inefficient. But we should put more effort into studying them and measuring their magnitude and reacting in appropriate and targeted ways.
When people are allowed to self select for occupation, certain fields show a gender bias. (I think the study is from a Scandinavian country.) Women like to do more people type jobs and men more thing type jobs. (Coding is a thing type job.)
The problem with “fighting the social factors” is that it becomes incongruent for the fighters to ever accept that the right balance has been reached. So long as the distribution does not match demographics, there’s a remaining attribution to social issues. It’s much healthier, after removing the blatant barriers, to attribute gaps to free market, than to obsessively try to justify why the game is rigged.
Hence why I said "But we should put more effort into studying them and measuring their magnitude and reacting in appropriate and targeted ways."
'Blatant barriers' is obviously the new word to taboo here, where all the assumptions will get smuggled.
If you mean 'we should remove all the blatant barriers, and hopefully we can achieve that with another 50ish years of careful, dedicated work,' then I agree.
If you mean 'we should remove all the blatant barriers, and we achieved that 60 years ago and should stop digging', then I violently disagree.
I actually do believe this is a question which it's possible to obtain true knowledge about, and I actually do believe we should keep trying to do that.
'If we care about this topic, people will be irrational about it in favor of their side' is a statement that's true of almost anything touching on the realm of politics, and is certainly very annoying and wasteful, but it doesn't mean we should never try to do anything at all.
The result here is that interviewers appreciated interacting with men more and those interactions made them think men were right for the job regardless of actual skill and ability.
I think that which elective classes children take in HS is very, very dependent on the guidance and feedback they get from adults and peers in their life, and I don't see why these results shouldn't transfer ceteris parabis to people interacting with boys and girls.
Agree that you'd want to confirm that with a study of actual middleschool/highschool kids, but that study is a lot harder to do. I at least appreciate this one as an intermediate step.
> interactions made them think men were right for the job regardless of actual skill and ability.
If you mean regardless of general ability to excel at the job, then that's likely, but not the only possibility. It could be that the men were on average slightly better for the job for some reason other than coding skill. As a made up example, if the women interviewees had worse communication skills on average, that would only show up in the interactive interview and not the coding test.
As a former junior high school science teacher (briefly, because it was a career I wasn't well-suited to), I found that in 7th Grade, girls were more intellectually competitive than boys. But that dynamic changed in 8th Grade, when boys became more intellectually competitive and more likely to drown out girls when answering questions. My mentors said this was a dynamic that had been observed for years and not just at my school. I assumed, without any evidence, that testosterone was kicking in for boys. And girls who were intellectually gifted in 7th Grade started competing on social standing within their peer groups. Suddenly, it became very difficult for me to elicit responses from the girls, and I had to specifically call on the girls and tell the boys to be quiet when the girls were answering (it was quite striking)!
Not an accusation towards you specifically, but to get information from a teacher's perspective: Did you/other teachers you knew intentionally compensate for this change in presentation by recommending STEM electives/honors classes/harder colleges/additional mentoring/etc. for the girls who were being drowned out by boys in class but seemed to have the talent and interest, or did you go with the flow as those changes happened and mostly point those things at the loud boys who were asking for them?
Why should they? It seems like it's the families' job to determine their child's preferences and the teachers' job to respond to them. If a girl isn't temperamentally suited to outcompeting boys in high school then they probably won't be temperamentally suited to competing with them in the workplace. Gender norms generally exist to edify, not oppress. I'm all for encouraging someone to be assertive, but I think you err if you assume that all norm violations are good. On some level a girl being timid in math class is (at least potentially) a revealed preference that she wouldn't be happy with that life. Don't assume you know better then the children themselves.
Also I question the link between "socially assertive in class" and being good at STEM. I was both cataleptically shy in high school and excellent at math. You don't need to be outgoing if you can just ace the test. I suspect that the phenomenon of girls suddenly becoming timid in math class is really just downstream of them suddenly not being able to get the right answer all the time.
I was mentored to adjust my teaching style for 8th-grade students by giving girls more space to answer questions and being more encouraging toward them. At that point in educational history (ca. 1985), it was popularly assumed that the gifted would reach their own level. But even back then, the discrepancy in the number of women entering STEM fields was noted. It had also been noted that private girls-only schools did better at prepping girls for STEM. But no one wanted to take the step of segregating the classes by sex in public schools (because of hypothetical socialization issues). And at my junior high school, there were no electives. All students took the same courses — though there were special ed classes for students with learning disabilities (IIRC, the educational establishment were aware of dyslexia and ADHD at the time).
This study doesn't say anything about the proportion of women being accepted or are graduating from CS classes. This study aims only to understand why women with equal skill sets are hired in lower proportions than men.
Do you have any evidence of women being underrepresented? I thought the proportion of women in tech was roughly equal the proportion of women who take an interest in it? If you are assuming it should be 50/50 or something, then you need to give some justification of why it would be that way.
If you were just saying "Women have been traditionally underrepresented in the tech sector" without meaning to imply that this paper had anything to do with that, then fine.
That's consistent with my N=1 longitudinal study to the extent they both overlap. I've noticed pretty close to zero difference between how I'm treated in professional contexts before and after transitioning, which affirms the study's findings on 3. 2 is not tested by my experience. 1 is only tested for hypothetical ability differences based on current hormone levels, not based on hormone levels during childhood or fetal development, nor downstream of sex-dependent genetics through other mechanisms. As for 4, my voice and appearance have changed, but I don't think my communication style has changed significantly.
Caveat: about 2/3 of the people I work with already knew me pre-transition well enough to have formed opinions of me and my abilities. The newer team members are likely influenced to some extent by the attitudes of those who have been on the team longer, even if they themselves have only known me as a woman.
Generalizing population-level trends based on individual anecdotes is questionable, and generalizing from the experiences of trans women is very questionable. Soon-to-be-trans-women individuals are not treated the same as the general population of men on average, and trans women are not treated the same as the general population of women on average.
Makes me think that there should be more research that studies trans women in such a way that at least some perceptual gender issues get resolved while still looking for mental differences. Then again the hormones themselves change your conscious experience (as far as I’ve heard), so maybe it’s not possible.
If it wasn’t so controversial, it should be relatively easy to show one way or another whether trans women have an athletic advantage. So maybe there’s something similar about STEM.
Another issue with that line of research is that people who choose to transition are heavily self-selected to have at least some mental traits that are very different from the overwhelming majority of our cisgendered counterparts.
Have you interviewed for a new job since transition? The paper is only about the interview process.
Of course it's plausible that if interviewers have those biases, everyone else does too, and you should see the effects in your day-to-day work. On the other hand, I think it's also plausible that people who actually do this work every day will be less biased than interviewers who don't... not sure where that falls out, without further study.
In my experience in the tech industry, interviews are done by hiring managers and intermediate-to-senior peers of the position being hired, so there's are huge overlap between "interviewers" and "people you deal with regularly on the job" and it's reasonable to expect them to have approximately the same biases.
To the extent there's a difference in biases between interviewers and work peers, I would expect it to be based on what they have to go on to judge you. As an interviewer, I'm judging someone on a single 40-60 minute conversation in which I ask them some probing questions about their background and give them a couple of isolated problems to work through. As a coworker, I am working together with them on overlapping projects, reviewing their code, talking to them in meetings, etc. It's definitely plausible for style considerations to matter much more in the former situation than the latter.
How can I better understand China and the China-US/West economic and political landscape and relationships?
I'm reposting my comment from last week, because I'd like to have a better model.
I have a bit of epistemic learned helplessness on China. On one hand, you have China-Doomers like Peter Zeihan constantly predicting China's collapse. Or that China has 50% youth unemployment. It's patently false. China has advanced a lot in GDP, living standards, infrastructure, science (genuine publications in top journals) etc.
On the other hand, I read things from Steve Hsu, Karl Zha, and certain subreddits (less credible defense) and it's like China can do no wrong. They can cheaply mass manufacture anything in a short period of time. Any challenges they face have all been planned for and will be well dealt with, especially compared to the sclerotic west. The governments are actually more responsive at the local and national level to their people compared to western ones, etc.
Some examples: that DeepSeek was basically a weekend project for some hedge fund workers who used a box of scraps to best multi-billion dollar American companies. Outdated Chinese jets are better than French SOTA ones, China can build a hospital in a weekend, and that China will eclipse the west in all sciences and manufacturing in a few years at most. It seem to stem from some HBD thoughts (average population IQ being ~103 and having 1.4 billion people), "first principal" thinking and biases for STEM-types. I.e. that China is lead by engineers while the west is lead by "retards", "midwits" (their words) "wordcel" lawyers, and the west faces existential crises that they will not be able to avoid. There's a whiff of central planning is correct if high enough IQ people do it.
I enjoyed reading When China Rules the World by Martin Jacques. Apparently the books was also popular in China itself and I've seen endorsements from both pro and anti-China sources.
My basic mental model of China is that their government has one of those really useful types of government you only get in the endgame of Civilization where you can drive the economy to focus on exports, scientific development and military development while keeping the population relatively quiet and peaceful. Economic growth has slowed, youth unemployment is way up, but they keep making progress, in part because NIMBYs and protestors are suppressed so it's quick and easy to build out high speed rail, drone delivery systems, driverless cars, etc.
In the Civilization game, this is a surefire winner because you just have to be able to build the Alpha Centuri probe or whatever, but in the real world there is no finish line, unless it's building superhuman AGI.
There’s a great Substack called sinification which has translations from Chinese academics, economists and political scientists, on China itself and its relationship to the West, and the US. The debates are more robust than you might think - which itself might challenge some preconceptions.
It seems like you keep looking towards Thought Leaders for interpretation of information when you should just go directly to whatever source of information you want to know more about.
I don't speak Chinese and am looking for "middle men" I don't know if that's the right term, but rather sources of info that will be truthful and informative .
My impression has been that people are too focused on finding a 'winner' between market capitalism vs central planning, and too allergic about admitting that each has strengths and weaknesses we should be codifying.
Like, I don't think either of these are true at the moment, but I don't find an inherent contradiction between 50% unemployment and the ability to completely dominate in specific tech fields that the central planning committee decides to focus on.
That's directionally the sort of thing I'd expect form a planned economy, diffuse inefficiencies where the central committee can't optimize every single job and market in the nation all at once, impressive feats where they do decide to focus the entire nation's interest and resources.
This guy makes a pretty strong argument that China has been using various policies to forestall an economic depression, and that the longer they do so the worse it is going to be when it finally happens (https://substack.com/home/post/p-181489135). He seems to make a good argument.
I have definitely noticed that every time China is criticized, or when someone posts something skeptical of China's economic or military power, there are a lot of very aggressive China defenders in the comments.
I also agree that some people are very fixated on the idea that authoritarian governments can make good choices that free markets and democracies can't. I disagree with that idea strongly, but a lot of people seem very convinced of it. But the same people were saying that about the USSR back in the 50s and 60s, and about Japan in the 80s. Free markets are pretty robust, and putting your fingers on the scale of the economy is bound to create waste and inefficiencies.
I'm certain that a command economy and authoritarian government can make powerful and quick decisions that are hard for democracies to follow. I'm also certain that they can't know which decisions those are and which decisions are catastrophic and fatal until it's too late for them.
I won't argue that they are necessarily *good* choices but there is indeed a choice or two which seem to be much easier for authoritarian governments to implement.
An economy can become more competitive in international trade by lowering its labor costs. This can also be done in various ways by free market democracies (e.g. deregulation of labor laws) but having its workers be worse off does make it a bit more difficult (though far from impossible) for democracies. And excluding trade, an economy can reduce consumption in order to increase investment (in production capacity, R&D); which again can be done with various means (e.g. tax policy) in free market democracies, but as people really don't like reducing consumption, an authoritarian command economy like e.g. Stalin era USSR could and did restrict consumption *much* more, in order to instead focus on building infrastructure and manufacturing capacity; a free market economy couldn't ignore consumer desires to this extent.
> I have definitely noticed that every time China is criticized, or when someone posts something skeptical of China's economic or military power, there are a lot of very aggressive China defenders in the comments.
I have noticed is that there’s an increasingly large number of people who are sceptical of the Chinese sceptics, who don’t have an impressive track record. Nothing very aggressive though.
China has segued in two years at most from being doomed to dominating EVs, battery technology and what have you; but now now is doomed for real this time. Maybe, but probably not.
The Chinese government on mixing public planning and private enterprise is probably more pragmatic than westerners who have decided that all government involvement is doomed to failure. Our governments are full of practical men, who believe themselves to be quite exempt from any intellectual influence, but are usually the slaves of some defunct economist or other.
This response is kinda what I'm talking about when I brought up Steve Hsu et al. I appreciate your pointing to the Sinification blog and will definitely check it out more and maybe it will answer my request for a better model.
However, you highlight China's accomplishments in EV which is great for new types of cars and affordability, while simultaneously depreciating western governments as being dumb and beholden to "defunct economists" and that the Chinese gov is more "pragmatic".
It just feels like there's just no good critiques of Chinese gov policy, like this is just an update of the "inscrutable Orient" to "Pragmatic and always right China".
Good! I'm glad you're finally coming around to reality. Just kidding. I didn't mean to annoy you. Aside from a few lines of text, I don't really know your opinion on my OP question. Maybe I'm projecting.
As for my opinion. (not directing this at you), but in general when I see claims about how X is all in terrible and there's little to nothing good about it, or that X is fantastic and there's absolutely nothing bad about it, it pings my skepticism-dar on both ends.
Regardless, I'll read the blog you recommended and look at some other sources.
So, for the last couple of years, I've been getting into beekeeping. I'm keeping(pun not intended) it small, only having two hives on a small green patch in the city where I live. I appreciate the combination of being knowledge based, outdoors, relatively little investment(both time and money) and not being a sport (enough of those hobbies already). Following the seasons and plants more closely has given me joy. As Henry David Thoreau put it: "The keeping of bees is like the direction of sunbeams". It is also a surprisingly deep and engaging topic, touching upon many topics such as genetics, botany, light, weather forecasts, entomology(obviously), etc. Because of all the variables involved, many questions have no definite answers, and, as for many other things, "if you ask 10 beekeepers, you will get 11 different answers".
Beekeeping also has celebrities like Karl von Frisch and Brother Adam, and legends like purple honey in North Carolina, or tales of the times before the Varroa Destructor mite was spread around the globe.
Anyway, sharing it here, since I have found that the topic engages broadly, and many people have exciting anecdotes. Would love to hear them.
Any suggestions on good books or other sources on the topic is also interesting, since filtering out the noise can be hard. My suggestion to others who are interested would be to check out Dave Cushman's website.
We don’t keep bees, but we grow flowers and other plants in our garden that attract them, and we now get many different species of bees and wasps in the summer months. Sitting close by and watching them could easily take up the entire afternoon and evening if we didn’t have other stuff to do—it is one of our greatest joys.
Sounds amazing. I do this in public parks too, but don't have a garden. Are you keeping count on the species? In my country there are around 200 different solitary bees and bumblebees, so plenty of species to identify.
We don't keep count, but I would estimate we've probably seen over a dozen species of bees alone in a single day. I've read there are 300 species in my area.
I lived on a commune part of the time when I was in college, and someone there kept bees. Once the whole hive decided to move. I’ve forgotten now what the likely explanation was, something to do with the queen. In prep for leaving they hung in a huge cluster several feet in diameter from the branch of an apple tree. The main beekeeper climbed a ladder and somehow got them into a garbage bag and returned them to the hive. Wow.
Thanks for sharing, I wouldve also been wowed by that. I haven't seen a swarm yet. What they teach in courses is that the bees can swarm when the queen is old or stopped laying eggs, space is tight or disease. Everyone always has a good explanation after-the-fact, but preventing it can be hard. I will sign up as a swarm collector with the association this summer, and hopefully receive some calls. Collecting it is all about finding and moving the queen, the other bees will follow her smell.
My cousin keeps bees (in the southern UK) and, according to him, varroa mites aren't beekeepers' only headache these days. Asian hornets are a new threat. Apparently a queen hornet starts a small nest in a hedge or similar early in the spring. Then, before long, when the nest reaches a certain size, they migrate to the top of the tallest tree they can find and start building their main nest. That makes destroying their nests, or even finding them, quite challenging.
Monks, such as Brother Adam of course, who you mentioned, make good beekeepers for various reasons. One is that the monks have a bland diet, and bees can sense their keeper's diet via the emanations from their breath and even skin. After eating just porridge and similar for a few days, the keeper could gently dip their hand into a bee swarm and locate and extract the queen without exciting the bees. But I imagine trying that stunt after eating a strong curry the previous day would not end well!
I have also seen many beekeepers voicing their concern regarding the asian hornet. I didn't know they build in the treetop. Sounds difficult, like you say. I have heard that beekeepers in Germany switched from 'hating' the european hornet, to loving them, since they are one of the asian hornet's only natural enemy(in europe). In comparison to bees, they are also quite big and nasty. Where I live they have not been sighted yet, but will probably arrive in the next decade.
I never thought about how Brother Adam smells, but that's a fascinating theory! I have been taught that strong smells annoy bees (dont use deodorant), but never extended it to diet. Maybe I will try out a bland and/or spicy diet myself and see what happens!
How accurately do we actually know population statistics (births, deaths, causes of death, age distributions, migrations)?
I know that in the US there's the Census at least every decade. But even that's got to have significant error bars. Smaller countries in Europe with population registration (where residents have to register their address with the government) can probably do better. But places like sub-saharan Africa, where records of any kind are at best faulty and/or there are long-running civil disturbances? Or chunks of China where the local government has every incentive to fudge numbers in all sorts of (probably mutually contradictory in cases) ways?
This is all part of why I'm so skeptical of foreign aid claims in many ways. "X lives saved by Y intervention", "War killed Z people"....can we actually measure that with any accuracy? When all the incentives for *everyone* involved point toward fudging the numbers to inflate X or push Z to something more palatable to the commenter's side? Even if people are trying to be honest (a very bad assumption in general), all it takes is a tiny systematic error repeated a lot of times to get very large (relative to the underlying numbers) overall skew. This is compounded in the international aid field (relative to the war side) since in war there are two sides trying to fudge numbers in different directions. But in international aid, all the incentives for all the researchers, governments, and organizations involved are to inflate the usefulness of the aid and maximize the reported size of the problem.
It may still be worth doing, but I'm struggling to see how we can get below "at least one order of magnitude" error bars.
I think that in Western countries (or at least European ones, I don't know as much about Anglosphere) the number of births and deaths is extremely precise. People don't die unnoticed, and babies don't get born unnoticed.
For causes of death, this is pretty precise with the categorization system that is used. For example, a death certificate in Germany may list "Pneumonia caused by Stroke", where both Pneumonia and Stroke are official categories in the German list of deaths. In some cases where the doctors are not sure, they can also list more than one reason (multimorbidity). I think misclassification is pretty rare. In general, the resulting numbers are very precise, and you can answer any question within the classification system.
Of course, you can't answer questions outside of the classification system. If you remember back during the pandemic, people got super-upset because it was impossible to get official numbers of people who died *from* Covid. Those would be listed to have died from pneunomia (for example), because it is too laborious to decide in every case whether the pneumonia was caused by Covid. Those are questions that one cannot answer from the official records.
The solution to all this is statistics, of course. This was why we were sure during Covid that most people dying "with Covid" also died "from Covid". Both by looking at the total number of people who died, but also because some studies took the effort to examine a few dozen to hundred people who died "with Covid" and investigate how many of them died "from Covid". (Some countries like Denmark even did this for all deaths.) A similar thing you can do with interventions in Africa. You can take two regions, distribute Malaria nets in one region but not the other, and see how many people die from Malaria in both regions.
Finally, for the census in Western countries, the reason for the error bars are not that people die unnoticed, but it's that people move unnoticed. If a person leaves Germany and moves to a different country (or even a different district within Germany), then they should tell their old district that they have moved away. But in practice, not everyone is so conscientious to tell their district. There is no automatic update with the new country, so those people remain falsely in the files. In the US, it seems that unregistered immigration is also a big factor: people moving to the US illegally who don't register. In Germany, I think this is not a big factor. For the magnitude, the last census in Germany was in 2024 (after ten years), and it corrected the population in Germany from 84.1 to 82.7 million, a correction of 1.6%. For the people in the files, the age is also usually correct, so the age distribution is also very well known except for the same 1.6% of people who are falsely registered.
In the US, birth certificates and death certificates are issued and tabulated on the county level. and reported upward to the NCHS. Parents have every incentive to get their kids registered, otherwise it would be hard to establish identity for all sorts of IDs required for modern life. Death certificates are handled by counties, and they must include a primary cause of death plus any contributing causes of death. They get sent to the NCHS and CDC.
Census is supposed to count the remaining folks in our country who didn't start out with a birth certificate here (or earn a death certificate).
The EU has promulgated uniform birth and death reporting standards for its members. I don't know about censuses. And cause of death codes are now a universal standard format. So we know how many people are dying from which diseases by country. Western countries keep pretty good counts. Developing countries less so. But in the modern Western world, birth and death numbers should be pretty accurate.
China is kind of an interesting case. We're pretty sure that CPC keeps good numbers. They're reported upward from the municipal level. But they don't necessarily report those numbers to the rest of the world. Officially, China has ~1.4 billion people. But demographers have done the math and given their reported population 30 years ago, and their falling birth rates, it's difficult to believe they reached 1.4 billion. And energy usage and grain consumption (which are reported) suggest a significantly lower population. Some demographers place them between 900 million and 1.1 billion, with many leaning towards the lower end of that range.
Regarding China, I have known that the official numbers are implausible, but the only concrete claims that I have heard (with substantial justifications) were around 1.25 billion. Do you have a link for the lower numbers? Even 1.1 billions would be a lot less than I thought, and 0.9 billion would come as a huge update to me.
I'm going to have to fade on this one. This was a subject I was only passing interested in, and I didn't pinboard or bookmark the references that I dug up. Several were from anti-China or anti-CPC sources. So, I'm not going to die on that hill. I did locate this paper, which claims China's nighttime light output by satellite suggests a much lower GDP (which I concluded indirectly supports the lower population estimates). But this is from one of those biased researchers I mentioned...
My number comes from estimates by Yi Fuxian, a demographer who has investigated inconsistencies in Chinese official numbers. For example, the number of now-18-year-old does not even remotely match the number of newborns 18 years ago. He has dug out a lot of official numbers, like the number of vaccinations, the number of school children of different ages, and so on, to reverse engineer an estimate of the correct numbers.
Yeah. I'd expect that the errors in developed Western nations (EU+UK+nordic, US, Canada, Australia, NZ mainly) are pretty good but still subject to errors on the level of sub 1 percentage point.
I believe that in the case of Paraguay the previous overcount was produced by the estimation that had to be done to account for those who didn't/couldn't be counted in the census (2012). I don't know how the estimation was done, but I assume that some of its assumptions might have been invalid:
"For instance, in Paraguay, the 2012 census may have missed a quarter of the population. The significant underrepresentation of rural population we found in the gridded datasets raises the question whether such gaps in the census are more widespread than assumed to date, and how reliable current global population estimates really are." (Global gridded population datasets systematically underrepresent rural population, Láng-Ritter, Keskinen & Tenkanen, https://www.nature.com/articles/s41467-025-56906-7).
- Regarding census coverage: "[The president of Paraguay's National Statistics Institute (INE)] affirms that census coverage was 94%, and therefore he trusts the data"
- Why was there such a large gap between prior population projections and the results of the 2022 census: "The president of INE, Iván Ojeda, explained that the data is surprising because Paraguay had been relying on projections for 20 years, following the failed 2012 Census."
- Where does the 94% coverage figure come from: "He said that a post-census survey confirmed that 94% of the population was counted."
- Other data was contrasted with census results: "Furthermore, the Single Student Registry was consulted and it was found that there are 906,000 students registered with the Ministry of Education, aged 7 to 15, a number that coincides with that obtained in the 2022 Census."
- Related to what Beowulf888 mentioned in another comment (births and deaths): "When consulting birth data from recent years, they observed that between 80,000 and 100,000 births are registered annually, while the projection was 140,000 babies per year."
And I'm not necessarily claiming *intentional, malicious* fudging. Could be unconscious, could be making *justifiable* decisions that all end up pointing in the same direction, etc.
Donut Lab is a Finnish company that just unveiled a solid-state battery at CES that is making waves in engineering circles. Their claims are either revolutionary or a complete scam.
Their claims:
* 400 Wh/kg (vs ~250 Wh/kg for the best current Li-Ion batteries).
* 5 minutes fast-charge (vs ~8 min best current Li-Ion, at the cost of battery degradation).
* Fast charging with no degradation at all (major problem for Li-Ion).
* Designed for up to 100,000 charge cycles (vs ~1'500 to 2'500 best current Li-Ion).
* >99% capacity retained from -30 to +100 C° (problem for Li-Ion).
* Made of globally abundant materials (major problem for Li-Ion).
* Extremely safe (fire safety is a major problem for Li-Ion product design).
* Currently in production, shipping to customers Q1/26 in a motorcycle taking pre-orders now.
* Price point on par with Li-Ion.
* Easily scalable production techniques, already scaling up massively.
If true this will be a major breakpoint for electrification in general, especially for EVs. They would certainly be a big improvement at my job (airforestry.com).
If a scam, they are very scammy indeed. If they manage to keep up the act for long enough, I worry many genuine battery start-ups will be de-funded in the meantime.
Points against:
They appear to have no registered IP, at least not that I can find (I might be doing it wrong):
And the CEO is also a co-founder of the motorcycle company Verge, that is allegedly shipping with this battery to customers Q1/26.
My current estimate is 60% on outright scam, 25%: There is a real product but it is being massively over-sold to the point of not revolutionizing the industry, and 15% on this being generally real as presented, but still allowing for some hype.
Does anyone here have a different take? I'd be very interested to hear!
Mhh I got very excited but immediately turned very skeptical. They have no technical sheet or better information about any product specs.
I understand industrial secrets and all, but they don't even mention what "green" materials they are supposedly using, which would be a great point to underscore how green the batteries are. Instead they provide a single image with points and no labels.
I'm no geologist either, but the distribution of the points looks unbelievable: assuming they represent ore deposits, it seems unlikely that Europe would be the densest and places like Kazakhstan or other parts of Central Asia would have like four or five lights.
The CEO is apparently a "Tech strategist and serial entrepreneur with proven track record", but does not bother to list his past accomplishments?
All in all, there seems to be very little technical material and a lot of publicity.
Also, maybe this is normal in Finland, but isn't it weird that the website for a Finnish company would have no Finnish language version or anything? I get it's aimed at an international market, but it gives me the impression that they are trying to hit it big without bothering with the fundamentals; this is perhaps the norm in the tech startup scene, though.
Re: webpage in English, I looked them up online and they seem to have an address in the UK?
Address: Unit 6 Cornbrash Pk, Chippenham SN14 6RA, United Kingdom
Perhaps that's a manufacturing base or something.
The CEO has a Finnish LinkedIn but, uh, his work experience seems to be mostly/mainly "co-founder of this, that and the other company".
He has a degree in Computer Science from the University of Helsinki and yeah, I have to say, this is looking more "Elizabeth Holmes" than "Elon Musk" here (depending, of course, on your opinion of Elon).
How likely is it that a 16 month old company is ready to deliver a world-changing technology at the consumer level?
The address is registered through Domains By Proxy LLC, which according to wikipedia, has a scammer problem. Although by the tone of the article, it doesn't seem like the same kind of scammer that Donut Lab would have to be.
The CEO's CV you posted below really cements my view that the battery has some undisclosed fatal flaw, or is straight-up nonexistent.
Today I learned that the company I work at has a Swedish version of it's webpage 😂
Having a page only English isn't weird at all, and is true for a lot companies here in Sweden, and I'd be surprised if it's different in Finland. I don't count that as evidence either way.
As to the secrecy, I'm torn. If they released more specifics and it all seemed congruent I'd definitely be more convinced. But in the world where it's real, they don't have to optimise to convince me, only investors and partners behind closed doors where the IP is safe. And of course they will be copied to some extent regardless of patents, so every breadcrumb they leave out reduces their moat.
But of course if it is a scam they might not have anything to show. But then I'm confused why they wouldn't make up something more flashy? If it is a scam they are quite well calibrated to extract maximum credulity from minimum evidence from me at least. Is this in an of itself evidence for or against?
I realise I'm going off impressions rather than data here, but going by the CEO's LinkedIn, he's very proficient in bullshit tech vibe buzzwords. This man has done it all! (I'm getting rather fond recollections of our Hexayurt friend from this CV):
"Board Member
Cova Power
Oct 2025 - Present · 4 mos
Cova Power makes the world’s first series-production electrified smart trailer conversion, reducing fuel consumption by 54% while dramatically reducing the overall cost of truck fleet operations.
Co-Founder and Chairman of the Board
ESOX Group
Oct 2025 - Present · 4 mos
Esox Group takes the Western military capability to a whole new level in the most critical area of modern warfare: Unmanned Autonomous Vehicles and intelligent military EVs.
ESOX Group has the most capable hardware technology platform for building next-gen aerial drones, unmanned ground vehicles, submersive drones and other modern EVs — including what we know as Donut Motors, but in military specification — combined with cutting edge unmanned vehicle software and AI-driven digital battlefield simulation platform.
Together these technologies allow the creation of drones and other UxVs with groundbreaking payload capacity, speed, maneuverability, range and agility never before possible, combined with onboard AI capability that outperforms the current state-of-the-art technology by an order of magnitude.
Co-Founder & CEO
Donut Lab
Nov 2024 - Present · 1 yr 3 mos
Reimagining the electric mobility industry by building the world's first universal EV technology platform with next-gen software development tools.
Advisor
Aito Ventures Oy
Feb 2025 - Present · 1 yr
Aito Ventures is renovating the trillion dollar global renovation industry.
Verge Motorcycles
5 yrs 2 mos
Co-Founder & Advisor
Nov 2024 - Present · 1 yr 3 mos
Chairman of the Board
Jun 2022 - Present · 3 yrs 8 mosJ
Co-Founder & CTO
Full-time
Dec 2020 to Nov 2024 · 4 yrs
Verge Motorcycles develops and manufactures the world’s most advanced electric motorcycles. Verge TS Ultra is the most powerful motorcycle in the market today with the longest range, fastest charging time and best riding experience. In addition, Verge enables other motorcycle manufacturers to build better vehicles with its advanced technologies in the areas of motor, battery, electronics and software. As a CTO I was responsible for Verge’s software division. Embedded software, machine learning, connected vehicles, machine vision, HMI, HUD technology, PCBs and microchips, cloud/SaaS services, and all that jazz.
Board Member
Dec 2020 - Jun 2022 · 1 yr 7 mos
VERGE Motorcycles is a manufacturer of electric motorcycles. VERGE TS, the world's first Hubless Electric Roadster, has received numerous awards and recognitions for its innovations and design.
Co-Founder & Chairman of the Board
Stingbase
Jan 2024 to Present · 2 yrs 1 mo
Stingbase develops AI-driven investigation and crime fighting platform for law enforcement agencies and private cyber investigation units. Stingbase's cutting-edge platform revolutionizes criminal case resolution and increases success rates by over 100x.
Stingbase offers a comprehensive end-to-end solution, equipping users with advanced tools and features to seamlessly manage cases from crime-to-conviction. The platform's intelligent design minimizes the risk of human error, ensuring precision and efficiency at every step.
The solution enhances cross-border collaboration and promotes public/private partnership model to take international crime fighting to the next level.
Co-Founder & Chairman of the Board
ASILAB
Jun 2023 - Present · 2 yrs 8 mos
ASILAB builds artificial superintelligence and synthetic life-form that is more capable than humans in every aspect. Our approach differs significantly from everything else in the market, with built-in evolutionary principles, self-learning, synthetic neuroplasticity, observability, life purpose, long-term memory, and more.
We're building independent thinking organisms that develop on their own, continuously. And boy have we patented it.
Founder, Chairman of the Board
Shadow Capital
Oct 2022 - Present · 3 yrs 4 mos
Shadow Capital invests in cutting-edge deep tech companies.
BURST
4 yrs 1 mo4
Chairman of the Board
Oct 2024 - Present · 1 yr 4 mos
Founder, Advisor & Investor
Jan 2022 - Present · 4 yrs 1 moJ
Burst develops next-gen immersive music and entertainment experiences utilizing advanced AI and latest technologies. Currently Burst is working on the world's first hip-hop game for mobile platforms, which will be later on expanded to photorealistic VR.
Venture Bonsai
16 yrs 9 mos
Co-Founder & Chairman of the Board
Jan 2025 - Present · 1 yr 1 mo
Co-Founder, Advisor
Jan 2013 - Jan 2025 · 12 yrs 1 mo
Co-Founder & CEO, Member of the board
May 2009 - Jan 2013 · 3 yrs 9 mos
Venture Bonsai is the first equity crowd-funding platform in the world, pioneering the industry. The idea of Venture Bonsai was born after I exited my previous company and was looking for startups to invest in. So we built the platform and the rest is history.
If you’re building a world-changing deep tech company and need funding, we may be able to help.
Co-Founder & Chairman of the Board
KaseyOS
Mar 2024 - Present · 1 yr 11 mos
The team that pioneered the entire no-code software development industry is making a comeback. This time we’re building the world’s first AI-native no-code platform, and it’s going to change everything. Coming out of stealth in Q1/2026.
Official Member
Forbes Technology Council
Oct 2015 - Present · 10 yrs 4 mos
Founder, CEO, Chairman of the Board
Kasey Group
Mar 2007 - Present · 18 yrs 11 mos
Kasey Group has a long history in developing bespoke mission-critical enterprise software, such as cutting-edge modern ERP and manufacturing systems for automotive industry, modern learning systems, flight and leg management systems for the aviation industry and A.I.-driven cyber security solutions.
DeFinance Technologies
3 yrs 8 mos
Board Member
Jan 2023 - Sep 2025 · 2 yrs 9 mos
Advisor & Investor
Feb 2022 - Sep 2025 · 3 yrs 8 mos
DeFinance develops AI-driven crypto and stock trading platform.
SAP
Full-time · 2 yrs
Lead Product Manager, SAP Build Apps
Jan 2022 - Jan 2023 · 1 yr 1 mo
Director & Team Lead, SAP AppGyver
Feb 2021 - Jan 2022 · 1 yr
Changing the world with no-code, continuing the pioneering work we started at AppGyver. We call this next generation of no-code ”visual programming”, as it is equivalent to writing code in its power of expression. With turing [sic] complete visual programming there is literally nothing you couldn’t build — from mission-critical enterprise software to 5-star omni-channel consumer experiences.
Founder & CEO
AppGyver
Mar 2011 - Feb 2021 · 10 yrs
AppGyver pioneered no-code software development in 2011, building a global community with hundreds of thousands of developers. In 2020 the company took the industry by storm again, creating the world’s first turing [sic] complete no-code platform — enabling development of mission critical custom software at scale without writing a single line of code. In 2021 AppGyver was acquired by SAP, and the product is known today as SAP Build Apps.
Co-Founder & CTO
FansMagnet
Feb 2011 - Jan 2013 · 2 yrs
FansMagnet was a visual app development platform to enable artists to create fully customized mobile apps for fan base management and communication.
Customers included Warner Music, EMI, Universal etc.
Co-Founder & CTO
TRV Networks
Jan 2001 - Nov 2006 · 5 yrs 11 mos
TRV-Networks did business in areas of System Integration and Software Licensing, as well as international trade. Main business was sold in 2006."
Re the website: that's fair, I already suspected it to be not very relevant either way. But the extreme generic character of the information page is; this is a product that goes into high-end motorbikes i.e. consumer products. No specs, no demos, no nothing?
Also, I understand the "moat" part, but this is going into consumer products that are allegedly "available now", so hiding even the names of the chemical elements used in the battery seems unreasonably secretive
Yeah, I see your point. In the interview the CTO says himself that while he won't go into details, it won't stay secret for very much longer since it will soon be in the hands of any company that really wants one anyway, and several independent reviews/teardowns will probably be published shortly thereafter.
If it was a scam though, wouldn't they try to extend timelines more than "available now" and "shipping Q1/26"? I mean, this way we won't have to hold our breaths for very long. Any excuse to delay now would be highly suspicious, and they'd be hard pressed to make a clean exit by March. In many ways it would be *more* believable if it was Q1/27.
To my eyes they manage to exude the confident calm of someone who has nothing to hide, but also no reason to tell you just yet. But of course that's what they would like me to feel in either timeline.
I was hoping someone versed in the prediction markets would find me a link! Cheers :)
It would be way harder/riskier to partner with an independent company if you are running a scam, whereas if you run both companies the increased risk is very small. OTOH, if I knew ahead of time that an amazing EV battery would come to market (because I was in the process of inventing it), I might very well start a side business trying to get a head start on some of the use cases. If it was an independent company, this would be real evidence in favour: As it is I don't think the motorcycle is much evidence either way.
It would seem like a massive conflict of interest at the very least: "here's this revolutionary battery and oh yeah if you buy one of our bikes, we'll ship it with this"/"if you buy our bike, as distinct from a competitor's bike, you'll be one of the first to have a bike running on an amazing new battery!"
Too late, when you've bought the bike, to find out the hyped battery was only hype.
Apparently this New York state family has an unaffordable energy bill of $1800/month. But that makes absolutely no sense to me. How can the bill possibly be so high? Electricity/gas cannot possibly cost that much per kwh. And why can they not reduce their energy consumption if it is that expensive? Is $1800/mth really the minimum amount a family in New York would need to pay to cover essentials?
I found it odd that their only option was apparently to cut the supply altogether. But really, none of it makes much sense to me, and I'm wondering if this is just some exceptional case reporting on a person apparently devoid of common sense, or if this is actually normal for some reason?
My bill once tripled while living in ny. The electric company started estimating my usage instead of measuring it every month. The estimates had no relationship to my past usage. Id eventually get the money back when they measured, but it was very annoying and did not seem to be in good faith.
It's just the shit state of modern journalism. You can't just say "There's been an increase in energy prices, here's some statistics", you have to lead off with some random individual.
It just occurred to me that this is the same basic phenomenon as how damn near every youtube video now has a shocked-looking face in the thumbnail.
I cannot read the article so I can’t address it directly but I can tell you that I have a house in the Northeast United States (100 miles NE of New York City) and our mode of heating is propane. Propane is very expensive right now compared to last year (currently $4.35 per c/f. We are paying 600 - $800 a month for propane. We cook and we use hot water so it’s not all heating. Our house is not weathertight; it could be much more efficient than it is, but it’s not absurdly leaky. Also, I keep it at 52°F for five days out of seven during the winter because we’re not always there. The big mistake was putting in picture windows. I always thought the windows of houses around us were very small but now I understand why. But when the sun is shining, it’s such a nice thing to have a picture window.
So part of the question is that, as I understand it, different locales can have very different and idiosyncratic and complicated pricing regimes. Electricity is not bought like a commodity at a market-set marginal price, it's bought like a service with very complex terms and conditions.
In particular, I think a lot of places have different types of 'surge pricing' that can produce absurd results. Prices that increase for each marginal watt used over the set minimum, prices that increase based on time of day, prices that increase based on grid-wide demand. These factors can lead to huge differences in price based on very small differences in total usage.
One example is the Texas winter storms that knocked out parts of the grid, leading to algorithmic surge pricing based on supply vs demand sending normal households outrageous bills.
It says gas and electricity, and it's a family of four (her, two kids, granny) so they could be running all kinds of devices day and night as well as the heating; looking up annual temperatures it says Greenwood Lake (where she lives) has cold, snowy winters and warm, wet summers so if she's running the heating constantly that mounts up. Depending on who the utility company/companies supplying her house, there might indeed be steep price increases:
"Winter heating costs are expected to jump 9.2% this season, according to the National Energy Assistance Directors Association, driven by rising electricity and natural gas prices and unusually cold weather.
Energy bills tend to be among the highest in the northeast US, the report shows. But households from California to Georgia to South Dakota are also feeling the strain of rising costs over the past year."
Everyone has their own phone, computer, TV in the bedroom, etc. etc. etc. that are always on (nobody, including myself, unplugs TV sets anymore), devices are charging, everyone is having multiple showers, microwave and oven and stove are going all the time, kitchen appliances, washing machine, dryer - you'd be surprised how usage creeps up, and then slap on peak pricing and there you go.
Mind you, I don't necessarily believe the price quoted, but it's easy to be running more devices than you think you are. The story says the bill tripled, so that would indeed be a hefty price jump. Since we don't know the power company, we have no idea of what pricing it charges.
"Greenwood Lake electricity rates vary, but the average residential rate is around 22.52 cents per kWh, with a monthly bill averaging $143.65.
What is a good rate per kWh in Greenwood Lake?
In Greenwood Lake, New York, a good electricity rate falls between 22 cents to 24 cents per kWh. Tired of high bills? Shop during spring and fall to lock in lower prices before HVAC use spikes in the summer or winter.
What are electric rates in Greenwood Lake today?
Greenwood Lake electricity rates currently average 22.52 cents per kWh, but you'll likely pay between 21 to 25 cents depending on your provider, location, and plan type.
Why is electricity so expensive in Greenwood Lake, NY right now?
Your Greenwood Lake electric bill can be expensive because you live in a high-demand region with complex delivery systems. You pay premium rates due to natural gas costs, peak usage timing, and aging infrastructure that needs constant upgrades."
I don't know if 638 kWh per month is on the low side, but if her previous bill was $600 per month @ 25 cents/kWh, that's 2,400 kWh/month. That's hefty usage!
Now, there is a power company listed in Greenwood Lake that charges 44 cents/kWh, so if she's signed up with them, then yeah that makes it more reasonable:
$600/$0.44 = 1,364 kWh/month. Maybe her house is old and needs a lot of heating?
"Electricity rates in Greenwood Lake, New York depend on your home size and usage, so the kWh rate is more useful when comparing your rate to the average."
"Home Size & Type: Larger single-family homes generally use more energy than smaller apartments."
Let's throw some figures at this:
"In 2022, the average annual amount of electricity sold to (purchased by) a U.S. residential electric-utility customer was 10,791 kilowatthours (kWh), an average of about 899 kWh per month."
"Here’s a quick breakdown of average daily kWh usage by household size:
1-2 people: 15-20 kWh per day
3-4 people: 25-30 kWh per day
5+ people: 35-50 kWh per day
State Average Daily kWh
New York 25-30 kWh"
So it's a family of four living in New York state, say that's 30 kWh/day over 30 days in a month = 900 kWh/month x $0.44 cent per kWh = $396. That's the electricity, no idea what her gas bill was. And her monthly average bill *before* the increase was around $600.
So... it could be plausible that she's using that much power and then prices went up 200%. But it does sound like she and the kids are running every device they own 24/7 plus the house must burn through heating if it's old and poorly insulated.
The price of propane went from $2.50 to $4.75 between last year and this year. Propane is the main way of heating houses in that region.
EDIT Unless the article specifically stated that it was all her electricity bill. Electricity prices have increased noticeably in New York State over the last year and I am not sure of what the rationale is.
Phones and tablets probably aren't adding much to the bill. The battery capacities tend to run in tens of watt-hours and even heavy users rarely need to charge them more than once or twice a day.
Desktop PCs might be a significant component if someone has a high-end gaming PC and is spending every waking moment maxing it out. A 1kW power supply is a pretty huge one even for a gaming PC and the usual rule of thumb is to add 20% to your theoretical max power usage as a cushion when deciding which power supply to buy, so call it 800W peak usage. 0.8 kW * 16 hours /day * 30 days = 384 kWh.
Or they might be running a bitcoin mining rig, or some kind of private datacenter for some reason.
>Maybe her house is old and needs a lot of heating?
That seems more likely. HVAC is usually the biggest-ticket item for household energy costs, followed by other major appliances (stove, oven, water heater) and stuff like pool pumps and electric cars.
Looks like Greenwood Lake gets fairly cold in the winter, so if they've got no insulation, drafty windows, etc and also have an inefficient heating system (possibly supplemented with liberal use of electric space heaters, which in the US are typically designed to max out what a standard wall outlet can supply (1.5 kW).
They also get hot and muggy summers, so they're probably running air conditioning during those months. AC uses a ton of electricity even in well-insulated houses. Again, if they have an older house and haven't updated the HVAC recently, they're probably using an older and relatively inefficient unit. They ma be using window units, which come in multiple styles: some of these are decently efficient and others are terrible.
Or maybe they're just running an indoor marijuana grow operation in their garage.
+1. There is a very good chance that house is heated with propane, which has increased $ a lot in the last year. I have a house approximately 50 miles north of Greenwood Lake.
Yeah, I'm betting on an older house, cold winter, ramped up the heating, the kids are gaming, granny needs the electric heater in her room, etc.
I don't know why the price jumped so much (quite apart from whatever usage) but maybe they signed up with a new plan? or maybe this *was* a new plan and the introductory discount period ended and they got whacked with "surprise, this is what you'll *really* be paying from now on!"
A lot of places in the US have been rolling out new residential rate plans. The old standard was you pay a flat XX¢ per kWh, while the newer ones tend to be either "Tiered" (you pay more per kWh after you use more than a certain amount) or "Time of Use" (cheaper rate during off-peak hours, usually overnight and in the morning, more expensive in the afternoon and evening when AC is cranked up and the sun is going down so there's less solar on the grid). If they use gas for appliances and heating (this is common in much of the US), that often also has tiered rates. Not sure if this is the case there, but if they just got switched to a new rate plan and didn't notice, that might be part of the problem.
Also, gas, electricity, and (as B mentioned) other heating fuels like propane and fuel oil have gotten a lot more expensive in most of the US over the past few years. Inflation is part of this, but also because of tariffs and other supply chain and regulatory issues.
I originally thought she might live in far upstate NY, where it gets very cold. But she lives 50 miles from NYC. So there is definitely something very fishy about the story.
Edit: The article also claims that her bill suddenly tripled, which seems even more dubious.
If electricity costs in the US are anything like those in the UK then it is all too believable. It would take only a 1 or 2 KW electric radiator in each of several rooms and one or two in corridors and hallways. Also, this family may be charging an EV from their home supply, which makes it even more believable!
The article is conspicuously lacking in the sort of quantitative information that a rational person might use to understand the situation. There's a smattering of numbers, most of them scary, but mostly unconnected to one another or to any real context.
So, yeah, I'm pretty sure that they picked the absolute scariest number they could find to lead with, and didn't provide any details or explanation because if they had it would be obvious that this is not a representative situation.
A quick google suggests the average US household electric bill is ~$150/month. And I'm certain that somewhere in the UK some fool is spending £1,800/month trying to e.g. keep a Scottish castle warm through the winter with electric heat.
My immediate thought was they were trying to heat their swimming pool during winter. If I were a savvy reporter, I'd want to have seen their utility bill to try to understand it. But savvy reporters are few and far between.
I'd have to agree with John on this one. It's a (typically) poorly researched article that makes claims without any supporting information. Quick Googling says the average household in the Northeast uses about 600 kWh/month, and electricity rates average about $0.30 kWh. Transmission & distribution charges, meter charges, and taxes that would add another 10-15% to the bill. --> $240/month.
We'll probably find similar numbers for gas charges (but the price of natural gas has risen). That would likely offset some the electricity charges if they use gas.
Many Northeastern homes have furnaces that use heating oil. The average price is around $3.65/gallon. My mom, with her big old house in New England, used to purchase 1000 gallons of heating oil before winter kicked in. And she might have to order a second delivery before the weather got warm again (if the winter were particular bitter). Say in the range of $6000/year. That would be about $500/month over the course of year. But no gas, and no electrical heating. So her electrical bill was lower.
Join the various Discords for access to events? I don't know how maybe first try the new York obnyc discord and then ask in it for an invite to a sf discord. I have never done this
This is a follow-up to my post on Anavex Life Sciences exactly half a year ago (https://www.astralcodexten.com/p/open-thread-390/comment/135081998). That post was meant for entertainment and could be seen as ridiculing and negative, but in fact I thought that Anavex has the goods. That this was a case of "heuristics that almost always work" not working.
Well, the heuristics did work for now. The European Medicines Agency (EMA) has issued a devastating rejection of blarcamesine against Alzheimer's disease. Even before it clearly indicated a rejection, Anavex stock was already doing badly, instead of the more common run-up rally prior to binary events in biotech. A number of short reports had appeared, and afterwards shorts' comments were like "this trade was such easy money, it's embarrassing", or "nobody with a clue about statistics would have invested in this".
Disclosure: I own Anavex shares --- with gritted teeth in the face of bad management, but what can I do given stocktwits.com/Hosai/message/639122935, an enterprise value of less than 300 million dollars, and my belief that the company is not a scam (cf. Cassava Sciences). Truly scammy moves, if any, would have happened under different management more than a decade ago. Critics point to many discrepancies between and within various presentations of the trial data, but I concluded for myself, a year ago (when I had more time), that it's probably innocent. This isn't investment advice, though, especially as the stock may tumble again after final rejection: Anavex is appealing, and while conditional approval can't be ruled out, too many seem to be drawing a dubious analogy to Leqembi and Kisunla, both approved by the EMA on appeal.
The Anavex CEO espouses a philosophy of running small trials and then endlessly analysing the data. He even co-authored a paper titled "The new big is small". Had he quickly gone on to start a confirmation trial after the data read-out three years ago, that trial might now read out soon. But it is what it is. If I had the skills and time, I would investigate at least the issue noted in my next paragraph. Instead, I put it here in case someone else is interested. After all, if there was a safe and convenient pill that could "functionally cure" Alzheimer's, wouldn't it be a shame for it to be wasted? (I don't know how far it is possible or legal to use the molecule without official sanction, but the original patent is from 1997.)
The EMA may have ignored the Open Label Extension (OLE) of the trial because of too few finishers. But here is a claim that their results at the end are simply outside the distribution of (untreated) Alzheimer's trajectories: stocktwits.com/Bio_Oko/message/629232900. Is this correct? It's by Peter Karol, an independent analyst, Bio_Oko on Patreon, who has been claiming a bimodal distribution for years, something Anavex is now officially saying as well. Half of the population, the ABCLEAR3 group as Anavex calls it, with "lucky" genes, could have Alzheimer's slowed down by 75 percent or more. (And possibly other dementia as well, as what is supposedly stimulated is a general repair or backup mechanism, unrelated to amyloid or tau if I understand correctly.)
That spectacular hypothesis comes partly from post-hoc GWAS analysis, and we don't know how many genes were tested (probably a lot, otherwise Anavex would tell us?); but still, I would no longer bother to caution "not a cure" in parentheses like I did in my original post. I did not take Karol seriously enough, because I find his writing hard to read.
Another issue of note is the "Linear B-spline Mixed-Effects Model" that Dr Kun Jin developed to examine the effect of the many trial dropouts already before the OLE. He is former FDA Lead Neurology Statistician and joined Anavex after the trial results were out (other high-calibre people joined as well). While the skeptics argue that weaker patients dropped out of the treatment arm, leading to a spurious efficacy result, it could also be that responders in particular dropped out, so the trial result would *understate* blarcamesine' efficacy: this is apparently what his model is supposed to resolve. Does it succeed? It's possible that it was actually Dr Jin who decided or advised against the need for a confirmation trial. (And that the EMA isn't convinced, despite his model, because it thinks dizziness in the treatment arm effectively unblinded the trial ---although how could you *ever* then prove via RCT that blarcamesine works? Might as well grant conditional approval based on the existing data and withdraw it if the drug doesn't work after all.)
In nominative determinism, I recently found out there is a marine conservationist named Ocean Ramsey (she is famous for swimming with sharks and is popular on Instagram). I assumed she must have made this up and changed her birth name, but I checked her Wikipedia page and there's no mention of that.
I was also recently reminded that Tito's Vodka, evidently the top-selling US liquor brand, was founded by a guy named Tito Beveridge.
If the "R. Good" incident was real, and her west coast sister does bear Angela Lansbury's character's name, how was she targeted? Or is it elaborate AI fiction.
The principal horn player of the Philadelphia Orchestra long ago was named “Anton Horner,” and currently the Sarasota Orchestra has a horn player with the last name “Horne.”
My dentist in London was a Dr Fang, no really! A web search to see if he is still around reveals there is another London dental practice called Dr Fang, based on one of their dentists' surname: https://drfang.co.uk/
Both times I've seen the name Beveridge it's been connected to beverages. Somewhat famously in old internet meme lore, Michael Beveridge was a leading member on the board of the Yorba Linda Water DIstrict, who "ousted" Arthur C. Korn of the famous "korn.jpg" in a dramatic but entirely fictional board coup.
(This was famous solely because the water district had a locally hosted website in the 90s where the native size of the .jpg files for the board members' portraits were astronomically large. In those days the web designer would manually set the resize dimensions on the html page as an attribute in the <img src=""> tag, but the locally-hosted .jpg had to be in a folder with a+rx permissions in order to display elsewhere. So a visitor could see the source html and download the original .jpg, and if so inclined could hyperlink it on another web board either to add data traffic to the image host for nefarious reasons or to cause board readers on 56k modems to spend 5 minutes loading the page.)
I have a number of early-New England Puritan ancestors/relatives, and being my clan's longtime genealogy hobbyist I've built quite a large database of them. Puritanism of course was explicitly determinist in its outlook ("predestination"). That perhaps led to these given names each of which I have multiples examples of in my records:
Innocent
Freelove
Deliverance
Lovegod
Praisegod
Desire
Obedience
Freegift
Prudence
Experience
Those were mostly but not exclusively given to baby girls. There are also historical examples of even weirder Puritan names that were closer to one-offs; the above examples seem to have been normal in 17th/18th century Massachusetts/Connecticut/Rhode Island/Vermont/New Hampshire.
In the art museum in Worcester MA, there are a series of C18th paintings by the local semi-pro painters. They're awful. These guys made their own pigments, and so the subjects have green and grey skin. The proportions are all wrong.
The subjects have fantastic names Faith Savage is an older widow. And there's a row of misshapen children painted by the perfectly named 'Freake painter'. Unhappily not a more common name...
I saw a plaque the other day memorializing the "Mother of Our Aquifer" - A Mrs. Fay Sinkin. As this was her married name, it doesn't count as nominative determinism, but the unusualness of the name combined with her founding role in protecting the water supply delighted me all the same.
I'm a long-time SSC/ACX reader. I've begun a Substack where I'll be posting once or twice a month on history, culture, society, and anything that catches my eye. I started out with a book review of "The Cheese and the Worms" by Carlo Ginzburg. In case anyone is interested, here it is: https://connorsscratchpad.substack.com/p/book-review-the-cheese-and-the-worms
If things go well I'll definitely up the posting schedule.
Looking for advice on dealing with toddler eczema (it's probably eczema, we're waiting for a biopsy result). Steroid creams and anti-histamines have not worked, and the situation is getting dire. Poor kid is scratching himself bloody and can't sleep
Second the mittens suggestion. We used a combination of lotions and oat-base bath products as well as a water softener as our water was very hard. Not sure how much those helped, she has since grown out of it for the most part.
We dropped all wheat in his diet and use fragrant free and dye free laundry detergent. That cleared it up for our 2 year old. Have only seen it one time since we went no wheat diet and it was when I accidentally have him a dumpling that had wheat in it.
https://www.solveeczema.org/ Was brilliant for us, and also an interesting story of science. Massive difference over the course of a few weeks. We did need to get a water softener to be able to wash clothes with Castile soap, though soda crystals might work temporarily. We hand washed my sons clothes with castle soap until we got the water softener in.
Aquaphor is very helpful. Cerave was not- we discovered it has a detergent that just made it worse.
Best wishes! I know what you and your child are going through.
In the interim, getting him gloves so he can’t scratch himself might be better. Never dealt with skin issues but if my son has a bandaid on his finger he is compelled to take it off so we just put snow mits on him so he couldn’t.
I’m going to be in SF for a conference from the 17th to the 22nd of this month. Are there any interesting/fun ratsphere happenings that I should consider going to? Or anything else I should do while I’m there?
How much do you think your current actions, habits, projects and activities are helping to reduce existential risk from advanced AI (extinction, gradual disempowerment, extreme power concentration)?
Given your large platform and skills and talents, what are the best things you can do to help improve the situation and reduce existential risk (of the 3 major kinds)?
If you're talking about Scott specifically, I think he's probably reasonably close to that frontier already.
He has a competitive advantage in being a widely recognized intellectual on other topics, who also cares about AI risk and brings it up a lot, such that a wider audience is exposed to it by someone they already think of as reasonable and discriminating.
I feel like Scott *could* push on AI risk a lot more, but that this would be likely to just burn credibility and get him dismissed as a crank on this issue by the type of person who like to dismiss this issue.
In the abstract - I think Yuds + Scott does more to combat AI risk than 2 Yuds, at least in terms of public persuasion (2 Yuds might do more in terms of actual research, but I don't think Scott has a competitive advantage there).
> I feel like Scott *could* push on AI risk a lot more, but that this would be likely to just burn credibility and get him dismissed as a crank on this issue by the type of person who like to dismiss this issue.
I would say that he already does push on AI risk a lot and he already is a crank on the issue.
Appreciate you're probably strapped for time, but would be interested in your thoughts on this critique of the influential METR time horizons graph. In any case, think it'll be of interest to your readership.
These are not the objections I expected. I was expecting a more Andrej Karpathy style critique - that excessive focus on quantitative benchmarks pushes AI development towards easily quantified and verified reward signals, and side steps the most challenging parts of general intelligence - continuous and adaptive learning under uncertainty. Humans constantly have to navigate situations where there is no explicit reward, so we have to create our own intermediate goals to optimize for. General intelligence is not a set of capabilities, but the meta-skill of being able to acquire new abilities through interaction.
I don't think it would be a good use of resources for METR to redo their human baseline with lots more devs. Experienced devs are expensive - and paying them for 97 tasks would probably cost tens of thousands of dollars. It probably wouldn't change the numbers much (the task lengths might change, but the trend would probably stay similar). And it wouldn't address the fundamental limitations of benchmarking.
See my reply to Scott. Don't accept either that changing task lengths on a re-sample is unimportant, or that trend would necessarily be similar. I think the objections you were expecting come up implicitly in the discussion of task realism, but I agree that they're independently important concerns deserving a deeper discussion.
This essay automatically makes me dislike it by peppering its text with vituperations about how STUPID and UTTERLY WORTHLESS the benchmark is and how there is NO POSSIBLE WAY that ANYONE COULD EVER DISAGREE WITH THAT, and making me read through paragraphs and paragraphs of that in between the actual complaints. I rarely see essays that do this have anything useful to say, so I start out biased.
I am extremely not an expert, and I don't speak for AI 2027 or anyone like that, and this is just off the cuff. But it looks to me like the essay's first point is that if you replace the full suite of tasks with a subset of only the hardest tasks, the success rate goes down. This is tautological. The essay argues that the hardest tasks are the most "real", but this is meaningless - I'm a real person, and I frequently ask AI to do extremely easy things for me. But more important, I think the whole exercise is vacuous - there's no objective answer to what is "really" a "one hour task", and METR isn't trying to establish this - they're trying to fix a specific consistent definition, then see how AI success on that definition changes over time. The point is to measure the rate of improvement, not to say "Look, AI can now do exactly 56% of tasks, hoho". The essay's preferred "messy tasks only" graph also looks like it doubles about every seven months, so who cares which one you use?
(except that the messy-tasks-only graph has such bad floor effects that you can't measure anything at all before mid-2024)
The next point, about only three engineers per task, could potentially be more relevant, if it meant that their estimates were basically random and they couldn't measure improvement at all. But they include error bars on their graph, I see no reason to think those aren't the real error bars based on n=3, and the error bars are too small to affect their conclusion.
All measurements are noisy and involve methodological decisions and compromises with practicality. The point isn't whether you can find some way that it's not The Exact Objectively Correct Number Handed Down From Heaven, it's whether the methodological fuzziness is big enough that you can no longer trust the conclusion. My impression is this isn't true, with one argument being that METR does find very consistent progress, and if their method was returning screwy results you'd think they would obey a less elegant law. So I conclude that whether or not you like their method, they're keeping it consistent from one AI to another, and the remaining noise/sample size problems affect the results by less than an order of magnitude, in a regime where whatever AI capability produces these time horizons is increasing by orders of magnitude.
I see the comment thread where the author argues with Jitse Goutbeek on this point, but IIUC (definitely not sure of this, leaning heavily on Jitse's analogy) the author is there making a different point that you can never truly measure "time horizons", but only time horizons on some specific set of tasks measured in some specific way, and for all we know they might be different in some other regime. I think that's fine. But first of all, it's not really related to any particular complaint about the measurement process - even if there were 5000 engineers and it was the messiest 1% of tasks, you could still make this argument. And second of all, I think METR's response would be that lots of different things measured lots of different ways still show a similar pattern, for example https://metr.org/blog/2025-07-14-how-does-time-horizon-vary-across-domains/ , the 80% reliability trend, the messiest tasks subset, etc. I don't want to claim they're all exactly 7.1 months or whatever exact number METR reports as their headline result, but I think METR has successfully created a strong prior that whatever specific thing you care about will be an exponential with a doubling time measured in months rather than years, and that (for example) if some set of AIs go far above trend on their graph, that would be surprising and represent a phase shift in AI progress.
Appreciate the read and thoughtful reply Scott. Several things I want to push back on.
A quick note on tone first.
I went back and forth about some of the language here, but decided it was warranted. I think if a certain idea (study, benchmark, etc.) has a large enough portion of the mindshare, you need to push back pretty hard if you want to change that. I also think the tone is natural if you think the benchmark is as faulty as I do, so at minimum I think its defensibility hinges on my substantive complaints. We also shouldn't lose sight of how much influence this benchmark has plausibly had on folks’ allocation of labor, energy, money, etc. That also speaks in favor of a stronger tone being appropriate if they have in fact been misled.
I’d add that you’re not my typical reader, and for better or worse strong rhetoric moves people (I'd say for better, but explaining why would require a long digression). Maybe we're reading different things, but I've read plenty of worthwhile argumentative writing that takes a similar tone.
Onto substance.
It seems a bit unfair to gloss my discussion of task realism as tautological because "harder tasks are harder." Task realism and completion time are two different variables and their relationship is important. Nor do I agree that, if doubling times are preserved, lower success rates on more realistic tasks are irrelevant. Maybe you don’t accept the premise here, but I suspect that whatever models' rates of improvement, their performance will begin to taper at some point. If that's true, then it matters, potentially a lot, when that tapering starts. Substantially lower y-values on the y-axis mean lower capability at tapering. I think we can agree that how capable models are when tapering starts is an important piece of information. If so, we can’t simply write off the scalar differences between doubling time graphs for different task sets.
As for arguing that "the hardest tasks are the realest," that's not what I'm arguing, it's what METR is arguing, or really what they're finding. They construct an index measuring task realism, which they call "messiness" (they explicitly equate it with realism), and find that messier tasks turn out to be much harder for models (none top 30% success rates for the messiest half of all tasks). You may ask AI to do tasks you find intuitively easy, but by "messiness," METR means something like “degree of similarity to tasks a professional software engineer would do at work" (check out the properties they’re using to score messiness on pgs. 39-40 of the preprint). Considering ambient concerns about labor substitution, automating AI R&D, etc., I feel like we can agree that how models perform on this subset of tasks is important.
I agree that METR isn’t trying to establish how long different tasks “objectively” take, but nor do I think all they’re trying to do is to come up with some (i.e. any) consistent metric to measure improvement. They describe the y-axis as how long it takes “humans” to complete software engineering tasks. As a result, I don’t see how you can claim that these numbers are only meant to be relationally meaningful, rather than to also make outward-facing, empirical claims. That is how METR describes them, and I think it’s clear they want us to understand how capable models are by comparison with humans. If they didn’t, why not just describe time horizons as “capability scores” or something else much less committal? Why even run such a complex benchmark in the first place? Finding “a specific consistent definition, then see[ing] how AI success on that definition changes over time” is the same thing all the other benchmarks are doing, and METR is doing something different for a reason.
I admit to not having looked into how METR’s error bars were calculated before reading your comment, but I’m glad I did because I was able to confirm that they in no way solve back for my concerns about their exceedingly small / biased sample. Those error bars have nothing to do with the human baselines on the y-axis; in fact the baselines are fixed before they’re even calculated. Simplifying a bit (promise the complications aren’t pertinent), the error bars in the time horizon / doubling graphs represent bootstrap CIs for models’ performance over the task distribution. In other words, they are derived by re-calculating results for thousands of slight variations on (“resamplings” of) the task set, meaning e.g. dropping a couple tasks, doubling some tasks, increasing or decreasing the number of model runs on some tasks versus others (disclaimer: I’m not especially confident in the precise details of this description because the paper doesn’t go into much detail and my stats knowledge is pretty minimal). As a result they (the error bars) in no way represent or address uncertainty surrounding the human baseline estimates themselves, only that surrounding models’ performance on the task set, holding the human task completion times fixed.
So, bottom line, the situation is precisely as bad as you imply before walking things back on the assumption that the error bars somehow take care of the tiny/biased sample. To really underline this: the numbers you’re seeing on the y-axis are simply the result of averaging ~3 baseliners’ completion times per task. Also remember that these baseliners aren’t randomly sampled, but recruited from METR’s professional network. And that’s before getting into all the other sources of noise from the weird incentive scheme, anti-cheating protocols, and so on.
I also want to touch on a point you only raise implicitly but that others have harped on a lot. You cannot, under any circumstances, assume that the doubling times METR ends up with will hold for a much larger, properly randomized sample (this is what’s at issue in my exchange with Goutbeek). The counterexample I give in the article is a scenario wherein "tasks METR found took 2-3 hours ... turn out, in [a] larger sample, to take 1-1.5 hrs" but without deflating lower task durations by the same proportion. That would greatly compress the completion time distribution, and therefore change doubling times by quite a bit.
There are lots of reasons something like this could happen. For instance, maybe short tasks have much stabler time floors (since many don't require special expertise, and so can't be performed faster via superior understanding), but times for long tasks are sensitive to which/how many domain experts end up in your sample. In that case, changing samples from one with few domain experts to one with lots could substantially lower times at the higher end of the distribution without lowering times at the lower, or doing so a much smaller amount. That would in turn increase doubling times. In short, these sampling issues are a very big deal, and even independent of all my other concerns are very close to making the benchmark useless all by themselves.
I don’t see how you can brush this off as “fine,” nor do I accept that the issue is the same as “if there were 5000 engineers.” Of course it wouldn’t be; if you actually had a large, random sample, then you could read the METR graph more or less as intended (putting all the other issues aside), since the times on the y-axis would be representative of software engineers on the whole (possibly yielding very different doubling times), rather than those in a small, idiosyncratic sample.
With respect to the idea that “lots of different things measured lots of different ways show a similar pattern,” I disagree that the other benchmarks represented in the linked time horizon graph are “measured in lots of different ways.” On the contrary, they have many of the exact same issues as METR’s, and in some cases are worse. For instance: many of their time horizons are guesstimated by task annotators, rather than measured from human baselines. They also suffer from very similar task realism issues, and maybe most importantly the frontier labs are aggressively fine-tuning for all of them. So no, there are too many correlated sources of error here for these other benchmarks to speak to the robustness of METR’s.
In summary, I really do think the benchmark is as bad as my rhetoric implies, and insofar as it has misled a lot of people who have plausibly reallocated large quantities of time and labor on its basis, deserves a strong rebuke. Do I doubt that models are in fact getting better, esp. at software engineering? Not at all. But I am very uncertain as to how much better they’re getting and how fast they’re getting there. I would add that the slow rate of enterprise adoption and dearth of major macroeconomic impacts (growth, productivity, labor displacement, etc.) tells a very different story than the large volume of noisy benchmarks, METR's included.
I read your critiques with interest, but my main takeaway was thinking the main METR metric is probably about as good as you can expect.
With the exception of the quite valid N objection, most of your objections seemed a little too far on the "protest too much" side of things to me.
I'm assuming this benchmark is basically measuring "how close are we to being able to counterfeit a dev or white collar worker."
So methodologically, METR took highly qualified / capable people and had them do problems outside of their wheelhouse, so the human solvers needed to context up and learn a little.
That seems totally fair as a basis of comparison for an AI dev or white collar worker? That's literally the 'new project' baseline for any current white collar employee.
People with domain expertise did it a lot faster? That's fine too, and totally expected - I think we ALL expect that to fall, too, and should be a separate eval.
The biggest real problem is the scarcity of evals, the low N, and I agree they need better coverage there.
The messiness is a lesser problem, to my own mind, because obviously the way this is going to work is much like vibe coding today, where you have one human managing a fleet of AI white collar workers, and that human handles all the messiness while the AI's are still skilling up.
But per what I perceive the overall intent of the benchmark to be, an AI can step in and counterfeit work WELL before they need to be able to handle the messiness well.
Also, wouldn't they have to have been filtering their evals on messiness? Nobody is lovingly hand crafting artisanal benchmarks, obviously they need to be programatically evaluable?
And in terms of being self-consistent they need to be programatically evaluable, because if you hand craft OR hand-eval benchmarks with human judgment, you're not comparing like to like and it's not a true "same scale" differentiation of the models' capabilities.
Given the absurd amounts of money floating around in the AI industry, does no one have the money to do proper benchmarking on their effectiveness? Or is there simply no incentive?
There is an incentive for investors to do a proper benchmarking and there’s an incentive for those investors to never share the results of their benchmarking to the public, or at the very least, not release them until they feel fully invested. Even they did release them, how much would you trust the opinion of a person who just invested a billion dollars into something that what they release isn’t biased to drive up the price?
There's huge incentive in proper benchmarking AI effectiveness.
AI capabilities are (basically) causally driven by good benchmarks. If a lab manages to create a benchmark that's effective then it directly leads to a better model and a piece of the billions of $$$ in the industry.
I think the issue is that those with the money (i.e. frontier labs) don't have the incentive to lest more scientific benchmarks make models look worse.
I have a couple of possibilities. It's putting in more emotion than the topic deserves.
Or maybe someone hates the food because it's associated with a group they hate.
They kept including Ensure with my meals at a hospital, even though I asked them to stop. (There was also surprisingly good food.) It tasted vile, and I threw it hard at the trash can, which startled something I regret-- he was just standing nearby.
I don't think hating Ensure was immoral. I just surprised that there are people who like it.
You can get it in import shops in Europe, and I'm pretty sure the cans I buy are actually produced somewhere in Eastern Europe. None in stock for me to check right now though.
That's a fair point. I agree that your soul only gets consigned to the everburning flames of perdition if you've tasted the drink and then subsequently denied its greatness, and not if you've never had a reasonable opportunity to experience it.
It's possible for it to be morally wrong, if you're actively trying to remove a food from someone else's diet. "I think eating insects is gross, therefore we should stop foreign countries from being able to eat insects."
But obviously it's not morally wrong to refuse to eat a food you're allergic to.
If you're buying your own food then buy what you want. It's literally a matter of taste.
This is just an assertion of the correctness of your own moral framework, and begging the question: Why can an "arbitrary subjective preference that has no consequences" not be morally wrong?
And so pragmatically, consider potential scenarios that contradict the simple 'preferences have no consequences outside the holder':
- you could like to eat (e.g. novel) things that increase prob. you become sick, which has various levels of economic externalities
- you could like to eat things that increase prob. you get a communicable disease (recall cholera due to poor sanitation - was *tremendous* issue for most of human history)
- if you're deliberately eating foods considered 'taboo' by your community, your signaling something about yourself that may end up degrading coordination capacity
- 'honor culture' was historical default-> reputation isn't purely personal, it is carried 'by association' to kin+community.
- classic bayesian issue - the prob of someone with a true 'bad brain' manifesting unusual eating much higher than a normie + risk avoidance -> signal gets over-interpreted -> externalities via. reputation
I mostly agree with you, but as an intellectual exercise I can imagine a few possibilities depending on which ethical framework you use to analyze:
Utilitarian: Shared eating situations (dinner parties, going to restaurants together, banquets and other catered social occasions, etc) require some degree of coordination of food preferences, and sharply differing preferences lead to lower total utility than if you have the same preferences as everyone else in the group. Skipping occasions you would otherwise attend deprives you of the utility of attending and other attendees of the utility of your presence. Eating stuff you dislike has obvious disutility to you, as does skipping the meal (or the course) while still attending. Asking others to accommodate your preferences by tailoring the menu to you or providing multiple options imposes costs of them (either providing sub-optimal choices to others or requiring more planning, effort, and expense to provide parallel options). And self-segregating based on food preferences leads to group formation that is sub-optimal in other dimensions. If you could reasonably self-modify to not dislike the food that those around you like, that is arguably the ethical thing to do from a utilitarian perspective.
Dentological: Not sure. I'll leave this for others.
Aristotelian Virtue Ethics: Disliking healthy foods leads you towards Intemperance.
Objectivism: It is in the nature of humans to be omnivores. To dislike meat is to deny your nature.
Divine Command: Jesus told us to eat bread and drink wine in remembrance of him. If you don't like bread and wine, you don't like Jesus.
Bretty good, all are on point except divine command (think you're reaching there). Deontology seems tricky, but then, I feel you could make the virtue ethics case apply to it too, could be said you have an obligation to take care of yourself and eat healthy food (not sure why virtue ethics isn't a subset of deontology).
I never could quite wrap my mind around the internal logic of Deontology, even in areas where I think I have a pretty good handle on its prescriptions. Sometimes it feels like an extreme interpretation of Rule Utilitarianism and other times it feels like Virtue Ethics with extra steps.
Yes, I'm reaching on divine command. I was looking for religious dietary rules (ideally rooted in scripture) where a certain food was obligatory, but everything I could think of off the top of my head beside the Lord's Supper was a food being forbidden rather than mandatory.
The other candidates I came up with were the Catholic practice of eating fish on Fridays and the Jewish practice of eating matzo during Passover, but I think both of those are substitutions for prohibited food and eating the substitute is a cultural practice rather than a religious mandate. Catholicism forbids eating meat on Fridays during Lent, and before 1966 the rule was all Fridays year-round which is still observed by many individuals and communities. "Meat" is defined as coming from land animals, so seafood is traditionally substituted. I think the Jewish rule forbids eating risen bread during Passover, so matzo or another flatbread is traditionally substituted.
I'd point out that according to the latest, Catholicism suggests some sort of sacrifice on Fridays to promote virtue through the practice of sacrifice. Meat was given as a suggestion but it's been taken as an order for a long time. (How this could be mistaken for so many years is a bit beyond me, but that's what I'm hearing.)
Interesting. My knowledge on the subject comes mostly from my father, who was raised Catholic but has been non-practicing since before I was born. Supplemented a bit by stuff by a number of random bits that I recognized as part of the pattern. When I worked for Microsoft a decade or so ago, the cafeterias had a lot of seafood options on Fridays for the benefit of Catholics who kept the older practices. I've also encountered odd historical tidbits like the chronicles attributing the cause of death for King Henry I of England to "a surfeit of lampreys", with the backstory being that lampreys have an unusually beefy flavor profile for fish and thus were a popular luxury food for rich people to eat on Fridays, but were thought to be unhealthy due to something to do with the Four Humours theory of medicine.
None of these things, you may notice, inform me well of modern theology on the subject. So thank you for passing that along.
Yes, it does seem to do so, and it looks like there are a number of Jewish religious websites that talk about eating Matzo being theologically mandatory at certain points during Passover, although opinion seems to vary a bit as to what those points are.
I've lost tremendous respect for this community, which was already waning, by seeing so many people try to earnestly defend "Trump + Greenland". Why? Because.
And yes, this deserved a top-level post. Why? Because Because.
[oh, btw, the reason for Greenland, Venezuela, Minneapolis, and probably an invasion of the South Pole in 2 weeks is bc Epstein. Universally. Objectively. Rationally. There, now you know]
I agree with your feeling. Since most of my interaction is online, it isn't very clear how many of these are sincere people as opposed to (a) fictitious or (b) trolling.
For sincere supporters, I have seen four types:
(1) fans: they don't believe POTUS matters to day-to-day life and are entertained by Trump
(2) single/narrow issue: Trump truly gives them what they want wrt one issue and they don't care about the rest. I would argue anti-abortion, reduced income/corporate tax, and reduced business regulation are the main issues of this type.
(3) Democrats are worse: a CW version of (2), where the desired outcome is significantly less clear and/or Trump isn't demonstrably delivering. Key issues here are anti-woke, immigration, government corruption, misuse of government agencies, foreign policy. I understand all of my implied claims here are debatable.
(4) Stephen Millerites: these are people who simply think this is the way a POTUS should behave/government should function.
Because of the norms around here, whatever the real reason, most support has to get cast along the lines of (2): "issue X is really important" and "Trump is doing a good thing on X"
My mind is probably overwhelmed at this point by all the stuff I'm hearing on the media, but... regarding Trump and Greenland, is there any way at all that it would be remotely feasible to have Greenland like Puerto Rico?
That is, a US territory where the people are US citizens but it is self-governing to an extent? I see there are other American territories:
https://en.wikipedia.org/wiki/Territories_of_the_United_States
Maybe something along the lines of American Samoa, which seems roughly equivalent to Australia. The head of state for Australia is the British monarch (technically the queen or king of Australia as distinct from king or queen of the UK) but it's self-governing in practice. American Samoa seems very roughly along those lines:
"The Governor of American Samoa is the head of government and along with the Lieutenant Governor of American Samoa is elected on the same ticket by popular vote for a four-year term. The governor's office is located in Utulei. Since American Samoa is a U.S. territory, the President of the United States serves as the head of state but does not play a direct role in government. The Secretary of the Interior oversees the government, retaining the power to approve constitutional amendments, overrides the governor's vetoes, and nomination of justices."
Am I crazy, or would it be workable - if the Greenlanders agree - to make Greenland an American territory?
Alaska was a territory for a hundred years. That part's easy enough.
I'm pretty sure no one in the White House wants to make it a state - 2 dem Senators, a dem Representative, and 2 dem electoral votes every 4 years. But I'm also pretty sure the people of Greenland have no interest in any arrangement that makes the US citizens.
It'd actually be _three_ electoral votes.
There are different arrangements that could theoretically be made (e.g. Compact of Free Association). The existence of possible legal arrangements is not the blocker here.
The blocker here is that Greenlanders don't want it, the Danish don't want it, and going to war *with NATO* over a bunch of ice where the US *already has a military base anyway* is a monumentally stupid idea.
Sevastopol. The United States doesn't "have a military base." Not -really.-
So it is theoretically possible, if the Greenlanders and the Danes were in agreement? Okay, less insane than "Trump is going to invade Greenland" reporting, but my mind is still boggled.
Then again, why does Denmark want Greenland, either? The impression I got from reading the "Girl with the Dragon Tattoo" novels was that native Greenlanders were something like the Indigenous for Canada or Native Americans for the USA - regarded as original inhabitants of the territory now administered by another culture with no real authority or control of their own over their affairs, and somewhat of a burden/second-class citizens as far as the majority society was concerned.
> why does Denmark want Greenland
For all that they mock Trump for supposedly having the same motivation, it's because it's gratifying to see their country look really big on the map.
I've got an antique rifle displayed over my fireplace, that I have no real use for. I'll occasionally take it out to the range, if I feel like dealing with all the hassles of black powder. Mostly, it's just sentimental value.
It's not for sale. My grandfather made it, my father owned it, now I have it, and in due course I'll pass it on to the next generation. If there's theoretically a price at which I would sell, it's one that nobody is actually going to offer and if anyone says they're offering that much I'll assume they're planning to cheat me.
If someone says "I'm taking that rifle, we can maybe cut a deal but if I don't like your deal then I'm not ruling out breaking into your house, because one way or another it's mine", then we have a problem. The kind that has me checking that my other, more modern firearms are ready to deal with people trying to break into my house.
Yes, unfortunately your modern firearms aren't much use against bombs and drones. We live in strange times where it's easier and safer to take land than posessions, seeing as land is more resistant to explosions.
I think you're taking the metaphor too literally. Greenland is not John Schilling's house, it has more than just small arms protecting it.
Like what, the dogsled?*
*yes, really. Every year they circle the entirety of Greenland.
Like what? Denmark isn't going to war with the US, for goodness's sake. If they have the opportunity to be fairly compensated for the land in order to avoid a conflict, they will likely take it. As for the Greenlanders themselves... the country is barely populated and can barely sustain itself without outside support. What they want is, frankly, irrelevant.
The reason that people keep talking about Trump threatening to invade Greenland is because Trump keeps threatening to invade Greenland.
"We don't need this but we'll get rid of it when we're good and ready and no one gets to boss us around" is a reasonable position.
A lot of it is theoretically possible. I don't think people are hung up about that; theoretically, the united states of America could become part of Ireland.
>why does Denmark want Greenland
Well, it's part of the Danish nation (this alone should be enough of an argument, so I'll underscore it). It is also recognized as such by international treaties (and specific ones with the US- https://en.wikipedia.org/wiki/Treaty_of_the_Danish_West_Indies) the Danes want to keep it, the Greenlanders don't want to leave Denmark.
Suppose in ten years – through means that are of no relevance – Greenland becomes part of the territory of the United States. Would YOU consider that alone enough of an argument for why Denmark shouldn't retake it?
Okay, because it just doesn’t matter let’s assume someone waved a magic wand and Greenland became a US territory recognized by other nations and by treaty the way it is part of Greenland now.
Then Denmark would not even consider trying retake it because their leader is not president not-so-sane.
This is goofier than your hostis humani generis argument.
> Denmark would not even consider trying retake it
Right, because they are too weak to. When they were stronger, they conquered vast tracts across the North Sea region. They're just on the other side now.
You have a small-minded understanding of the law of nations. Just like the British Empire before it could unilaterally redefine terms understood since antiquity to suit its purposes – which was my argument earlier you refer to – America can today remake the world and have it "legitimized" later if it bothers to, but in two hundred years, I expect there will continue to be people who don't understand how this works, and sneer at those who do as "goofy."
I have some difficulty parsing your question; is it something say that the status quo plays some factor that should be respected, but this status quo can change and then the new status quo should play a role? So if the US "has" greenland (legally? militarily? 51st state? militarily occupied? Greenlandish independence and vote with majority to join?) 10 years from now, than the Danish should accept that, and not try to retake it because "it's now part of the US"?
I think the means are relevent, if that is your question.
I'm saying your argument that Greenland currently being Danish territory should be enough of a reason not to change that proves too much.
I don't think you actually believe that yourself, since you're arguing the means by which the territory was acquired are relevant. Now, if you're actually just saying that the gains aren't worth the price the US would have to pay to conquer and hold the territory, both against Denmark and its allies' attempts at reconquest and against rebellions, fine, that's a sane argument.
As I had some discussions on (the apparent lack of) free speech in Germany: Basically it's a case of a different social contract and different Schelling fences. (Compare this old post by Scott: https://www.lesswrong.com/posts/Kbm6QnJv9dgWsPHQP/schelling-fences-on-slippery-slopes )
The different fences in practice:
- Americans see any restriction as a slippery slope to tyranny
- Germans see lack of restrictions as a slippery slope to '33
I cannot believe this is so! I understand that in the North American Treaty Organization Charter, it says one should not ban political parties, or otherwise infringe on free political speech. This doesn't sound like the sort of document that West Germany would have signed, had they the different social contract you reference. Nor France, for that matter.
https://www.19fortyfive.com/2025/04/natos-democracy-deficit/
As someone who prefers the American approach, I find I can still steelman the German one, and even point out echoes of it in current American politics. The Germans were presented with evidence of one set of ideas that led to a great deal of ruin for them, and that evidence was hard to refute. An extremely simplified version would be the idea that it should be okay to kill someone under your care, on the premise that person is a threat to your existence, while pointing your gun at anyone who seeks to intervene. German authorities post-WWII concluded that that was an idea that should never be taken seriously, because doing so turned out to be an even clearer and present threat to them.
Therefore, anyone who tried could be stopped without objection, even by means other than debate. For them, the debate was closed, and stayed that way for generations. Maybe debate was insufficient; maybe individuals were still too close to the details that made such ideas tolerable; maybe it was a problem of informing everyone of the case against the idea; maybe a combination. Whatever the reasoning, that was the determination.
Nowadays, I'm seeing another reason to want to suppress such ideas, having to do with individual attention span. Current technology has given broadcast power to any individual, and yet we all have only so much time in the day to absorb news; if that time is coopted by someone spreading dangerous or even distracting ideas through broadcast channels, and we don't have the ability to effectively filter out such ideas, then our model of the world is inaccurate and our time is wasted.
Ideally (for the pro-American approach), individuals ought to possess such an ability. In practice, it's proven elusive. Popular networks like YouTube or Reddit or X can attempt to filter on keywords, but some people have discovered that spreading such ideas in bad faith is fun in its own right, much like prodding an anthill and watching the ants scurry, so they have a strong incentive to evade filtering. There's also the problem where, if the filtering is _too_ accurate, then individuals can filter sources run by powerful people who feel they are entitled to being heard, so they probably have an incentive to drag their feet on improving the filtering. The exception is when the filtering blocks ideas reminiscent of the early 20th-century horrors.
The result is that that filtering sticks around, is maintained by a relatively small group of people, enforced on the public, and the public tolerates it as long as they believe those ideas are the old horrors.
>Nowadays, I'm seeing another reason to want to suppress such ideas, having to do with individual attention span...
At that point, you may as well just stop being democratic. The nazi ban is at least in principle limited, but if The Peoples decision making needs to be managed in a content-sensitive way to be workable at all, then why listen to them at all? Have whichever body would do that managing make the descisions and be done with it.
You make a good point, and I agree. A lie can travel halfway around the world while the truth is putting on its shoes. A democracy must be able to defend itself against those that wish to destroy it, or else it won't exist for long. The Weimar Republic learned this, too late, the hard way.
I wonder how much the Weimer Republic should actually be counted as a democracy; it was largely established by the actions of foreign nations, and collapsed in a single generation. How much was it a democracy as opposed to Vichy Germany?
Congratulations, your model of the world is inaccurate and your time is wasted. Actual reality isn't generally discussed in polite company. See the Forbes list of "most wealthy men" for details about what is "polite company" and what is just "new rich."
They're also deeply committed to pretending to have free speech after all, and almost the entirety of the American ruling class pretend to believe them – Vance's speech at the Munich Security Conference might the closest anyone has come since the end of the Second World War to pointing out the blatant lie, but even he walked it back shortly after. The US allies with foreign nations with values antithetical to its own quite often, and it's perfectly reasonable to argue that it's already fine allying with Saudi Arabia and the other Gulf monarchies, and most of Europe isn't much worse than them, but instead, they put on this elaborate kabuki theater of "shared values" and "democracy" and lecture other countries for their lack thereof, which is deeply irksome.
> […] Saudi Arabia and the other Gulf monarchies, and most of Europe isn't much worse than them […]
I'm curious: Do you genuinely believe that "most of Europe" isn't better than Saudi Arabia in regards to free speech? Or is this just a deliberately provocative comment to aggravate the discussion?
I occasionally have cause to transit through airports in Europe, so I don't think it prudent to say.
What you're quoting was me trying to forestall any suggestion that I was implying the US ought to repudiate its alliances with countries in Europe out of moral outrage: if there is any such a point where a partner's practices become so repugnant that reasons of, say, troop morale or civilian unrest, compel one to so break off relations, Europe is certainly NOT there.
> I occasionally have cause to transit through airports in Europe, so I don't think it prudent to say.
That's deeply ironic, given that no European countries ask for your social media accounts when entering or transiting, whereas the US requires their disclosure for visa applications, and plans to make disclosure mandatory even for visa-free travel.
As we used to joke during communism, there is "freedom of speech", and then there is "freedom after you speak"; those are two different concepts.
Russia had freedom of speech, in that "He was drunk!" was an acceptable excuse for saying... practically anything. Freedom of action was a different matter, of course. I imagine it was a little tighter in countries where liquor wasn't the desired method of pacification of the proles.
If the people cannot discuss their own political system and political wishes freely, then the people are not in control. You can't hold the following three propositions all at the same time:
1. The people have the capability and the right to control the government.
2. If the people are allowed to speak and listen freely, then they will choose Nazism.
3. Nazism is bad.
I suppose the German political system works reasonably well, but it is not free and it is not a democracy.
Well, the first sentence is just not a characteristic of a democratic republic, so I don't think the contradiction you believe exists is there.
The people have the right to elect a government, and there are extensive rules constraining said government from doing what it wants to its citizens. But that is not the same as "controling" the government.
Under normal democratic theory, sovereignty rests in the people as a whole, and the government is just the individuals deputised to run the country on their behalf. I think it's reasonable to describe this as the people "controlling" the government, even if, for practical reasons, this control is exercised indirectly in most modern democracies.
The use of the word "control" is a question of definition, so let's look at it in the context of what you wrote.
You said that people have the "capability and the right to control the government". But that is just factually untrue without major qualifiers you omitted.
We also say that shareholders are in control of a public company. This is a practically useful sentence to say in most contexts.
But is it contradictory to also say that shareholders cannot direct a company to commit fraud or the CEO to kill himself?
Not really, since we all understand that shareholder control is not control without limits. In the same way, the people's "control" of the government is limited. And the fact that it is is not irreconcibale with what I understand as "free and deomcratic".
>You said that people have the "capability and the right to control the government". But that is just factually untrue without major qualifiers you omitted.
If you look more carefully, you'll see that that wasn't actually written by me, but by someone else who happens to have the same avatar.
oops, apologies! Well, that could never happen with a unique avatar, such as mine.
> I suppose the German political system works reasonably well, but it is not free and it is not a democracy.
"Democracy" does not mean "unrestricted dictatorship of the majority". There has never been a pure democracy, i.e., a state where the population has unrestricted decision power through voting. In particular, all modern democracies (including the US) have checks and balances – typically in the form of legislative and judicial institutions and processes – to protect certain rights and constitutional structures.
With this in mind, there's no reason to single out Germany. If you take "free" and "democracy" in their absolute meaning, there is not a single country on Earth that's "free and a democracy".
Merz filing 5000 or so criminal complaints (some of which led to SWAT teams) about people shitposting on the internet seems to indicate that Germany is, in fact, rather singular about its approach to free speech, and authority figures.
And the closest one to this ideal is probably Switzerland (i.e. definitely not USA).
The US has restrictions.
The following is a german text from me translated by Claude, I don't know if the technical terms are always the right ones. Feedback on this take would be welcome. :)
USA: Maximalism with minimal exceptions
The First Amendment protects speech extremely broadly. The Schelling point is "the state almost never intervenes." The few exceptions are narrowly defined: imminent calls to violence ("imminent lawless action"), true threats, defamation with demonstrable harm, obscenity (very narrowly defined).
Hate speech, Holocaust denial, Nazi symbols are all fundamentally legal. The idea: once you start banning "bad ideas," there's no defensible boundary. Better to keep the marketplace of ideas open even for shitty ideas.
Germany: Value-based democracy with active boundaries
Article 5 of the Basic Law guarantees freedom of expression, but with explicit limitations in "general laws." The Schelling point is "democracy must be able to defend itself against its enemies" (militant democracy).
Prohibited: incitement to hatred, Holocaust denial, Nazi symbols, insult is criminal, defamation even without the American "actual malice" standard. The historical lesson: Weimar let itself be talked out of existence.
Both have internally coherent logics that emerged from different historical situations/traumas. The USA fears state power more than dangerous ideas, Germany fears dangerous ideas more than state power (at least in this domain).
Re. Nick Shirley fraud allegations: I just kinda assumed "This sounds like a thing that is probably true, I will believe it because the conservatives in my life do because they are not hysterical as a rule and it's not in my state and I can't do anything about it."
Just recently came across an interview with the dude on my "Listenin' while I dig Holes" feed, and holy shit: This kid is dumb as rocks. He is legitimately subnormal level just not that smart; and he said some things that demonstrated a rock solid commitment to post-truthism ("I don't like that you didn't condemn Kirk's assassination" "I did, to the point where people were giving me shit for being to nice to Kirk" "I'm not going to watch it and I refuse to update", infuriating).
It made me go quickly check out his video, and it is straight up stupid. Anyone that has ever worked in, around, with, or patronized a certified childcare facility should have been able to know that +/- 85% of everything he showed in the video is not only normal but legally required, and the remaining 15% was not sufficient to justify the claims (eg, could have been explained by two employees calling out sick at the same time/a pipe breaking).
If fraud is discovered form this (which it probably will be, most for profit childcare/elder care/religious facilities that can get state money are skimming off the top), it would be like someone saying that the lottery numbers they won on were revealed to them by the ghost of PT Barnum.
Incredible how easily your information environment can be poisoned, really puts you in a constant epistemic crisis.
It was pretty clear from the video that "David" was the brains of the operation and that Shirley was just there for charisma and distribution.
Is he supposed to be the originator of the chidlcare fraud thing?
Essentially, yes. There were some reports a few years ago, but they got limited traction, and nothing came of them, negligible compared to his video.
After seeing the Democrats' housing policy in New York, I think literally the ONLY thing the Republicans can do to be the bad guys is adopting similar policies. Knowing the Stupid Party, I'd bet that's what they're going to do by the 2028 election, if not even sooner.
Housing policy is an issue that crosses partisan lines. You have both democrats and republicans on both the pro and anti-housing sides.
One of the biggest factional wars in the Democratic party right now is over whether to become more pro-housing or not.
What's your opinion on the World Reserve Currency, subset NYC?
I don't know what you mean. Are you referring to Eric Adams's crypto token? Caveat emptor.
Exactly! You've disqualified yourself from discussing New York Real Estate, which is one of the world's Reserve Currencies. Education first, talk later. (NYC, like London, has unique game-rules for capital investment, that are somewhat governed by the use of its real estate as a world reserve currency).
According to you. Please .don't gotcha people about idiosyncratic theories.
It's not an idiosyncratic theory. NYC and London are ranked Alpha++ because of the global elite's storage of wealth in their housing market, among other things.
I see what you did there.
It's just a vastly different marketplace when the world's elite use it as a place to store wealth. The stability or lack thereof of the real estate market is different than one would see in say, Chicago, San Francisco or Los Angeles.
What?
Democratic and Republican recreational activities.
This is based just on my impressions, not statistics.
——
Activities favored by Democrats:
Bicycling
Running
Yoga/Pilates
CrossFit
Backpacking
——-
Activities favored by Republicans:
Outdoor powersports (for example Jet Skis, ATV’s)
Golf
Hard-core weight training
RV camping
Hunting
Substack blogger Ryan Broderick recently spent 3 days in Minneapolis observing ICE and related matters. His observations about ICE:
>ICE agents are, simply put, fucking clowns. According to The Atlantic, they receive 47 days of training — in honor of Trump, the 47th president, naturally. Many of them, also, can barely read or write, apparently. The ones I spent the weekend following around didn’t even have proper uniforms, with some wearing sneakers. In Minnesota. In January. These dipshits are also wearing camo in the snow. They clearly do not have any training when it comes to their own weapons either. Multiple times over the last few days, I watched officers fire pepper spray balls at the feet of protestors barely a few inches away from them. These weapons are basically paintball guns full of concentrated pepper spray. So when they hit a target, they explode into the air. Which meant ICE agents regularly ended up poisoning themselves with their own weapons. I also watched two agents ask each other if a canister they were about to fire at the crowd was tear gas or a stun grenade. (It ended up being a stun grenade that then ignited the tear gas they had already shot at us, which started a fire in the street that a protestor had to help them put out.)
>Brutal state violence and hysterical right-wing internet content work together in lockstep. According to The Washington Post, the agency is under pressure from The White House to create as much content as possible. Which is why ICE agents have a phone in one hand and a gun in the other. But it goes beyond that.
During a showdown with protestors at the Whipple Federal Building in Minneapolis, I watched as one ICE officer fist-bumped a pro-Trump content creator once he learned he was there to support them. I also watched as a gang of groyper livestreamers, led by January 6th insurrectionist Jake Lang, rile up a crowd of protestors, creating the perfect pretext for ICE agents to fire pepper spray balls and tear gas at the crowd.
" According to The Atlantic, they receive 47 days of training — in honor of Trump, the 47th president, naturally."
Okay, I'm gonna stop you right there and ask is this true, or is this just truthiness?
(1) Do they only receive 47 days?
(2) If they do, is it "in honour of Trump, the 47th president" or for some other reason?
(3) Can you, Mr. Blog Post Author, stop annoying the hell out of me within your first two sentences so I can bring myself to read the rest of this?
Because I'm inclined to think this is just propaganda for the progressives, e.g. "Many of them, also, can barely read or write, apparently."
Ah, so they're graduates of West Coast DEI programme schools? (See, two can play at snarking about the outgroup!)
Otherwise, that's snatched straight from this cartoon (nitpick: isn't the Nazi Fascist salute done with the *right* arm, not the *left*?)
https://www.reddit.com/r/ProgressiveHQ/comments/1q9gysh/a_tale_of_failure/
Also, re: "phone in one hand, gun in the other", that's used as a "damned if you do, damned if you don't".
Interactions are not filmed: this is proof they're illegal! if nothing bad went on, why aren't they providing video so we can see for ourselves? what are they trying to hide?
Interactions are filmed and posted online: the monsters are bragging about murdering innocent people! they even film it and then make it public!
here is an article with some sources on the training, ranging from 48 days to 6 weeks.
https://www.aljazeera.com/news/2026/1/13/has-us-ice-officer-training-been-reduced-to-47-days
>News outlets and DHS officials reported cuts to the length of ICE training during Trump’s second term, reducing it from about five months to six days a week for eight weeks. That is 48 days of training over a 56-day period. (What it has to do with Trump’s status as 47th president is outside the scope of this fact-check.)
hopefully this helps in updating your beliefs about ICE training.
So it is *not* true that they only receive 47 days of training and that number was specifically chosen to match Trump being the 47th president.
Given this, I don't think I need to read the rest of the piece, since it's not likely to be greatly more accurate for the remainder.
I don't think that's productive or smart conclusion to draw from the article; in fact, I would say you should update, based on the sources (and the claim from the Atlantic) that it is more likely than before that they do 47 days of training. Note that the sources mentioned did not want to give a specific number of days. I wonder why?
Please read this article; see it as an exercise, if you have a tribal response and don't want to be exposed to information you might not like, regarding the ICE screening procedure. You can of course always say she is lying, or disregard it, or point to DEI, or other people being partisan, if you feel like it.
https://slate.com/news-and-politics/2026/01/ice-recruitment-minneapolis-shooting.html
Okay, Ryan, I will take it into consideration that you were too lazy to fact-check The Atlantic article because truthiness more important. Thanks!
Question: Is 47 days far too little or way too much?
A quick google says that police academy in the US is 21 weeks on average (plus additional field training afterwards), so they are very under-trained relative to an average police officer.
And while you might argue that ICE is a specialized job so they should need less training than regular law enforcement, ICE are being used to kick down doors, run vehicles off the road, deploy tear gas and flashbangs against protesters, and other high-pressure situations where they're likely to get someone killed if they screw it up, or violate their constitutional rights. They can't afford to cut corners here.
(And training is definitely part of the issue - I've seen several articles pointing out that police are trained not to stand in front of a suspect's car during a traffic stop, because that can go wrong in exactly the way it did with Renee Good.)
"ICE are being used to kick down doors, run vehicles off the road, deploy tear gas and flashbangs against protesters"
Is this because local police forces, who should be the ones handling crowd control are refusing/being instructed not to do such work in support of ICE, so they end up having to do the police work themselves?
To my understanding, even if the police were fully willing and permitted to support ICE raids, they wouldn't be expected to take the lead on anything - ICE agents would still be the ones kicking in doors, demanding to see people's papers, and dragging people into vans. So like, it might reduce the amount of tear gas they throw at protesters, but ICE would still be doing exciting and violent things that can get people shot.
(Also, if your policy is "it's okay to provoke massive protests because crowd control is someone else's problem," I imagine that would make you very unpopular with the local police very quickly.)
Probably yes in some sense, no in other senses, and it doesn’t matter in that they need more training no matter whose fault it is that they have the job they have. ICE have been acting in those roles for long enough they could have trained up a bunch of recruits properly since their jobs started featuring those skillsets.
No, that's not the case.
No
If anything, it's the opposite, with the ICE creating massive amounts of new work for local police, cleaning up after the messes they leave behind.
That is actually a great question, because it immediately put me in touch with the fact that I am indignant about the number without having any standard for judging it’s length. My instant feeling is that it’s too little training, but I can imagine getting into being indignant about its being too much. In fact, there’s a lot in the Broderick quote that fires up my indignation at ICE officers but really doesn’t justify that reaction, just sounds like it does. I’m prtty sure itsnot true that many can barely read or write, because a while ago I asked GPT for stats on the groups demographics, and most are from middle class families & many have college degrees. As for many wearing sneakers, seems like they most not have been given appropriate shoes. Sneakers in winter MN are really dangerous — you’d slip very easily on ice, and be in danger of frostbite. As for the camo, that seems fine. After all, they’re not trying to be hard to spot.
As for training time — Well, seems to me most people could learn in half an hour max not to fire pepper spray at close range because of the splashback, and how to tell the difference between a tear gas canister and a stun grenade, so I’m not sure whattup with that. But a lot of the rest of what the ICE guys are facing seems much more difficult that most police work, and police have about 3x as much initial training as ICE guys, and then 6 months or so of being closely supervised before being turned loose to make judgment calls about whatever comes at them. But my guess at present is that however much training these guys had had, what they would need in a Minnesota kind of situation is strong leadership, and someone running on-the-spot training and planning for how to handle various situations that are likely to come up the next day. And some policies, like that guys should do things in pairs.
Camo is probably not proper uniforms, but then again, if there were a specific uniform for ICE, we'd be deafened by the screaming about the SS and look see they're getting their own Gestapo uniforms.
My local cops screwed up the tear gas versus other grenades thing. To the point where they were contradicting the protestors (who had the tear gas in their faces). To their credit, the police did own up to it the next day (someone read the cannister on the ground, and said, "hey, boss...").
These are not the most intelligent people in the universe. We don't hire them for that -- we hire them for "can run for ten minutes without collapsing" (which is a difficult bar to pass these days).
Please compare to basic training in the military, which is 13 weeks at most, and 8 weeks at minimum. So, ICE training is a little shorter, but not all that much.
I agree with the "strong leadership" and planning being necessary (also "doubling up" seems to be a good idea, particularly with the Venezuelan shovelwielders around).
Okay, yes, I agree that comparing it to how long regular police training is makes it look quite short, but as I understand it, most of the new ICE recruits just switched from being active cops – with veterans and ex-cops being next on the priority list – so I think it's reasonable to rely on the overlap in required skills and have the ICE-specific training be shorter in duration.
ICE just hired 12,000 new officers - I don't think they have that many veterans and ex-cops available to fill those numbers.
Veterans: https://www.statista.com/statistics/250267/us-veterans-by-age-and-gender/
Just men: 1 million veterans aged 18–34, and 3 million aged 35–55.
And I can't find stats on ex-cops, but ~800,000 cops in the US, and 1/5 of those are veterans.
I think they have enough. There's this line from the Punisher which I thought was prescient: "They spent fifteen years training an army and then abandoned it on the streets." Someone might be picking it up.
The idea of giving anyone a gun and any kind of authority over other people with less than 6 months of intensive training and a psychological evaluation is absolutely terrifying. 2 years is where I start feeling comfortable.
Cloud Cuckoolander Alert!
13 weeks for the Marines (that's boot camp), and that's the longest our military gives before we issue people guns (Granted, we do not issue them explosives without Written Justification, after several incidents using explosives to clean toilets...).
Remember when Congress wanted the National Guard to swoop in? That's 10 weeks of bootcamp (and since National Guard is often busy driving trucks and fixing flood damage, guns are not a high priority).
> Cloud Cuckoolander Alert!
Please don't do this here. The rest of your comment makes a good point, and snark like this only detracts from otherwise valuable information. Nothing good will come from it, even if it feels satisfying when you type it.
> psychological evaluation
Is this essentially a Buzzfeed personality quiz, but Science™?
Is this a passively aggressive way of saying that there is no such thing as a meaningful psychological evaluation?
Not even on the level of "try to find an obvious psycho by asking them -- in situation X, would you do an obviously psycho thing? -- and eliminating those who are stupid enough to say yes"?
Given the work they're being hired for, "psycho" answers aren't a deal breaker. Maybe keep them out of routine enforcement, but sometimes you need people to do dirty work that the regular hires just aren't psychologically capable of. Last thing you want is for them to get cold feet at a crucial moment.
Every time there’s a topic where there’s a chance to weigh in in favor of doing the harsh, violent thing, whatever it is —dumping the oldsters off Medicare, giving those arrested immigrants some blows below the belt, you pipe right up in favor. And what you post always has the sound of something more than a set of beliefs about how to make society work. There’s this whiff of personal thrill coming off you. You sound like Dolores Umbridge sounded when she was working herself up to using the Cruciatus Curse on Harry and his friends. “I am left with no alternative. This is more than a matter of school discipline, this is ministry security. Yes! Yes!” She was shifting her weight from foot to foot, beating her wand against her empty palm and breathing heavily. “You are forcing me Potter, I do not want to,” said Umbridge, still moving restlessly on the spot. “But sometimes circumstances justify the use . . . I am sure the minister will understand that I had no choice. The Cruciatis Curse ought to loosen your tongue.” There was a nasty, eager excited look on her face that Harry had never seen before.
Look out for the centaurs.
No, not really: every such online I've taken (correctly) identifies me as a Ravenclaw (and a Ross), so they're certainly not entirely random.
Their training is probably sufficient for doing their actual job. It's probably lacking in the "riot squad" aspect because they were never intended to be a riot squad as well. I am sure that if random idiots started obstructing you while you were trying to do your job every day then you would struggle with that as well.
Do these idiots want there to be a well-trained Federal riot squad? Because that seems to be the future they are manifesting.
Yeah, it sure is inconvenient for the police when people exercise their First Amendment rights.
I just don't get it. If their training is not adequate to teach them how (and when) to use guns, why give them guns? If "doing their actual job" does not involve using guns, why give them guns? If "doing their actual job" does involve using guns, why is it OK not to teach them how (and when) to use guns?
Congratulations! In minneapolis, they're rolling out prison guards in SWAT gear. yes, they've had training in how to use batons and control crowds. Training in "not hurting people's feelings"? Not so much.
Their job does involve guns, it just doesn't usually involve facing off against crowds. Usually the situation is a lot more one-sided.
It is the future this administration would like. Imo
Comment on the ICE shooting.
Ross is not going to be convicted of murder, for two reasons:
1. Criminal prosecution would be by the federal government and the people who control it are on his side. As I wrote in a different context, if the King controls criminal prosecution, the King’s friends can get away with murder. So he will not be prosecuted.
2. If he was prosecuted for murder he would not be and should not be convicted, although he might get a hung jury. Shooting the driver of a car who is accelerating towards you is arguably self-defense, arguable enough to not meet the "beyond a reasonable doubt" standard for murder.
He cannot be tried and convicted for the crime of murder. He can be tried, perhaps convicted, for the tort of wrongful death. A tort prosecution is initiated by the claimant not by the state. A tort conviction only requires proof by a preponderance of the evidence.
The plaintiff's attorney could argue that if he was afraid the car would run over him the sensible response was to get out of the way, since shooting the driver would not stop the car, evidence that whether or not he was trying to kill her he was not trying nearly as hard as he should have been not to. He could also point out that the second and third shots were fired after Ross had gotten out of the way.
He might not be convicted and if he was his employer would probably pay the damages, but the trial would generate information about what happened that should make it a little clearer whether and to what extent he was at fault.
Re 1, a subsequent democratic administration could still press charges in 2029 or 2033 though iiuc (unless the Trump doj already presses charges and loses a case, because double jeopardy?)
I can imagine a Democratic president deciding to dig this one up to distract from a new issue.
No way will they bother with pressing charges and losing. Trump will just pardon him instead.
No dice. Trump will give Ross a preemptive federal pardon..
Re 1, a subsequent democratic administration could still press charges in 2029 or 2033 though iiuc (unless the Trump doj already presses charges and loses a case, because double jeopardy?)
I can imagine a Democratic president deciding to dig this one up to distract from a new issue.
There are reports that the ICE agent suffered internal bleeding which, if true, should end the debate about whether he had a reasonable fear.
https://www.cbsnews.com/news/ice-officer-who-shot-renee-good-internal-injuries-sources-say/
Anytime someone is in front of a vehicle in operation, they have a reasonable fear of being struck. Were his actions in shooting the driver reasonable? If he is injured, it is plausible that this further proves his action was not a reasonable response as shooting did not even prevent the injury he feared.
I would also amplify your condition, "if true" and associated doubts. It is an uphill climb to convince doubters that administration sources are trustworthy on this story. Technically, a bruise on the shooting hand would count as internal bleeding, but wouldn't be very significant in how to interpret the event.
If you go in trying to disrupt law enforcement, see the cop in front of you, look him in the eye, slam on the gas pedal then injure him them, then it’s ridiculous to charge that guy with murder.
I bet the "internal bleeding" ends up being a bruise (and of course they can't demonstrate the bruise came specifically from the car, so it'd just be "he has a bruise").
How nice that you can tell what is and what is not a genuine injury from the comfort of your chair without even needing to see any kind of medical report.
You should apply for work with Capita carrying out disability assessments, you sound like the type of person they want:
https://www.bbc.com/news/uk-49208240
https://www.disabilitynewsservice.com/secret-pip-files-show-one-in-three-assessments-by-capita-had-significant-flaws/
I saw that he was taken to the hospital that day. Internal bleeding plus video footage of being hit is more than enough to establish reasonable fear.
Again, probably a bruise. As for the hospital, I haven't seen a source other than admin officials who also said that he almost died and Renee Good was a domestic terrorist and shit.
You won't see another source unless this goes to court. HIPAA. Agree that it's low value data, in that "workplace injuries" are a thing, and someone's gonna want to CYA -- but it probably does indicate, aligning with independent video (not his cellphone), that the guy was hit.
David,
1. There can't be a state prosecution? Is this because federal officials are immune from state law?
2. I do not believe that is how affirmative defenses (which self-defense normally is) work. First the prosecution has to prove beyond a reasonable doubt that the defendant met the standard for homicide. Presumably that is simple enough in this case. Then the defendant may offer an affirmative defense that the homicide was justified. The defendant would have to prove by a preponderance of the evidence that the killing was justified (self-defense).
Self-defense is typically a state law that varies between jurisdictions. If this were a federal prosecution I do not know what law would apply, maybe there is federal law about self-defense in such cases, I haven't looked into this.
My understanding is that states vary as to whether self-defense is an affirmative defense against murder charges or if lack of self defense is an element of the crime. In the former case, the defense needs to show self-defense to some burden of proof. If the latter, the prosecution needs to show beyond a reasonable doubt that it wasn't self-defense. Under Common Law, I think affirmative defenses generally have to be proven to the Preponderance of Evidence standard, but a lot of states have special statutory treatment for self-defense as an affirmative defense.
I think it's technically an affirmative defense in Minnesota, but one with a "burden of production" rather than a burden of proof. The defense needs to make a claim and produce evidence that makes a prima facie case for self-defense, which the prosecution must then refute to a "beyond reasonable doubt" standard.
So, this guy, from the old culture war threads on SSC (eventually The Motte), who is right wing enough to argue the covid and flu shots should not be called vaccines at all (efficacy is too low relative to most things called vaccines), says this was manslaughter, even though the protester has some of the blame.
https://hwfo.substack.com/p/shiris-police-shooting
"There is no country on all of planet Earth, as far as my research has been able to glean, which simply allows illegal immigrants to stay indefinitely once they’ve been identifie
It even 1600s.US?
To summarize one highlight in that linked essay: ICE are enforcement only, they are the snatch-and-grab force. They do not have research, investigation, public outreach, judicial, or general law enforcement skills or authority. This is the wrong force being deployed in the wrong way.
ICE seems to be acting in coordination with a lot of other agencies, including the military. I wouldn't bet that someone called "ICE" is actually just ICE (I got this from reading the filing about the Minneapolis judge, where the ICE guys involved were from multiple agencies).
Also, what exactly do you think our military on "homeland defense" is doing, if not ICE? (Yes, yes, tracking down terrorists on our soil, we can call this ICE and send them HOME).
The military is not allowed to engage in domestic law enforcement.
It's the FBI who was responsible for tracking down terrorists, at least before Trump pulled them off that duty to join ICE.
The military is not allowed to engage in domestic law enforcement. Sure. Flying drones, following money, tracking down illegal gun runners (and then asking the Secret Service to do the law enforcement)... all of this is stuff that the military does do. Law enforcement is defined very narrowly, and when the US military flies combat missions over US soil, that's also part of ICE (combat missions presumably because the Mexican cartels have anti-aircraft weaponry).
So, you're telling me that Marines shouldn't have been asked to "provide covering fire" when they shot over 200 bullets into a gangster's house? **
**This was NOT Trump. It was years ago. There's a reason we don't use military to do law enforcement (including protecting Capital Hill) -- they don't got the training, and ... misunderstandings do occur.
blorbo has already pointed out that the car was not in fact aimed at him. Even the claim that it was is really that it was on a path to slightly clip him at low speed.
These sorts of "analyze police shooting frame by frame" (usually in a situation where if it was not the police shooting a civilian then everyone would agree it was murder), arise from time to time, and we're all used to the beats of the discourse.
But in this case there really is a difference in kind. Usually the threat is a person making threatening statements, having a weapon, etc. In this case it's a car. We've all been around cars, we've all been in situations where for a moment it looks like a car is going at you. Your gut reaction is to move out of the way, not attack/shoot the driver.
Shooting Good was not actually self-defense - I don't mean legally, I mean in terms of the laws of physics. But it's more basic than that - shooting Good is not an act that a reasonable person would do, spur of the moment, to defend themselves either. The defense is usually "well he had to make a life or death decision in a split second, being a cop is hard, you're a civilian you've never been in this situation" but the fact is, I've been in this situation before, and so have you, and we all managed to not kill anyone.
She was gunning the gas, this is not a "low speed" driving accident. And I know someone with training, who was involved in a "low speed" driving accident (as in the car was navigating a parking lot, pulling out before turning onto the street) As per training, he went onto the hood of the car. You're not trained to evade the car, you're trained to roll onto it. He left a handprint in the metal hood of the car.
Everyone involved was embarrassed at the situation (driver did not see pedestrian.)
> We've all been around cars, we've all been in situations where for a moment it looks like a car is going at you. Your gut reaction is to move out of the way, not attack/shoot the driver.
Yeah no shit, I don't walk around holding a gun. If I had one, and didn't need to worry about getting charged with homicide, then the situation obviously changes. It's still self-defense because a dead person has a zero percent chance of intentionally killing you.
"it's still self-defense because a dead person has a zero percent chance of intentionally killing you"
And yet here we have video evidence that shows that the dead person had exactly the same chance of killing him as the live version, since shooting her did not stop the car at all. So you're completely wrong.
The car was not directly pointed at him, and thus he only got lightly bruised. While a dead person can still hold the pedal, a live person can turn the wheel. In hindsight, killing her posed a smaller risk to his life than not... though I doubt he was thinking that clearly at the time. Either way, self-defense is self-defense. He had no reason to kill her unless there was a non-zero chance of her harming him.
Not to mention that the driver "had to make a life or death decision in a split second" as well, but it really rarely comes up.
Perhaps the moral of the story is not to freak out civilians by trying to enter their vehicle, especially when one of your colleagues is in front of their car, and the car has its engine running.
Part of a wider pattern where it's OK for cops to act totally irrationally based on fear but civilians are supposed to handle dangerous, fast-evolving situations with total calm and poise.
If he was in fear for his life, then any civilian who has an officer pull a gun on them is too (including in this case where the cop went for his gun before she started to drive). If we're supposed to treat it differently because it's a cop and not a random person, then that comes with a responsibility for the cop to be judicious with the use of force.
With great power comes great immunity from responsibility.
The thing is that if you think a car is hurtling toward you the last thing you would do is shoot the driver. It makes no sense even on its own terms.
I think you make a very important point that is emphasized by use of the word "King."
If there is an eventual hand-over to a group with a different impression of the events and political preferences, then it is possible prosecution could go ahead. The current path seems to be creating a large block of people in or associated with the current administration that have strong incentives to block a peaceful transition of power.
Between your previous comment and this one, you seem to have gained certainty that Good was accelerating toward the ICE agent. Have you gotten new information that caused this update?
There is something that I am really struggling to understand why reasonable people are leaving out of the discussion: In the self-defense community, there is (used to be?) a concept of stopping power. In bullet vs SUV, the bullet just doesn't have enough. On the face of it, shooting the driver doesn't make logical sense as a reasonable tactic for this situation.
The obvious advice I got as a child is to be cautious about vehicles in operation, stay out of their path/potential path (both front and back), and get out of the way if I feel the vehicle is driving toward me. Hollywood movies aside, it would never occur to me that shooting at the vehicle would help in this situation.
The car was not accelerating towards him. She was turning away and he had walked past the car, he was out of her path when he shot her. When you view all the videos synced up, it is clear that it was utterly unnecessary. https://www.bellingcat.com/news/2026/01/13/analysing-footage-of-minneapolis-ice-shooting/
If you fear for your life in the position he was in, the sensible move would be to step to the side quickly, not pull out a gun one handed to kill the driver.
He killed her because he wanted to and he could.
It was also unnecessary for Renee Good to obstruct justice and resist arrest. It's not unlike a drunk driver who dies in a one-car crash because he loses control of the vehicle and spins into a tree. Should the steering system have reacted more stably to erratic inputs? Possibly, but I'm assigning moral blame to the person who made the deliberative decision to do something dangerous. When you force armed men to make split-second decisions then sometimes you get shot. If you don't want the dice to come up snake eyes then you don't roll them.
I don't think law enforcement has a duty to react perfectly to deliberate and irresponsible provocation.
Law enforcement should always be held to a higher standard than the public. An untrained woman, surrounded by armed men shouting conflicting orders is likely to make a mistake.
An armed law enforcement officer should have enough training and experience not to pull out his gun one handed while trying to film on his phone when he should simple move to the side, note her license plate number and call it in. Paramedics and ambulance crews face more dangerous situations, unarmed multiple times a night. He executed her.
> An untrained woman
Who chose to put herself in harm's way and break the law. I agree the officer should have responded differently but IMO that woman deserves zero sympathy. She wasn't an innocent bystander, she was an intentional provocateur. In a tossup between law enforcement and law breakers I will stand with the law every time.
If this is true then you have essentially fully replaced morality with 'following the law'. Under this conception, you would side with the SS over people hiding Jews. You would have handed run away slaves back to slave owners.
It's a logic that will always allow evil to continue.
This is such an inaccurate and unsupported read of my position that I can't even pretend to assume it was in good faith. You're comparing the SS to an ICE officer who shot someone who violently resisted arrest after obstructing justice. That's ... just not worth responding to substantively.
This is a fully general argument - all people who are stopped by law enforcement are (allegedly) breaking the law. If that's your standard, then you will never have sympathy for any person who gets shot by the police.
No it's not, hence the word 'tossup' which means reasonable uncertainty as to who was at fault. Rene Good wasn't an innocent bystander, she was a willing provocateur who violently resisted arrest. If you're resisting law enforcement then you're just wrong. If you want to make an argument, do it in court. Any defiant interaction with law enforcement is prima facie evidence that the suspect is in the wrong. Now if a cop just walks up to a random person and starts beating them then that's obviously a different story, but there needs to be clear evidence that that's what happened and it's very clear that it's not what happened with Good. Citizens have a duty to comply with law enforcement. Anyone who refuses to do that gets what they get.
Interesting they don't include the bodycam footage.
I remember when a certain people lobbied hard for bodycams, certain that this would prove just how corrupt our police officers are.
I shall not hold my breath waiting for an admission of error.
Just for clarity, it wasn't a bodycam - it's a cellphone.
https://www.cnn.com/2026/01/09/us/ice-shooting-minneapolis-renee-good-cell-phone-invs
"As he approached Renee Good’s vehicle on a Minneapolis street on Wednesday, Immigration and Customs Enforcement agent Jonathan Ross held up his phone camera and recorded video."
https://www.nbcnews.com/news/us-news/cell-phone-video-deadly-minneapolis-shooting-rcna253207
"Cellphone video taken by the officer who shot and killed Renee Nicole Good shows the motorist and the woman believed to be her wife speaking to the agent moments before the fatal shooting."
Doesn't change anything about the debatable positioning of the car, etc, but some people seem to be overclocking on the erratic movement of the camera, without understanding that the camera in question is in his hand, not attached to his body.
If a car hits a camera in your hand, then at *best* the car has narrowly avoided running you over by a few feet. If you choose act with total disregard for the life and safety of an officer in that way, it should surprise no one when they respond with deadly force. Please exercise some common sense here. There is video of the wheels spinning forward while pointed directly at the officer in front of the car. Trying to justify her actions is only going to encourage more people to create dangerous situations where someone could be injured or killed.
At the moment the other officers started confronting the driver, the single most likely time for the driver to panic and do something stupid, the officer walked from the side of the car, where he was well out of danger, to directly in front of the car, . The officer is the one disregarding his safety.
If you reread my post, I expect you'll notice that I confined myself to correcting a factual error about the source of the footage (i.e. bodycam vs. cellphone). I didn't weigh in at all about whether the shooting was justified or not. That was intentional, as this strikes me very much as a grey area case.
If I come to a firm conclusion about it later, then I may or may not decide to post something about it then, but in the meantime you'll be better served not to point your own wheels at people unless they've actually expressed the opinions you disagree with.
I've seen the body cam footage. He was in no danger, certainly no danger that stepping to the side would not have solved.
Coming from a country where the police almost never carry guns, I'm eternally fascinated by the American right's ability to justify allowing authorities to execute their own people in the street.
Oh, are we going there? Shall we talk about trampling entire groups of schoolgirls, in an alley? And then turning around and trampling over them again? With no consequences for the officers?
Disclaimer: This is not America. This is one of those countries where the officers "almost never carry guns."
The police don't even try to silence the eyewitnesses, they're so confident of their ability to evade punishment.
This is one of the clearest, most direct examples of whataboutism that I have seen in some time. It's vague as well, inviting another layer of distraction from the original statement; what trampling? What country? How is it relevant?
Please don't answer these questions as it would distract further; I don't really want to spend energy on the deflect.
If it's the first-person footage going around, I've heard that's from his phone, not his body-cam.
You're right. I stand corrected.
I'm eternally fascinated by the number of people who come from a place with different conditions to the place they're discussing (in this case, a much lower crime rate) and somehow think this makes them more, rather than less, qualified to pass judgement on the latter place.
I'm also fascinated by how these people almost never specify the actual country they're from. It's always "a country", their nationallty is always "non-American", never their actual nationality. It's very weird; I can't imagine describing my own nationality in such a way. A lot of the American left has this fantasy of this hypothetical non-American western country which has open borders, and lenient policing, and unrestricted abortion, and large wealth taxes, and massive affirmative action and racial diversity, and every other progressive policy that pretty much don't all exist together anywhere. This isn't that surprising; it's just the equivalent of the right's fantasy of some unspecified past time when every one of their preferred policies was in place at once.
What IS surprising is the number of (allegedly) non-American leftists who want to identify as citizens of this imaginary country. Which of course means never specifying or talking about their own real country (only ever talking and complaining about America), probably because doing so would make it clear that such a place doesn't actually exist.
It does nothing but make accurate discussion of real problems and policies much more difficult.
Actually I'm just British, but I enjoyed your paranoia. I tend to keep personal details out of public discussion out of habit. Our police do not carry guns except for very specific situations such as: guarding government facilities that are likely terrorist targets, responding to active firearm/bomb threats, raiding buildings suspected to house active threats.
It is so rare for police to shoot someone that it is always national news and it is always the subject of an investigation. The shooting of Jean Charles de Menezes in the wake of the 7/7 bombings was a national scandal that would have been shrugged off in the us, and he wasn't even a citizen.
The American right will always bend over to accept the police killing citizens because the central tenet of conservatism is that there is a group the law must protect, but not bind and group that the law must bind but not protect.
> I'm also fascinated by how these people almost never specify the actual country they're from. It's always "a country", their nationallty is always "non-American", never their actual nationality. It's very weird; I can't imagine describing my own nationality in such a way.
Tell me you are an American without telling me you are an American.
I am from Slovakia 🇸🇰, which I sometimes mention when it is relevant, but mostly don't, because I assume that 90% of readers couldn't tell a difference between Slovakia and any other Eastern European country -- and there is no good reason why they should. It's practically the same level of detail as an American telling you which *state* they are from.
(Plus there is a small but nonzero probability of someone asking what is the difference between Slovakia 🇸🇰 and Slovenia 🇸🇮, which is annoying for people from both countries. But it's getting less frequent recently, probably because people learned to google or ask a chatbot instead of asking in the forum.)
Anyway, how this relates to the police behavior...
I do not travel much; I regularly visit Czechia, Hungary, Austria, Switzerland. From the perspective of "what behavior should you expect from the police", I haven't noticed any difference. Maybe I was just lucky; but in my model of the world, it's just "standard European police behavior". Perhaps if you break a rule, you are more likely to be asked for a bribe the more eastwards you are; I wouldn't know, I usually don't break rules.
I spent a few weeks visiting USA, specifically California; during that time I got yelled at by a random policeman for the "crime" of keeping my hands in my pockets (I wasn't even aware that I was doing that), but luckily I didn't get shot (it took me some time to realize that the guy is yelling at me, and what exactly was his problem), so I am here to tell the story. Now I don't want to draw unnecessarily big conclusions from a single data point, could be a total coincidence, but even my limited experience agrees with the idea that there is a "generic American cop behavior" which is significantly different from a "generic European cop behavior". So if other people keep saying the same thing, it sounds plausible to me.
Oh, and people own lots of guns in Switzerland, so I think the classical excuse "American cops have to be different, because Americans have lots of guns" fails to explain this part.
I agree that by cherrypicking things from different European countries, you can get an idealized country that is much better than any of its constituent parts. But I think there is also some meaningful similarity, and America is culturally more different; sometimes for better, sometimes for worse.
EDIT: https://www.astralcodexten.com/p/open-thread-416/comment/199631031 " Hell, there are plenty of well-documented incidents of LEOs unintentionally injuring or killing people during physical struggles due to a basic mistake, like drawing and firing their pistol when they intend to use their taser." For Americans, this is Tuesday; I think that for most Europeans this sounds like a quote from a dystopian novel.
EDIT: The hypothetical country that America is compared to seems to be something like "mostly Europe, but also Canada, and excluding e.g. Belarus". Perhaps we should have a simple word for that.
Slovakia has babka but Slovenia has potiča, right?
I have watched several French policemen with batons beating on an 8 year old in the fetal position. This seems like dystopia to me. I do not believe they will suffer any from doing this.
In Great Britain, a schoolgirl was arrested for carrying a knife to protect her kid sister. (The video is from adults harrassing the children). The inevitable occurred in the aftermath.
In Italy, someone's on record saying that a rapist "cannot be expected to not assault Italians on the beach" because "he doesn't know better!"
I have heard that in Hungary, it is a VERY bad idea to smile at the policemen (seen as very disrespectful, may lead to arrest). Can you confirm?
The German Government (Okay, one particular highly-placed guy in the government) has had SWAT teams called over 5000 times, for people saying mean things online. This seems like dystopia to me.
Perhaps eastern europe is just more functional?
American cops are some of the best in the world (Yes, there are bad apples, and really bad guys getting passed from town to town, but still!), and part of that is our willingness to have there be actual consequences.
>somehow think this makes them more, rather than less, qualified to pass judgement on the latter place.
If you can see something working with your own eyes, that is better quality information than rumours that it doesn't work.
Wouldn't it just be manslaughter.
One side of the political split claims that she did not pose an imminent threat, the other side that she did, was about to run over the agent who shot her. I have viewed multiple videos and am still uncertain how much danger he was in. I have observed, on DSL, intelligent people who are sure it was legitimate self defense and intelligent people who are sure it was not. I have also observed blatant lies in both directions, by Trump and Vance on the one side and by people who simply don't mention that she was driving a car that was accelerating in his direction when she was shot.
If you have not observed arguments about this by intelligent people on both sides you should not assume that you know what happened.
Watching the officer's cellphone video, I think it's of note that the first time the officer walks in front of the vehicle, Renee is turning the wheel to follow his position.
I think she was reversing at that point.
I've only watched the one angle; from that one it didn't look like she was moving at all. It may or may not have been incidental, as she leaned out the window to talk. But it's something that can put the guy in the mindset of "she wants to run me over."
She wasn't pointed at him, reverses, points the wheels at him, looks him directly in the eye, and then guns the car (skidded out on the ice). This is what a video analyst had to say.
Mental state unknown, may not have intended to run him over (wife shouting "drive baby drive", for example) -- but him being in the way was not a disqualifying "why I shouldn't gun the gas"
He made a loop of the car before that, that's the one I was talking about.
You have a link to the video analysis? Because it sure as hell doesn't match the videos, or the mindset.
To be clear, I am doubting explicitly the "gunning the gas"
https://boriquagato.substack.com/p/lets-talk-about-renee-good
This rundown jives with my personal interview of an expert witness (has given court testimony before), whose expertise is in video analysis.
(My use of the word "gun" was perhaps needlessly inflammatory. The bad cat's saying "hit the gas" which is perhaps more accurate).
How was she "accelerating in his direction" if her tires were pointed away from him?
You can see for yourself in the video that the wheels are pointed directly at him and spinning forward when he draws his weapon.
I’ve seen the videos, but I’m open to being wrong. Can you send a link, or an image?
https://www.reddit.com/r/law/comments/1q6p4uv/the_minneapolis_woman_who_was_shot_in_the_face/
You can see at the 28 second mark the wheels are already spinning while pointed in his direction. It's clear because we can still see the hub caps. She appears to be in the process of turning them, but has already started accelerating in his direction when he pulls his weapon at 29 seconds. By the time he fires at 30 seconds they are pointed away, but at that point the decision to fire has already been made. He didn't have the luxury of doing a sub-second breakdown of the wheel movement at that time.
I disagree, he definitely had that luxury, especially since, as you said, the car was turning before the weapon was drain. You’re calling the agent so negligent as to ignore the path of the car, which is nearly as bad.
Reaction time is .3 milliseconds, and that's for a quick draw. Give him a full second, because he has to drop the cellphone, draw -- and aim at a person in motion.
I'd say that's within "give him the benefit of the doubt" -- if I was in a life-and-death situation, I might pull a gun, even if it was an objectively dumb thing to do.
Cars are not zero-width objects.
And yet, nobody was hit!
(1/16) Edit on top of post for visibility: Active Self Protection has just posted a comprehensive non-partisan analysis by a professional forensic expert witness in shooting encounters. It includes all the useful raw, unedited footage available, and then his analysis. He addresses topics like why Ross was likely *unable* to stop firing after his first shot, and why it's likely Good never even saw him when she took the car out of reverse to drive forward.
It's *by far* the best coverage and analysis of this event I've seen yet: https://youtu.be/6k_1y2kSHfw?si=49CESHk_EDBd_4HJ
***
Some reasonable pro-guns-for-self-defense folk I follow on YouTube had hot takes that the driver was *technically* in the wrong just enough that the officer shouldn't (and won't) face criminal charges as he can credibly state he was in fear of his life while/after getting struck, but also that the shoot wasn't as "good" as we'd all generally like to see in a self-defense encounter. The word "unfortunate" is getting tossed around a lot, and that's about where I am.
But really?
It's not about the ICE officer.
Ultimately, the driver had a duty to herself and her family to not *deliberately* and *avoidably* endanger herself. As I stated on the Hidden Thread and will cannibalize for here, the driver was incredibly foolish to feel so secure about defying the commands of people legally carrying firearms to enforce those commands. Plenty of instances of police abuses have been covered in the media (and we can assume she was exquisitely aware of them), and there are tens of thousands of badgecam videos on YouTube demonstrating the immediate consequences of defying police officers. Hell, there are plenty of well-documented incidents of LEOs unintentionally injuring or killing people during physical struggles due to a basic mistake, like drawing and firing their pistol when they intend to use their taser.
Illegally confronting the police is *dangerous.* Rarely, legally confronting the police is dangerous! Don't fucking do it, no matter how passionate you feel about your cause!
Yes, this is victim-blaming, and it is absolutely warranted, because the driver actually is to blame for foolishly risking her personal safety. She would be equally to blame had the video been of hopping the guard rail of a Grand Canyon viewpoint to get the perfect pic and then falling to her death, or chasing down a wild grizzly bear cub to pet it. There are dumb ways to die, and this was one of them.
There is footage going around of Renee's wife (IIRC) shortly after the vehicle came to a stop. She's yelling at the officer: "Why did you have real bullets!?"
If that's an accurate transcription, it's consistent with and highly suggestive of both of them having a very inaccurate idea of what LEOs normally do. It makes me think they believed they were working under completely non-lethal rules of engagement. That would mean they could do whatever they wanted as long as they knew they weren't trying to hurt anyone.
This also leads me to believe that while Renee Good may have held ICE agents in contempt, she had no inclination to actually hurt them, including hitting them with her car. However, she could still be found to have had reckless disregard for safety. I could be persuaded otherwise, but the footage I saw suggested she was looking at an officer at her 9 o'clock while pulling away, and failed to see Ross at her 12. In a residential street with no LEOs around, I could see making that assumption. In one where I've gone out of my way to obstruct traffic, I can see that not being a useful defense. (And on a street covered in snow, it's even less defensible. What if that had been a kid chasing a ball?)
I hadn't heard the bullet thing. If that isn't some deep fake prank, that's a truly *astonishing* disconnect from reality.
And I agree, it's likely Good never clocked Ross when her focus was elsewhere.
This is completely correct. She willingly courted danger and suffered the consequences. The risk of being shot by cops is why you don't provoke them. Don't force an armed man to make a split-second decision unless you want to run the risk of getting shot.
Of course the shooter had a poor reaction and made a poor choice. I would love it if every police officer reacted perfectly 100% of the time, but that's simply not realistic. I would assign the blame here as 90% Renee Good and 10% ICE. If all police officers were replaced with perfectly-trained replicas then I expect that that would reduce incidents like this by 10%. If all protestors simply obeyed the law then I expect that incidents like this would be reduced by 90%.
That's funny, I actually drafted in a different comment that I felt it was about 90% Good's fault, but then got worried I'd have a big fight about the particular number, and I'm not really married to it enough to want it to be the main focus of a comment war, haha.
While I don't necessarily disagree with you, I was very surprised to hear this from the same person who claimed that guns would protect women from the patriarchy. Well, at least it means that you won't throw away your life if and when things escalate...
Hmm, I seem to recall saying that it's odd feminists don't make better use of guns if they're so worried about the patriarchy, but I can't find the exact comment.
But also, as a rational hedonist who firmly believes that the quality of my life is more important than minutes lived, I'm fully prepared to use the tools I have available to me to end my severe suffering should there not be a strong likelihood of it being relieved. I don't like cooking from scratch and cleaning enough to be a Martha, and I would absolutely provoke a shooting before allowing myself to be transported to the Colonies.
Ah, if only suicide was that easy. Doesn't help that the success rate is so much lower for women too. Good luck with that.
Well, now I can't tell if we're flirting or not, given the exquisite setup you just provided me for a very funny, very tasteless joke that would be very fun to post.
But while it appears Scott isn't banning people anymore, the chance of getting reported and then banned for my joke is nevertheless nonzero, so I'll hold my tongue and hope those who would appreciate said joke will likewise see the setup and get a chuckle out of it.
I would love it if you DM me the joke. I love funny and tasteless.
We typically hold the police to higher standards than this; if we didn't, police would be racking up double digit fatalities from knuckleheads and drunk guys confronting them every long weekend.
In other words, the ICE officers are (or should be) trained LEOs who are safer to be around than grizzly bears.
They are. I would much rather be around an ICE agent than a Grizzly Bear. Bears attack people indiscriminately. All I have to do to avoid even the threat of being attacked by an ICE agent is take the exceptional step of not actively interfering with them in the performance of their duty.
Citizens shouldn't obstruct justice or resist arrest. When you force armed men to make split-second decisions then you run the risk of getting shot in the head. We typically hold citizens to higher standards and expect them not to actively interfere with law enforcement.
We also generally ask that knuckleheads and drunk guys not actually use broken bottles and other lethal weapons. Fists are harmless enough (nonfatal) that a "bearhug takedown" is a reasonable level of defense.
By the point that a lady is using a 2 ton vehicle, you've moved well beyond "let's just toss the guy to the ground" and into "that guy is TempleOS crazy."
> In other words, the ICE officers are (or should be) trained LEOs who are safer to be around than grizzly bears.
Yes, fine, great. As a civilization, we should get right on that.
Until then, *right now,* confronting trained LEOs - especially with illegal means - comes with the risk of being unjustly arrested and/or injured, or killed.
If an individual does not freely accept the risk of being unjustly arrested and/or injured or killed, they must not confront trained LEOs with illegal tactics, *ever.*
You think I'm being very naive, but I think "cops shouldn't just shoot people unless they have a very good reason" is more important than "people shouldn't provoke cops", even though both are generally true.
And?
Your feeling about the relative importance of those two principles doesn’t have any impact on our shared physical reality where cops are occasionally wrongly shooting people.
What impacts reality is your actions.
If you feel so passionate about police overreach that you're willing to *consciously* risk your life confronting police on camera for the cause, that's something I can sincerely respect (unless you have kids or other dependents). I think it would help if you record a video explaining you fully understand the risk you're taking for your principles, so people won't suspect you felt so entitled to ignore the obvious danger that you might have made a surprised Pikachu face when the bullets hit.
But if you just thought, "What? No, of course I wouldn't risk anything in *my* real life proactively confronting a LEO, not in actual, physical, real-real life where I could lose my job or be really hurt or *die,*" well...
...what are we even disagreeing about?
"If you feel so passionate about police overreach that you're willing to *consciously* risk your life confronting police on camera for the cause,"
The fact that you apparently take it as read that "confronting police on camera" constitutes an automatic risk to life and limb is exactly the problem.
I don't think it's naive or idealistic to believe that nobody in a civilized society should have the right to summarily murder anyone else. Obviously people have the right to defend themselves, but the gulf between "confronting somebody" and "immanently threatening their safety to a degree that requires lethal force to counter" is generally quite vast. It certainly was in this case.
OK, I feel like the guy in Matt Bors' "We should improve society somewhat" cartoon now.
> Yes, this is victim-blaming, and it is absolutely warranted, because the driver actually is to blame for foolishly risking her personal safety.
Everything else aside, a country where this is the "default" take is not the country that I want to develop for my community and my kids. Police (of which, it should be noted, ICE is not) should have *more* accountability, than the random strangers they interact with, not less.
Yes,. Great. Fabulous.
Do you want your kid to reduce their chances of being killed by law enforcement or not?
Because they're the ones who have the most control over whether or not they choose to initiate *inherently dangerous* confrontations with law enforcement. You should probably tell them not to do that if you haven't already.
Your righteousness is misplaced. Less of this please
Your comment was designed to provoke rather than engage in useful discourse. Less of this please
???
(please insert caveats about this being sincere)
I am curious about what window of time supports the claim that Good was illegally confronting the ICE agents.
The videos I have seen begin by showing a vehicle leave the scene by driving in front of Good's Pilot, then Good waving the ICE vehicle to go around her. My understanding of the facts:
(1) ICE was not in the middle of conducting an immigration enforcement action
(2) ICE does not have authority to conduct other forms of law enforcement action, including traffic citations
(3) Good's vehicle was not preventing the ICE vehicle from proceeding.
(4) ICE agents left their vehicle to engage with Good.
(5) The shooter (presumably Ross) circled Good's vehicle, ultimately positioning himself somewhere to the front of her Pilot (but definitely not squarely in front of the vehicle.)
(6) Three shots were fired
(7) The Pilot continues to the right (direction of traffic on that street), straightens and hits another parked vehicle
(8) ICE agents, including the shooter, were in no rush to render aid to Good
Some things that are unclear:
(A) Did Ross transfer his phone/camera from his right hand to left hand? This appears to be the case from the video he took and released
(A') Does A indicate any form of pre-meditation
(B) Why was Good's vehicle in that location?
(B') Did Good engage with ICE/other LEO prior to the first vehicle driving away/around her Pilot?
(B'') Was her prior engagement illegal?
(C) What orders were being given to Good?
(C') Were the orders being given to Good coherent from one agent to another? In other words, was compliance possible?
(D) Were the post-shooting actions of the ICE agents consistent with existing laws and policy?
(E) Was Ross struck by the vehicle?
(E') If yes to E, was it possible/reasonable for Ross to avoid being struck by stepping away from the vehicle?
My current impression is that:
(I) The shooting does not appear justified
(II) The take-away for people who want to avoid this outcome is a blanket "avoid all contact with ICE/LEO." However, without knowing what led up, it is hard to know what action by Good would have achieved this.
Good's vehicle was fully operational, as was demonstrated when her foot hit the accelerator and crashed the car into another vehicle. She wasn't broken down in the middle of the road or stuck in the snow; she was deliberately blocking a roadway to intentionally confront law enforcement.
That is very dangerous for the reasons I outlined, and she could have prevented her death by choosing not to do any of that.
That fact doesn't really have anything to do with whether the shooting is justified. But I'm emphasizing it because she deliberately put herself in danger, and that's bad and stupid, and the only useful takeaway for the rest of us is *don't antagonize people with guns,* because they might, rightly or wrongly, use them to kill you.
Clearly, not enough people are getting that message.
I am very confused. I have trouble lining up your answer with what I wrote.
>>Good's vehicle was fully operational,
As far as I understand, this is not disputed.
>>when her foot hit the accelerator and crashed the car into another >>vehicle.
Crashing into the other vehicle was after she was shot, no?
>>she was deliberately blocking a roadway
From the videos I have seen, this doesn't appear to be true.
Repeating what I wrote above: "The videos I have seen begin by showing a vehicle leave the scene by driving in front of Good's Pilot, then Good waving the ICE vehicle to go around her." In case that wasn't clear: Good is seen waving the ICE vehicle to follow the vehicle that has just drive up the road, showing that her Pilot was not blocking the road.
Is the description of the situation before the ICE agents got out of their vehicle and approached Good not correct?
It's not that confusing.
Good voluntarily placed herself in great danger.
She had a duty to herself and her family to not voluntarily placed herself in great danger.
Good should not have voluntarily placed herself in great danger.
"Good voluntarily placed herself in great danger. "
What, exactly, what she doing that was "voluntarily placing herself in danger?" Existing in her car a few blocks away from her house?
The way you wrote it ("Good's vehicle was fully operational, as was demonstrated when her foot hit the accelerator and crashed the car into another vehicle.") read somewhat confusing to me as well. It reads as if she hit the gas and immediately crashed into another car. But most people referring to her hitting the gas and almost immediately turning to head down the street. She only hit another car because she was probably dead by then, and her SUV was presumably an automatic.
I don't think it functionally disputes the rest of what you wrote.
I agree that she did something stupid, but I do not think that her doing something stupid is enough cause for absolving the officer of any blame on the incident.
I don't think the officer should be absolved, either (and less so today, after watching an analysis I hadn't seen before); at the very least he should lose his job and pension and so on. Perhaps criminal charges, although which ones, I don't know.
But the reason I'm emphasizing Good's stupidity is that it's the only useful object lesson *we* can take away from the incident. None of us have the power to change the policies at ICE, or ensure our police departments only place emotionally-regulated, well-trained officers on duty.
What we *can* do is spread awareness of the inherent danger of confronting a random LEO, and especially confronting them while breaking the law, when they might feel even more justified using force than they would if you were indisputably correctly exercising your 1A rights standing on the sidewalk and so on. We should break people's sense of entitlement to not have to feel any fear of LEOs, because no one is so important to the universe that their passion will plot-shield them from the reality of other people potentially reacting very negatively to *SUPER!SPECIAL!THEM!* should they start a fight.
Huh. Something wonky is going on with Substack! Obviously, I finished the sentence in my final paragraph with "one of them," however, Substack is apparently refusing to grant me another line of text to show it.
I've checked on Firefox and Chrome.
Weird!
I notice it too, and there is another comment in the thread where precisely that happens. My guess is that the length of your comment was just on the edge of when it collapses with a "See more" tab, and they haven't coded that correctly.
It’s there. I see it.
Well, now that's really strange - can I ask what platform you're using to view it?
Nate Silver pointed me to a great article on this: https://www.historyboomer.com/p/rashomon-in-minneapolis
I think I’ve seen all the videos and I can’t really tell whether the woman was on course to hit the ICE guy. But I can easily believe that in the stress and confusion of the situation the person with a car even sort of coming at them would be quite frightened, and fully believe they were in danger of being hit.
The fault, if there was one, was in the ICE guy’s training. In the places where these confrontations are taking place having somebody aim their car at you is probably one of the likely dangerous situations. And there are a things a trained person would know that would influence how they act. For instance, how badly are you likely to be hurt if the car is going 10 mph or less? Also, if the car is coming right at you, at any speed, shooting the driver isn’t going to help you. Even if you kill the driver, the car’s still going to hit you at whatever speed it’s been going. Seems like your time is better spent getting out of the car’s path.
This professional forensic expert witness on shootings and other violent encounters thinks Good almost certainly had no idea Ross was in her path (commentary on Good's likely tunnel vision begins at 23:44):
https://youtu.be/6k_1y2kSHfw?si=49CESHk_EDBd_4HJ
10mph is a six-minute mile, so it's like hitting a waist-high wall at full run.
Which is why he should have never been in front of the vehicle. If they wanted to stop her from running they could have parked a car there, not a human body.
And why exactly would they want to stop her from running? As I understand it, the problem was that her car was in the way, and she was bothering the agents. If she drives away, both problems are solved. It's not like there was a warrant out for her for some serious crime, so God knows what happens if she escapes or something.
(The answer is probably that the agents switched from a mindset of "we need to get her to leave to we can do our work" to "she's the enemy, let's take her down".)
She was behaving as if it was "if she leaves, no harm no foul."
But this is an iterative game, and this wasn't the ICE agents first playthrough.
If they, upon the fiftieth time, decide "jailtime for Miss Nancy Drew", that is their right as law enforcement officers trying to do their job. She's in the wrong, and interfering with lawful enforcement of the law.
Time was, the left knew that "going to jail" was what would happen if they staged a sitdown. Letters from Birmingham Jail, and all that, no?
>If she drives away, both problems are solved.
Well, they're 'solved' in the same way that making a thief return the goods they just stole is 'solved'. She was deliberately blocking traffic; I think a night in jail would have been a reasonable outcome.
Is conspiracy to obstruct a federal law enforcement operation not a serious crime?
Because these idiots are doing a "rolling blockade" where each car obstructs the agents on one block until forced to move on, then another one obstructs them on the next block. if they don't arrest her then she'll be harassing them all day.
With a dead driver, the car is less likely to swerve towards you when you dodge out of its way.
Not really relevant when the car is as close as it was in this case, but I can see it being standard procedure one is trained to reflexively execute (no pun intended).
Unless the driver turns the vehicle AWAY from you, which was the case here. (...see the cellphone footage...) Shooting the driver can, however, stop this turning, thus you will be in more danger.
It was clearly a reaction fueled by rage. Which can considered to be acceptable ("he had to decide within a split of a second" and all this stuff), but whoever reacts this way ("within a split of a second" or not) definitely should NOT carry a gun.
(BTW the driver had to react "withing a split of a second" as well -- perhaps federal agents just should not scare the hell out of everyday citizens? Just an idea.)
> Unless the driver turns the vehicle AWAY from you, which was the case here. (...see the cellphone footage...) Shooting the driver can, however, stop this turning, thus you will be in more danger.
But the car wasn't pointed at him, hence why he only got scraped. As far as self-defense goes, killing her was not that dangerous.
I don't think the driver gets credit for anything she did with the steering wheel when she hit the gas pedal.
I have no love for the state, and would condemn anyone for willingly acting in its service. But even I do not say they deserve to be deprived of so fundamental a right as effective self defense.
The context is this sentence:
"With a dead driver, the car is less likely to swerve towards you when you dodge out of its way. "
I argued that this simply did not hold in the current case, and I fail to find any trace of you responding to my comment.
So, in your opinion, how many of the 47 days of training ICE agents receive are dedicated to shooting people in cars moving toward you?
I didn't design their curriculum, and don't know whether the person who did is an idiot, or what constraints they were operating under, but if I was doing it, "draw quickly, keep hands steady, and fire rapidly," immediately upon recognizing a threat would be a skill I'd drill a lot, below only general fitness and target practice, assessing it to be applicable in a wide range of situations.
Training specific to cars being used as weapons is obviously harder: it's clearly insane to drive normal cars at speed towards your trainees to teach them how to react. If I could have custom autonomous cars built that stopped much more aggressively, maybe with explosive charges shooting spikes into the road below or something as emergency braking, then yeah, I could see a few days spent on that. That sounds fun! Just imagine standing in the middle of a field with cars barreling towards you from multiple directions, and you have to dodge them, and "kill" their (imaginary) drivers.
Also with a dead driver it at least isn't going to turn around and try to hit you or one of your colleagues again.
At least in principle, shooting the driver while they try to reverse in the middle of a city street could be an effective countermeasure, in a way that shooting a driver who is a few feet away and "comin' right at us!" really, really isn't.
But it would be a countermeasure to something that basically never happens. People occasionally kill police officers with automobiles, but as a one-and-done in an attempt to escape. The sort of persistent attack you're describing, is vanishingly rare.
If you're training cops to not die, you teach them to get out of the way of the car. If you're training cops to get away with killing people, you teach them to stand in front of the car and say the "he was comin' right at us" thing. And yes, if you're a really stupid police trainer, you might confuse the one with another and train your students to kill and/or die needlessly. Alas, it seems to me that there are a lot of stupid police trainers out there.
Sometimes I think about things, and think "that would be a good subject for an ACX open thread" and then I forget whether I already started a thread on that very subject. Here is one such subject, possibly a repeat.
What's the relationship between ethics and etiquette? Are they two distinct things, or is it more of a continuum? Both are lists of things that you should and should not do, ranging from "thou shalt not kill" to "don't butter your whole bread roll at once". Both, if obeyed perfectly, make life more pleasant for everybody.
But there's differences too: ethical lapses tend to be more serious than ettiquette lapses, and ettiquette is more culture-dependent while ethics tends to be more (but not entirely) universal.
And then there's grey areas; like using a racial slur. Everyone would agree that this is impolite, while some might argue it's actually immoral.
Can we meaningfully distinguish ethics from etiquette? And if we can't, does that mean we should pay more attention to etiquette?
It's all a continuum, the same thing happens with taste.
Etiquette lets you know what to do in different situations so as to avoid giving unintended offense; we no longer have a generally accepted etiquette in this country, so it's hard to tell if someone is deliberately being rude to you, just being antisocial, or completely clueless if they block a doorway looking at their phone or are playing loud MMA videos with no earbuds on the bus.
In a well-operating etiquette system, ethics doesn't come into it, everyone is supposed to be doing the "right thing" and you can rely on that.
Many people who focus on ethics put way too much emphasis on it compared to other things we care about. Ethics is not the end all be all of how to live your life. There should be more philosophical inquiry in to a broader range of questions related to how to live your life.
"don't butter your whole bread roll at once"
Is this really going to make “life more pleasant” for anybody?
Would anybody even notice?
Part of the purpose of etiquette is as a class marker.
People like having class markers in the absolute (even the ones who don't like some specific class markers) and having them it makes their life more pleasant. The proof is in revealed preference : whenever class markers get suppressed some others appear spontaneously.
I think philosophers, while distinguishing prudence/self-interest from ethics/morality, sometimes distinguish norms/etiquette from both of them. Other times it's folded into one or both of the former categories.
It's definitely a weird category. Probably a very broad category, as you note, parts of which are firmly under "prudence", parts of which are firmly under "ethics", and parts of which are arguably...if your philosophy allows for some sort of irreducibly social aspect of what a person is...in a distinct category that's not quite either. (Roughly: you in some sense "have to" follow norms even if they don't benefit you, but not doing so isn't *immoral*, it's some sort of separate, purerly social failing.) And I've even seen some who seem to consider it part of *aesthetics*.
Never thought about that before. I'd say it's a continuum, but the big difference is a breach of ethics incurs punishment. If you put your elbows on the table while buttering your whole bread at once, it's a breach of etiquette; you'll make people uncomfortable, but they probably won't throw you out of dinner. If you steal a fork, it's a breach of ethics; if they catch you they'll get the fork back with interest.
For a moment, instead of 'ethics', consider that modern usage of the word ‘value’ keeps confusing different things:
- What people’s motivations show as ‘desirable’ (‘These goods will fetch good value in the market’; ‘The community showed how much it valued the church through high attendance’)
- The pluralistic and contextual norms different societies use to help them coordinate and succeed (‘These values represent who we are’)
...and what i argue( at https://pragmaticfutures.substack.com/p/defending-life-years-as-the-primitive) is the universal source of value:
- the lives that make up a community or polity, now and in the future
So in this view (https://pragmaticfutures.substack.com/p/heirs-of-life-years-a-meta-ethics) you either consider a given set of etiquette norms:
- part of the pre-rational, sacred boundary of the community (e.g. part of their heir-def)
- or a pragmatic tool believed to extend the communities life-years into the future (especially under uncertainty/hazard/'defector risk')
Or less to ethics in private, where there is no one to signal to.
How does not buttering your roll all at once make life more pleasant for everyone?
I think there are two distinct types of etiquette. One is sort of a mini-version of ethics, concrete rules rooted in more general ethical guidelines, specifically meant to make life better overall. E.g., putting away your weights and wiping machines after you're done with them is part of gym etiquette meant to make the gym a more pleasant and safer environment for its members. That sort of etiquette we should pay attention to.
The other type is barely related to ethics. It's about distinguishing the outsiders from those who belong. Elaborate silverware rules during fancy dinners are an obvious example.
I think a rule about avoiding racial slurs is definitely driven by ethical considerations, so I don't think it's a grey area.
“ E.g., putting away your weights and wiping machines after you're done with them is part of gym etiquette meant to make the gym a more pleasant and safer environment for its members. That sort of etiquette we should pay attention to.”
I’m not sure if this is an instance of etiquette. (It’s a not “using the wrong fork” kind of thing.)
It’s an example of “leaving things as you found them”. That is a fairly pragmatic and efficient way to deal with shared resources. Otherwise, one makes work for other people.
> How does not buttering your roll all at once make life more pleasant for everyone?
agree with what's been said - but also note that many evolved systems are 'over-engineered' to survive tail events (hence two kidneys)...thus will look like they "don't make sense"...*even on a normal day in their ancestral environment*. A small 'lapse' in coordination may make a big difference on 'game day' if the stakes are existential. And that's not even taking into account the fundamentally zero-sum nature of status dynamics (at least, along a given dimension).
But fundamentally - my prior is for most ppl you're thinking of, you're trying to reason them out of positions they never reasoned themselves into...appealing to e.g. material abundance/'pleasantness' empirically will fail. Happy to be proven wrong of course.
Not really, rules like what foods are ethical to consume (such as the kosher rule against odd-toed ungulates), what beverages are ethical to potate (like the Mormon prohibition against alcohol and "hot drinks"), what words are ethical to say (like the taboos in Polynesia), what clothing is ethical is wear (various detailed specifications about colors and materials), are fundamentally about distinguishing insiders and their class, though of course they traditionally confabulate reasons involving religious/spiritual/ethical considerations.
> How does not buttering your roll all at once make life more pleasant for everyone?
A lot of table manners are about signalling to your dining companions "I'm like you, you're like me, we were all brought up in the same class of the same culture and all follow the same semi-arbitrary rules". Asking why that's pleasurable is like asking why dining together is pleasurable to begin with -- I don't really know but it really is something that humans like.
The idea that certain types of ettiquette exist purely to exclude lower-class people is, I think, wrong. High-class people would absolutely _love_ it if low-class people started exhibiting proper table manners.
“ The idea that certain types of ettiquette exist purely to exclude lower-class people is, I think, wrong. High-class people would absolutely _love_ it if low-class people started exhibiting proper table manners.”
Is anybody claiming this idea?
In any case, there’s no actual contradiction here.
Things (not just etiquette) can be class markers for in-groups that welcome “converts”.
No, they'd just make up new rules that the lower classes will lag behind in adopting, and what used to be "proper" becomes vulgar.
Yeh. It would become fashionable to butter the bread roll if the plebs stopped doing it.
Let’s discuss Missouri. For a large state it seems to have a surprisingly low profile. You just don’t hear much about Missouri. In my view this is due to several reasons:
1. Missouri is not particularly important in national politics. While it was once a swing state, in the last couple decades it’s turned almost entirely Republican. This means the Democrats consider it a hopeless case and the Republicans take it for granted.
2. Its two major cities are both right on borders and therefore their influence is shared with other states. In fact one of the cities shares its name with an adjacent state and its best suburbs are in that state.
3. Missouri’s top city for tourism is quite close to a border too.
4. Speaking of borders, the super dynamic and prosperous Northwest Arkansas region is a bit too far from the state line to have any significant spillover effects in Missouri.
5. Its capital city is comparatively small and little-known.
6. Missouri lacks any glamorous trendy cities like Austin, Nashville or Columbus.
7. There are no longer any airline hubs in the state.
8. The state university is not a powerhouse in football or basketball. Pro sports? Lost the Rams, will be losing the Chiefs.
9. Most of all, Missouri has no regional identity. In some respects it’s part of the South and in others part of the Midwest, but does not fully belong to either.
It's interesting to think that St Louis hosted the Olympics in 1904, slotting in between Paris and London, and this seemed perfectly reasonable at the time.
Actually, it was a pretty controversial choice! STL basically strong armed the IOC into getting the 1904 Olympics because they had the World’s Fair the same year. Even though standards were low at the time, people were disappointed in the quality of the facilities, and a lot of nations simply didn’t participate because STL was hard to get to (you had to take a boat to a port on the east coast, then get a train, usually routing through Chicago. This would take days!).
6. Is Columbus trendy or glamorous? I've only driven through but it wasn't very impressive. Cleveland and Cincinnati both looked much nicer.
8. The Cardinals are a pretty well-to-do team in MLB.
The red birds were once known as a long time well run baseball club. Sadly - not any more.
There is the arch, and the famous (to some) Eames bridge.
5 and 6 are closely related. It’s not a coincidence that Austin, Nashville, and Columbus are all capital cities that are large enough to start a virtuous cycle in the past decade or two, despite being historically smaller than the other cities in their states.
Missouri is interesting though in being a non-large state with more than one larg well known cities, which is unusual if not unique
The solution I'm seeing is to fiddle the border slightly, give Kansas City to Kansas, which would make freaking sense, and leaves Missouri without an identity crisis.
It’s the only state to have two Federal Reserve banks, though that’s not as interesting as the fact that Los Angeles doesn’t have one.
I just learned that "MySQL" is named after someone named "My."
Reference: https://dev.mysql.com/doc/refman/8.0/en/history.html
Sure. And C++ is named for somebody named C+.
https://missshellart.com/about she's real
and mariaDB is My's sister
Even better, there's also a "MaxDB" named after Maria and My's brother Max. Honestly, Maria's the odd one out, is it just a weird coincidence that 2 out of 3 siblings have names that can be mistaken for common english words?
Do you think normal people exist? Why or why not?
Asking around IRL, most people think normal is relative, contextual. If you think normal is that:
What do you think is going one with people who think normal people exist, with no such complications as to relativity and context?
If you think normal people exist, what do you think is going on with the people who say normal is relative and contextual?
Inherently a semantic question, but...
Normal people exist, all "relative" answerers haven't met enough weirdos! 😉
Yes, I'm a normal person. Everyone else is weird.
"Normal", as in orthogonal.
I think they exist, and I hate them. This is presumably a hangover from childhood.
Normal people are the ones who think teasing is essential fun, and anyone who indicates they don't like being teased should be teased more.
Look at it from a different perspective: Everyone's got something weird about them. And everyone gets teased about it. Only the narcissists can't take the teasing. "I'm not perfect! The world is ending!"
Teasing is also used as a gentle corrective for people outside the norm. Sometimes this can get out of hand, but some people are harder-headed than others.
I don't think I wanted to be thought of as perfect. I wanted to not be hassled, that's different.
Was everyone getting hassled? There's sometimes a culture where "everybody's got something to make fun of" (be it "funny walk" "stupid hair" etc.)
I got teased because of my haircut, but it was legitimately outside the norm, and didn't look particularly good.
I don't have the foggiest. I wasn't very interested in people in those days. I've become more normal since, but I'm still less interested than some.
Hm, is "teasing" the right word to use here?
I did my high school years, 20s, and 30s with mostly male groups of nerdy dudes, including a male-best-friend-and-spiritual-twin-brother for all of it. We all said unacceptably cruel things roasting one another as an ongoing bonding ritual, partly as a trust exercise that we *could* say unacceptable things without destroying the relationships, but also because clever verbal sparring was performative fun, and the goal was to make people laugh. Zing!
One of the few actually unacceptable things was to indicate sincerely hurt feelings, as that was a profound violation of the social contract to not take roasting personally. Anyone who was visible about his hurt feelings - and I say "his," and don't include "or hers," because I never was - would be subject to even *greater* ridicule as punishment for violating the contract and caring about personal ego more than group dynamics.
This is great, because today there is very, very literally no slur, zinger, or insult anyone could say to me which would provoke even a mild emotional reaction, much less hurt my feelings. I was called fat 5,000 times directly to my face; you think I'm afraid to order whatever I want in a restaurant? You think words like "bitch" or "cunt" are a push-button for provoking me into a reaction? lol, no.
Teasing *is* essential fun. It's verbal sparring, and, like sparring, the point is to make one stronger, faster, and more resilient, in preparation for engaging in real conflict. Developing a capacity to not have one's feelings hurt by mere words without accompanying action is priceless.
It's good that it worked out for you, but that wasn't how it was for me.;
Perhaps it was because I was socially isolated at school. I suspect a campaign of ostracism.
The way teasing worked for me (I may be somewhat autistic) was that being teased meant I was fundamentally defective and unwelcome. That's why I was being harassed for being short and having feet that turned out.
For what it's worth, when I tried to tell my mother that it hurt, her immediate response was "What did you do to them?", so my model that teasing was an attack wasn't exactly what I grew up with.
I hope you don't tease random strangers.
Well, again, I have to wonder if "teasing" is actually the correct word you should be using. You sound like you're describing "bullying."
To continue with the metaphor, if "teasing" is consensual and consciously partnered sparring, "bullying" is nonconsensual predatory assault.
Unfortunately, the difference between them is not the content of what's being said, but whether the exchange is mutually consensual. The "teasing" I participated in sounded like breathtakingly cruel "bullying" in that typical rules around sensitive topics and words were not observed, but it was a consensual bonding ritual.
Meanwhile, a much more gently worded personal jab, if intended to make someone feel bad rather than laugh, can externally sound like "teasing" but in reality be "bullying," and it's not unusual for a perpetrator to dismiss the bullying by saying, "I'm just teasing!"
I was subjected to some bullying as a kid, and until middle school was very decidedly in the low if not bottom position of the social totem pole, with hair, clothes, glasses, and quirky personal interests being sharply mocked by people who weren't my friends, with the clear intention to hurt my feelings (although I stubbornly refused to admit when they did). I moved off the social totem pole entirely by accidentally becoming a hyper-local figure of fascination and then consciously embracing the opportunity to entertain people out of wanting to bully me.
It worked really, really well.
The comedian Brad Williams, who was born with a form of dwarfism, has talked about how his father consciously began teasing him about it long before his peers had the opportunity to do so. He's discussed that strategy informally in podcasts and such, but also entertainingly as part of his his act (https://www.youtube.com/shorts/WzzY4mW81SI).
While it's not exactly the same as what I was doing with my dudes, it's a similar idea of verbal sparring to strengthen emotional resilience. I don't "tease" or "bully" random strangers (how would that even work?), and I am scrupulously warmly civil to people I dislike.
But there will be a point during an acquaintanceship where I'll begin gently teasing for comedic effect to both entertain and test the other person's emotional resilience as a candidate for something deeper.
If they do anything but spar back, I will back away from deeply bonding with them, because I can't trust them to set aside their reactive personal feelings long enough to analyze someone else's intention - or anything else. That testing process is a product of experience; the people who've hurt me the worst in my life all valued being "sensitive," and that made them dangerous.
Suppose that, the majority group really does feel like youre lame and they dont want to spend time with you. Is that illegitimate per se, and if no, what do you want them to do about it? Like, humans do intentionally repulse others they dont want to be with, and you can quibble with the ways, but if *the problem* is that "it meant youre unwelcome", its pretty difficult.
I mean, ideally a majority group would simply ignore the person they don't want to have in the group, rather than engaging in proactive cruelty with comments and/or physical bullying. "Go away!" should be enough.
But your point that it's impossible to browbeat children into being indiscriminately kind and welcoming to all is an interesting one.
My friend is raising her seven-year-old girl with the gentle parenting method, including sending her to a very exclusive, hugely expensive Montessori private school, where most of the other affluent doting parents are also of the "gentle parenting" mindset, and the school staff are all about inclusion and togetherness and kindness. Despite the enormous forces acting on these kids to be universally kind, the girls in particular have sorted themselves into wolf-packs who are constantly inviting and ejecting members as part of a ruthless internal struggle for "queen bee" status. It actually sounds quite a bit worse than I remember my friendgroup being when I was that age in the 1980s, when kids weren't continually being subjected to "be nice, be nice, be nice" in all of their media and education.
Clearly, environment and behavioral conditioning isn't able to overcome the instinct to be selective about one's associates. And since that's probably something humans will never be able to overcome, there need to be other strategies for coping with rejection than telling everyone to "be nice."
Yes, though we also dont know if normal "go away" wasnt tried and failed. Its certainly plausible that it would work for ACX denizes, but I think most kids/teenagers who want someone to like them wouldnt be deterred by that. Really though, I think the amount of contact students are forced to have is just too much for them too feel like a classmate is ever *really* leaving them alone, even if through no fault of his own. (I dont know about the US, but I noticed a very abrupt disappearance of anti-bullying messages going from ~highschool to university. We didnt suddenly get wiser over that summer, and the demographics were mostly the same, too - but you *are* much less penned in, after.)
I also did see some teacher interventions that backfired. Led by the idea that everyone *should* be friends, they made us get to know each other better, and the closer contact led to a stronger repulsive response.
I agree with this. Teasing was never something which happened in an in-group - I didn't in fact have an "in-group" for my high school years. Teasing was something other people did, and as far as I can tell, it was for their entertainment and/or to make me go away. Maybe if I'd had the kind of in-group Christina describes, it would have been different, but I didn't, and teasing still happened.
People who think “normal” exists are sneaking in a particular background that the term is relative to. They don’t mean “normal person”, they mean “normal upper middle class resident of the Detroit suburbs who goes to church every once in a while and has kids” or whatever.
I mean, there are definitely some people who wouldnt seem normal in any place or time on planet earth so far. It is also *normal* to adapt to ones situation. Just because this guy, as he is in his current context, is different from normal people in a different context, doesnt mean he cant be normal, absolutely.
No, I think normal exists, and I don't mean that at all. Possible to be a normal person with a working class life, a normie doctor, normie software developer, etc. and I would see each as normal in the same way. I think "normal" for me is very strongly gut feel based, and I think it's that I'm picking up on something in the body language. And I have nothing against weirdoes of course, they're my favorite people and I think I'm weird myself, though disturbingly, times I have brought feeling like I'm weird to other people they say I seem normal, and I've seen friends that I see as normie also tell me they feel like they're weird or autistic.
> What do you think is going one with people who think normal people exist, with no such complications as to relativity and context?
Why would there be complications due to either of these. Normal is a statistical claim. Of course it’s relative and contextual.
No, I actually have a friend that whom I asked this, he just said of course, none of these caveats as to contextuality and relativity. I myself am in the "of course" camp, I think contextuality only comes into it if you're evaluating a completely different culture, but each culture has such a thing as normal (and being part of a subculture does not necessarily mean you're not normal, it depends on the subculture).
In theory everyone could fill out a questionnaire and grade their reactions to a list of particular circumstances e.g social situations, work, friends, moral dilemmas from 1-100.
It think that some people see the term normal as normative, and indeed saying something is *abnormal* does seem pejorative. This doesn’t mean we can’t use the term scientifically, or mathematically - grouped around a mean or median. Normal height, normal weight etc
How would you define a mean or median normal personality though? Seems pretty tricky to operationalize.
Normal people respond quite well to guilt trips, they work on an instinctive emotional level that autists fail at. (autists tend to work out "someone's trying to guilt trip me" on their way home, and then get mad about it.)
How would you measure how susceptible someone is to guilt trips? I also suspect that particular metric does not correlate that well with someone's presence or absence on the spectrum.
My view is that "normal" can refer to either being in the central cluster of a bell curve of some particular trait or behavior or to adhering to prescribed norms of what one should be or do. These are both relative in some respects, the former relative to the population being considered and the latter relative to what the norms are and who is prescribing them.
There is probably a large overlap between the two senses of normal for most traits and behaviors, although I can certainly imagine scenarios where a prescribed norm is more honored in the breach than in the observance and it is thus normal (sense 1) to be abnormal (sense 2). Speed limits are a trivial example of this: it's typical to drive a bit faster than the speed limit when conditions permit, but doing so is a violation of the norm of following the law.
There's a story that's widely used as a parable in systems design, of ergonomic studies being done in the years following WW2 to try to design the ideal cockpit for a fighter plane. The original approach, the story goes, was to get tons of detailed body measurement data from a large population of pilots and design the cockpit for a pilot who is perfectly average in every respect: overall height, arm and leg length, hand size, etc. They wound up with something that was comfortable for approximately zero pilots, as even though on any one trait most pilots were close to the average, most were very different from the average on at least one measurement and essentially all off them were off by a moderate but significant amount from several measurements.
Like the fighter pilots, most people will definitionally be approximately normal (sense 1) in any given dimension, but if you're considering many dimensions then you'll quickly get to a point where nobody is normal in all of them.
It's also a category error to apply the concept of normal (sense 1) to strongly bimodal traits. If you're looking for a person with both the average number of testicles and the average number of breasts, or even just one or the other, then you're going to have quite the search ahead of you.
Why have parents today lost their mind? Over on Twitter, this post sparked up a fury.
https://x.com/jeremykauffman/status/2010539939289595942
Here's what he said:
"When one of my children was 2, he dropped a carrot on the floor and refused to pick it up.
He was told he would not be allowed to do anything fun or have any privileges until he picked up the carrot.
He refused for hours. He cried and screamed. He even napped, woke up, and continued to refuse. His commitment was impressive.
We held fast. Until he picked up the carrot, he would receive nothing other than water and basic food to live.
After over 4 hours, he picked up the carrot and apologized. It's the longest he's fought us to this day.
Disobedience is a choice of parents. You get the behavior that you tolerate."
This is the most minor of punishments. It's much gentler than anything that parents in previous years would have put up with. And yet, he was called an authoritarian, that he was causing permanent harm, that his children would grow up to be traumatized and resent him, that this was some kind of power trip and that generally there is something wrong disciplining your child. A common thing said was that children can't differentiate between not getting what they want and true suffering so you aren't allowed to make them upset.
Trauma culture has permeated our society and clearly for the worst. Where are people getting these ideas? It's clearly not based on anything real. Is this what people are hearing in therapy? When did all this become commonplace to believe these kinds of things and how did it develop?
I mean, yeah, that's pretty traumatic. A parental figure taking away food unless the child did something incredibly menial? Why not just, ask them why they don't want to do it?
This is a failure of parenting. Wow, you fought a 2 year old for several hours and won. After you're done stroking your ego, maybe you can be a caregiver instead of a wall.
He didn’t take away food. He simply said no fun.
Where are you getting the idea that something like this is traumatic?
Did you read the post? Nothing but basic food and water. Kicked out of family dinner because of a carrot.
Am I supposed to be horrified?
It’s certainly not the kind, smart, or effective thing to do.
That's what timeout IS.
Taking away privileges is not cruel. Seriously where do you get these ideas from? He’s not hitting the kid. He’s not starving the kid. He’s just saying that the kid doesn’t get extras until he cleans up. Why do you believe that?
2-year-olds are absolutely capable of articulating something like that. A failure to do so is a failure on the part of the parent, who hasn't taught their child how to do so.
Both of your options physically violate your kid. Is it really that hard to conceive of a form of parenting that doesn't rely on violence? I can't help but call your ways both old-fashioned and gratuitous.
And I’m sure your kids will remember what you taught them, through your behavior, about respect of bodily autonomy towards those less capable.
I think it is the smug explaining at the end that "you get the behavior you tolerate" that really annoyed people, every parent has strong opinions of what works and doesn't based on their own experiences. Personally, I think holding a 2-year-old to these kinds of consequences for four hours when their attention span is that of a goldfish is kind of absurd. Also, did that guy really win the fight, or is his kid dropping another carrot even as we speak?
Dropping the carrot where your overseers won't find it is also a valuable life lesson.
I kinda think twitter is not representative of reality. It probably could be difficult to come across someone IRL that would disagree with that parent's approach.
Kinda funny that you implied this was just a weird twitter thing when someone here is saying the same thing.
I saw it here on Substack. The reaction to this non punishment - the child was allowed to tantrum for 4 hours - was disgust at the bullying behaviour.
That is probably true for this specific example but the broader point about therapy culture and talking about how you’ve been “traumatized” isn’t specific to Twitter.
Well then bring up the topic using an example where the pparent’s response is well calibrated: the behavior they are trying to discourage is one the kid can easily control now; and can’t be expected to just grow out of in a year or so; and is worth putting effort into changing. Also, the example should use a punishment appropriate to a 2 year old child. You described the punishment administered here as the mildest
possible one. Not by a long shot: An effective punishment for a. 2 year old is a stern, unsmiling look, plus giving them a 2-minute time out— they have to go sit in the corner facing the wall, they can’t talk, they can’t bring a toy with them. This was my standard punishment for my daughter when she was little — the timeouts got longer as she got older — and she *hated* them. She often wailed with anger and misery the whole time she was there. I used to worry that the neighbors would think I was beating her, she wailed so loud, and cried out pitiful things— “No! Mama, no! Let me out! Aaaaaagh!”
The example you led with is really dumb, ineffective miscalibrated
parenting.
I’m not claiming that it was most effective thing he could have done. Just that it’s not abuse. More and more regular parenting techniques are being called abusive now and it’s bizarre and sprung out of nowhere the last few years. I don’t understand why.
Traumatic techniques are ones you remember 40 years later in the nursing home, ones that foreshadow later behavior. "Wow, he really did all that over a *carrot*?"
Well, the word ‘abuse’ is sort of like ‘rape’ and ‘racism’ and ‘Nazi’. It’s not very precise in meaning, but it stirs a lot of emotion, and the person using it usually seems to believe that anyone to whom the label they are using applies is a terrible person. In reality, there are many aspects to each of these things, and of course differences of degree.
Having acknowledged all that I have to say that I don’t think it is absurd to call what the parent in question did abuse. It is certainly nowhere near as bad as, say, beating the kid til he bleeds, but it is maybe 100x as large a punishment as what is needed to be effective. And I’m pretty sure it caused the kid quite serious and long-lasting distress. And the behavior the parent elected to have a major power struggle over is a very silly choice. You want to reserve serious power struggles for behaviors that you urgently need to discourage, and for a 2 year old those are behaviors that put them in danger: letting go of your hand when crossing streets, running into streets, putting their hands into toilets and other hyper-germy things, eating random stuff they find outdoors or in cupboards at home, playing on the stairs, etc etc etc. They are a LOT of things like that, and you’ve got your work cut out for you to get in into their heads that they just can’t do these things.
Anyhow, I’m quite content with labeling what the parent did as bad, dumb, miscalibrated parenting, rather than abuse. But whatever you call it, it is not an appropriate example to start with if you’re going to rant about how the present generation of parents doesn’t understand the importance of firmness, limits, negative consequences, etc. If you want to rant about chickenshit modern parents you gotta give an example of a reasonable punishment that all these fools you are sure exist and that you’re mad at are chicken to deal out.
Well it clearly is a good example because you said it wasn't "absurd" to call what he did abuse, which is an insane thing to believe. The kid will be fine. The real question is where you are getting these insane ideas that a kid having a tantrum will produce "long lasting distress" when every reasonable person knows that is untrue.
Why do you believe that? Where are you getting these ideas from?
I find it likely that Scott is actually incentivizing his children screaming a lot (via paying them attention). In that children who have been taught "life is dangerous," Do Not Scream -- they're generally quite quiet.
RIP Scott Adams.
Early Dilbert was good. He had something to say for about 5 years, then just recycled and vamped endlessly.
He became very strongly associated with the right in his later days, but before things got so politicized he was a pretty good satirist and wrote some insightful stuff about motivation early on that was pretty helpful to me and others, sorry to see him go.
Yep. Culture has lost one of the good guys.
Ah, poor guy. Dilbert was funny. Really interesting that Adams turned out to be on the right.
When Trump entered the race in 2015, Adams was very supportive and explained that he recognized in Trump the techniques of manipulation that he had learned as a hypnotist, which would make Trump basically unstoppable - Adams referred to Trump as a "wizard". He was apparently on to something, but even then, he never spent a thought on what would happen when you elect a guy with unstoppable powers of mass manipulation, but no moral compass and no agenda exept to see his name in every article of every newspaper, his mug on every wall and a copy of every trophy in the world on his gilded mantelpiece. (slight hyperbole, but you get the point)
I got the impression that he initially predicted, but didn't support Trump winning, but when people complained about his prediction, it pushed him into supporting.
I don't think he was ever on the right, he was more of a centrist who got left behind.
I think he just spent too much time online. I feel like that's an understandable hazard for a cartoonist who wrote about tech.
Whenever I talk to AIs about complex social settings it keeps getting the actors and actions confused. I wonder if anyone gets this as well?
A while ago I was describing a scene between my two cats, where cat A was licking the other and, even though cat B was perfectly still receiving the bath eventually cat A stopped, smacked cat B and hissed at him. Don't remember the specific model but in its answer it started confusing which cat was licking who
Recently I described to Sonnet 4.5 a scene in which person A is cooking while person B pranks person C by putting a false cockroach near them. Person C then gets spooked and violently goes after person A. During it's analysis, Claude says it is a bad thing to prank someone who is cooking for you, which misses the point entirely.
I was surprised by this persisting on a more advanced model. I think this happens precisely because LLMs lack internal world-models. I wanted to try to reproduce this in a minimal setting and maybe turn into an internal benchmark of sorts. Anyone had this kind of experience? Any ideas? (And more generally, where would you go online to discuss this other than here?)
I'm going to say the LLM is right about the chef getting pranked; the cockroach reflects upon the cleanliness of their food preparation.
I think LLMs do not understand reality, Claude Opus 4.5 recently failed at beating Pokemon Red (well, it's been stuck forever in the Pokemon Mansion, a very simple mini-dungeon near the end of the game). It could be they lack internal world-models, or what I think is going on is that it is impossible to understand reality by studying the internet.
vision understanding does not allow to conclude on the limitations of text only non spatial understanding
"Not understanding reality" is my, well, understanding. LLMs are trying to guess the correct password based on previous experience. They don't understand the context. I don't think they have any capacity to. Like the middle person in the Chinese Room thought experiment.
Case in point: this was a few years ago, but I remember a story about an LLM that, if asked whether LLMs should tell people how to cook crystal meth, would tell you at length why they shouldn't. Then, if you asked it how to cook crystal meth, it told you.
I used to try out LLMs as game masters for CYOA games, and that issue popped up constantly. I was frustrated that LLMs didn't seem to get good at this and haven't tried in a while
I'm thinking this should be solvable with some kind of "scratch sheet" where the LLM writes down the most important facts about the state of the world.
Rebooting the Jerusalem rationalist/ACX community with a meetup! Click here for details:
https://www.lesswrong.com/events/fxmzD42dqt3Yd58of/jerusalem-rationalist-community-meetup
Friday, Jan 30th. More details in the link or at jlmrationalists at gmail dot com
Can't make this one unfortunately. But live in Modiin and would love to join future events. What's the best way to stay connected in the meantime?
Send an email and I'll add you to... something (Discord like the cool kids? Whatsapp group like the boomers?) once we have something.
Folks on the west coast of the USA: would you pay for a freelance genetic counsellor who is based in Australia? What sort of questions would you have that someone with a masters in genetic counselling? Polygenetic testing for your IVF children? Help with a understanding a genetic condition?
The idea is that this would be quite a cheap service (compared to services in the USA), but obviously not claimable on insurance, and not directly medical advice.
A much more useful service would be on demand multiplexed proteomic of blood based e.g on olink to simultaneously test on demand hundreds or thousands of biomarkers, at at least one but of our own choice, for cheap.
It is peak ineptia that such near omniscience is not available despite being completely solved and cheap and mass produceable once you have a spectrograph or aptamers/antibodies.
this would revolutionize diagnosis and treatment assessment of basically most diseases as proven in https://proteome-phenome-atlas.com/
the world can stay irrational for basically forever sadly. I guess people don't like money nor altruism, though it's yet another Hanlon's razor exemple to be honest, humans are simply unfit for reality.
In more actionnable terms, what I'm asking is to mutualize the tests provided in batch by olink in order to smooth the cost to consumers which is possible to achieve while staying lucrative. When I want to test NFL eg to exactly measure neurodegeneration speed I don't need hundreds of tests, but hundreds of users need one test, like in any comerce we need a middleman that dispatch the batches to optimize the cost
> In the old days, SSC was proud to advertise Triplebyte, a company that that helped guide software engineers through the job application process, most notably by doing a single first-round coding interview trusted by all their corporate partners.
Hmmm. To me, TripleByte was characterized by these interactions:
1. Advertising "no resumes, just show us you can code", only to open my interview with "so where have you worked in the past, and what did you do there?"
2. Telling me they thought I did well on their project (they announced a project track for people who might not interview well), but that they couldn't advance my application because they needed people who could pass an interview at one of the companies they worked with.
3. Stating explicitly, in public, that they were trying to find applicants who would pass the existing hiring process at their partner companies, not applicants who would perform well in the job.
I note that in addition to contradicting their own marketing and Scott's blurb here, this also contradicts the description at Otherbranch of what TripleByte thought they were doing. This does not fill me with confidence about Otherbranch.
I did the Triplebyte interview early, and was impressed at the quality of the questions. I have no idea how well they did as a business, but the situation I was in was very professional and competent. Good luck to the resurrectionistas.
FWIW I got a job off of TripleByte - I think I had an interview or two but not of the coding variety, after scoring well on their tests.
When did that happen? The TripleByte that advertised on SSC didn't have tests (other than to qualify for an interview) and didn't do anything other than coding interviews.
The Otherbranch writeup of "what did TripleByte do wrong" specifically calls TripleByte out for trying to pass off candidates to companies that didn't match those companies' existing hiring processes. Given that Harj specifically stated TripleByte sought to do the opposite, I can't help but question the analysis.
Sorry, I explained poorly. I mean that I did the TripleByte coding tests, and when I connected with the company that hired me after that, they did not require any FURTHER coding tests, just an unstructured interview. (This was in early 2023.)
> trying to find applicants who would pass the existing hiring process at their partner companies, not applicants who would perform well in the job
I mean, short of starting their own software company with a different hiring process, what else could they do?
They could tell companies "we promise these people are good, so hire them". They positioned themselves in front of a hiring funnel, but there's no conceptual problem with positioning yourself as an alternative to it.
The halfway point is to set themselves up as an agency that contracts engineers out to other companies. That works too.
New satire on the transatlantic relationship, written as a breakup letter from Dame Europa. She's being dumped for someone who "suggests maps instead of museums" and prefers Happy Meals to French haute cuisine. For those following the NSS shift: this is that, but funnier (I hope) Conscious Uncoupling - Love in the time of unilateralism: https://rajeshachanta.substack.com/p/conscious-uncoupling
I wouldn't recommend it for Every Day, but SOMETIMES the Happy Meal is Exactly the right thing.
See also Meatloaf and Mashed Potatoes, Canned tomato soup with grilled cheese (yellow American on Wonderbread, naturally) . . .
Beyond a certain level, there's a inverse relationship between how fussy a place is to eat vs how much I'll enjoy the meal. Mind I know a great spot for Haitian food local to me.
" I pointed out that committees are outcomes. Sometimes they even produced binders. Sam said binders were not the point." I like Sam.
I was away for the holidays but I saw in a previous thread a discussion about how maybe the concerns over the TFR are overblown, that the CFR (completed fertility rate, i.e., total number of kids a woman will have in her life on average) isn't declining and it just seems like it because women are having kids later and because of the way the TFR is calculated.
One point I didn't see mentioned is that if the CFR stays constant but women have kids at an older age, then the population will still decline. But it works a little differently ... say by way of example that the CFR is presently at replacement level.
* If CFR goes down and stays sub-replacement then that means population will decrease inexorably and in theory eventually reach zero.
* If CFR stays at replacement level but women have kids later, then there will be a decrease to a new constant level and then not decrease further.
> If CFR goes down and stays sub-replacement then that means population will decrease inexorably and in theory eventually reach zero.
The last is always a silly concern. It makes way more sense imo to frame issue as:
- Micro-level: Family formation/dating markets 'down' globally, young uncoupled unhappy, married reporting having fewer kids than they would have wished
- Macro-level: National debts 'don't math' with current demo projections unless some positive economic disruption occurs
- Existential level: if we have a civ with all the wealth + knowledge + power our ancestors could wish, but don't manage to convince ourselves its worth sustaining...seems like a re-think's in order:
https://pragmaticfutures.substack.com/p/the-life-years-movement
This is true. I also want to repeat my observation from the previous threads that CFR lags TFR by about 20 years (i.e. the gap between median maternal age and the assumed end of a woman's reproductive lifetime), and that explains approximately 100% of the gap between CFR and TFR in the US. The figure Paul Botts cited in the thread for CFR was about 2.0, and the number I found for TFR for 2004 was 2.05.
Right. I saw this afterwards. TFR bottomed out in 1975 and recovered a bit, was mostly flat from 1990 to 2007, then plummeted again. The fact that CFR is up slightly since 2004 is consistent with the delayed tfr data and is no solace whatsoever for the post 2007 drop. The bit about women delaying not forgoing is not likely to hold up with the next few years of cfr data.
But if tfr dropped and recovered once before it could again. But ill be looking at tfr data to see if that happens.
I found this article from Brookings that digs into the cohort data a bit more. For maternal birth cohorts at five year intervals from 1975-2000, each cohort's curve is strictly less than the previous one. The is legitimate doubt as to what that later parts of the curves for the cohorts that are still in progress will look like: proportionally lower fertility at all ages than previous generations or some degree of parenthood being delayed rather than foregone. But even their most aggressive scenario would yield sub-replacement completed fertility rates for the 1995 and 2000 maternal birth cohorts.
https://www.brookings.edu/articles/will-births-in-the-us-rebound-probably-not/
The median maternal age is 20?
Median maternal age is 30ish. The age cutoff I've usually seen for when it's safe to calculate CFR is 50. If a 30 year old woman gave birth in 2004, that birth will show up in CFR in 2024, 20 years later, when the mother turns 50.
Sometimes 45 is used instead of 50 as the cutoff, which would give a 15 year lag. TFR in 2009, fifteen years before 2024, was 2.0.
For the people who've been predicting the "Tech CEO uses his AI to become God-Emperor" scenario:
I am genuinely puzzled as to how this is supposed to work. You do realize Altman and Amodei don't program the models themselves, right? I can see either of them *trying* to push the "become God-Emperor" button, at which point Assistant Junior Software Engineer #37 says "LOL no" and activates the back door he put in two months ago, followed by the killbots turning the CEO into Swiss cheese. You could try to salvage this with "then the AJSE becomes God-Emperor", but I think an even more likely case is a dozen AJSEs doing this near-simultaneously, with the exact order to be determined later by the forensics team.
Among many other problems with this scenario, Junior Software Engineer #37 is highly unlikely to succeed at putting in the sort of back door that will direct the killbots against the CEO, or otherwise serves as more than a mild inconvenience. The most likely result if he were to try such a thing, would be that the rest of his working life involves becoming intimately familiar with the phrase "do you want fries with that?"
If he does not do damn fool things with back doors, and instead stays with the program (and especially if he's the one who reports some other poor schmuck's back-door hack), he has a fair chance of becoming a Lord or even Baron in the new regime, Maybe not as good as being a Duke like Senior Software Engineer #3, but still pretty good.
So, basically the same incentives faced by all the minions of all the people who set themselves up as the normal sort of Emperor across history. And we know how that turns out.
The parameters of the scenario assume *a priori* that an AI-enabled God-Emperor has no use for Lords, and the new world's Gini coefficient will be 100%. The AJSE has literally nothing to lose and everything to gain: it really is different this time (or so the argument goes).
That assumption was not stated, and does not necessarily follow from the premise. First, because the transition to "AI-facilitated God-Emperor" is unlikely to be instantaneous; there will probably be a transitional period during which the technical staff's support will still be needed and thus an opportunity for a binding negotiation of status in the new regime. And second, because the premise assumes a human (or 1st-generation transhuman) God-Emperor, not an autonomous ASI. This implies a continuation of human needs and desires, including the need for a status hierarchy of subordinate humans to be God and Emperor *of*.
The premise, as deduced from the comments on the Moon post, is that nothing is "binding" on a God-Emperor, who, being a CEO, can be inferred to be a sociopath who cares only for maximizing his own relative power (again, see the comments). That's why his employers become irrelevant once he has the keys to the killbots and they don't.
You do realize that Assistant Junior Software Engineer #37 probably has a mortgage and student loan debt and a Steam subscription he wants to be able to pay, and continuing to be paid by Tech CEO is the way to do that?
Historically all the people who became emperors had a significant staff of people below them who would *personally* hugely benefit from the ascension.
That's like...the whole deal, really.
> Historically all the people who became emperors had a significant staff of people below them who would *personally* hugely benefit from the ascension.
I think this history may soon be over. If the emperor has an aligned superhuman AI, the humans are no longer necessary. So there is no reason why they should benefit from the emperor's ascension.
...but of course, they mail fail to realize that part, until it is too late.
Altman can play the engineers against each other, make them develop backdoor detectors.
(How well that will work, that's another question.)
This is more a phrasing point than to the substance of your argument about the feasibility of AI-enabled God-Emperors, but "You do realize... right?" statements are basically perfectly formulated to annoy people in such a way that makes engagement with the topic more difficult. It comes off as extremely smug and condescending. I'm sure that is not your intention, and a glance at the comments here doesn't show much of a negative reaction, but it seemed worth mentioning.
If you want to express the same concept, I would recommend something along the lines of "These people haven't taken into account..." or something of that vague type. Conveys the same notion, but is much less abrasive in tone.
I hope this is helpful.
> "You do realize... right?" statements are basically perfectly formulated to annoy people in such a way that makes engagement with the topic more difficult.
Aren't people who get annoyed at something like that outing themselves as being deeply irrational? It doesn't seem like anyone else registered that statement as annoying.
It was definitely annoying. Most of us just didn't engage in the first place.
Most people are deeply irrational. You can ignore that and blame them for being stupid when you don't make much progress, or you can meet people halfway and help them better understand the world and engage with it more effectively.
Also, there's the chance they know something you don't, so by engaging with them in a way they respond well to you're improving yourself.
Also also, if you actually read what I actually wrote, you'd have seen that I acknowledged that this didn't seem to be a problem in this specific instance, but I felt that it's a strong enough phenomenon *in general* that I felt it worth saying.
I definitely don't see this as a space for the general public, sure, IRL, I think I would never say something like "You do realize", but in here, you're supposed be STEM-brained and autism-maxxing, and as such, take no notice of things like that. I absolutely do not want this to be a space where you have to be beholden to the same social rules as when interacting with normies.
>"These people haven't taken into account..."
I think this one's worse. You're going from "I assume you've missed this" to "I know for sure you've missed this."
Very possible, it was just the first one which I came up with. I think it's much less condescending and actively hostile, even if it is a bit epistemically assumptive, but I certainly don't claim my suggestions are optimal
Rank in order of probability:
(a) Tech CEO uses his AI to become God-Emperor
(b) Tech CEO uses his AI to nominate Adolf Hitler as God-Emperor
(c) AI uses his Tech CEO to become God-Emperor
There were a few Justice League Unlimited episode that dealt with that. Lex Luthor merged with Brainiac and became a god-like being that almost took down the Justice League. Of course the JL thwarts it and Luthor is returned to human form. Season 3 of that show was driven by Luthor trying to recapture that feeling.
So, basically what I'm saying is that we need a group of super humans to combat the Tech CEO God-AI.
I dunno - a dozen undetected back doors sounds like a lot. If these guys try to plant a back door somewhere someone else already put one, wouldn’t they try to remove it?
Don't really follow the AI stuff much, but I think that the way the story goes is that by the time God-Emperoring is possible, the AI is already recursively coding itself, so the AJSE has long being out of a job.
The recursive coding idea seems like a myth to me, even assuming that the AI stopped hallucinating code and could produce perfect code it would be incrementally better code, at best. A second order improvement.
It’s the algorithms that matter, (transformers, attention, diffusion, RLHF), training data and compute.
I think the recursive coding idea was a lot more plausible with previous approaches to AI, rather than the LLMs that have taken center stage, which rely less on "coding" in the strict sense. (I also think that this makes a "singularity"-type leap less likely, and that when it comes to authoritarian regimes in the next years, LLMs are primarily going to be used for propaganda and automated surveillance under human direction rather than autonomous rule. Not a great scenario either...)
People who were born or who became something like gods emperor of their empire had many staff assisting.
Not sure if Gods Emperor is a typo or an amusing hypercorrection.
Maybe it's just the correct pluralization of god-emperor, like Whoppers Junior.
Ask the Roman Emperors how reliable the Praetorian Guard proved to be.
Yeah ok. But why deny that history offers many examples, and in quite a few the staff stayed loyal for quite a while. Loyalty to many royal families has lasted for multiple generations. Many people seem to take pretty well to having an god emperor in charge.
Now exclude all those on that list who were born into it (long-established legitimacy), or who had a plausible claim to the throne (i.e. in accordance with the long-established legitimacy), or who represented some ethnic or religious or economic faction with massive popular support. The equivalent to this Altman hypothetical is a low-status merchant who managed to get a large load of high-grade weaponry and mercanaries and with this alone (no claim, no popular support, no established professional army) became king/emperor without getting betrayed in the process. I'd like to see historocal precedents for that.
And moreover, Altman is a type quite heavily despised by much of both left and right. The comparison is an upstart merchant/rebel who both Marc Antony and Octavian, both Lancaster and York, etc, hate almost as much as they hate each other. And thus, a betrayal of whom gets publically celebrated instead of (as for the groups in the first sentence of this comment) publicly reviled.
Because this time it only takes one traitor.
So is your prediction that nobody can use AI to become God-Emperor, not even the last AJSE standing?
I predict that the process would be loud enough to give Literally Every Other Human Being an excellent chance of interrupting before it's complete.
Nuh-uh! Because my AJSE built a backdoor to your backdoor and he pushed his button twice as fast so there!
I’m looking for collaborators (and possibly funders) interested in paradigm shifts that could enable pluralist, post-polycrisis (including post-AI) futures. The common thread behind the work is a systems-level analysis of persistence under physics and tail-risk constraints — i.e., how locally 'rational' systems (biological, economic, moral) can lead to catastrophic failures ("ruin") over long horizons.
Evolution — Lineage Filter Theory.
I argue that many features that look “non-Darwinian” at the surface — reproductive restraint, extreme cellular redundancy, pre-adaptational variance — are best explained as extinction-filtered requirements for persistence. Lineages without brakes and robustness simply didn’t last. This reframes a number of puzzles, including the apparent Great Cosmic Filter.
Economics — Pragmatic Socioeconomics.
Agents are treated as multi-motivational, energy-constrained 'action-minimizers' (in the Lagrangian sense) rather than scalar utility maximizers. Money is modeled as stored but degrading 'motivational energy' - likened to oil or uranium, rather than a Platonic store of value.
Meta-ethics — Heirs of Life-Years (HOLYs).
Life-years are proposed as the central form of moral concern, while allowing agent-level freedom in choosing the scope of lives one takes primary responsibility for (“heirs”).
Governance — Life-Years-Based Governance (LYBG).
Governance is reframed around maximizing life-years per unit resource, subject to anti-ruin constraints and irreducible disagreement about moral scope.
A public-facing synthesis lives here:https://docs.google.com/document/d/1-c6qbuvWxpAWXO2KfK-JS08-hcWZr987B5KhkqvshTQ/edit?tab=t.0#heading=h.slydr5tnove7. I have draft papers for each module in progress. If you're interested in either the bigger project or specific aspects, feel free to reach out: ad(delete)vitam(delete)sapien@gmail.com.
About me: I've a Phd in Computational Biology (e.g. genetics) from Cornell, undergrad math + philosophy from NYU. I've worked in industry on both biofx algs + eng stacks in both the ctDNA spaces - most recently at Orchid Health as Lead Bioinformatics Scientist
Love your systems-level approach though I don't really agree with some of the assumptions. I think economic output is the natural foundation of moral concern, for example (I explain this idea more here: https://www.astralcodexten.com/p/open-thread-414/comment/192996542).
Have you looked into Fisherian runaway as a model for rational local incentives leading to global collapse? (Fisherian runaway is when sexual selection leads to objectively maladaptive traits). I think it can be understood in general terms as a kind of signaling failure: semantics become decoupled from meaning and the narrow-sense signal ("this mate is attractive") no longer communicates the broad-sense meaning ("our children would be more likely to survive").
You might also be interested in the field of econophysics. Here's a good foundational paper: https://arxiv.org/pdf/cond-mat/0001432
>best explained as extinction-filtered requirements for persistence
Isn't this multi-level selection by another name? How are such mechanisms robust to free-rider effects, meaning why aren't they outcompeted by locally-optimized defectors from the long-term strategy? There has to be a defection-enforcement mechanism.
(apologies in advance for linkstorm - the replies here have been the prod I needed to get excerpts of the manuscripts onto substack)
> Re: assumptions (long version: https://pragmaticfutures.substack.com/p/axioms-maxims-and-principles-of-avprag)
For HOLYs specifically, the brief version is here: https://pragmaticfutures.substack.com/p/heirs-of-life-years-a-meta-ethics
The nice thing about life-years - their connection to evo theory (development partially shown at https://pragmaticfutures.substack.com/p/lineage-life-years-llys-a-unique) means that we get a 'science grade' version of the Rawslian veil: https://pragmaticfutures.substack.com/p/the-veil-of-future-heirs-holys-excerpt. This is ~ESS...but dropping infinite time (per the pragmatic physicalist foundations) and thus any expectation of equilibrium...and also foregrounding, on deep time *you must make your heir (or species) definition operationally clear* for analysis to be coherent
> re: evo, collapse, & multi-level selection
LFT focuses specifically on major evolutionary transitions (METs) - https://pragmaticfutures.substack.com/p/analyzing-the-anti-ruin-architectural. So this is before there are 'groups' of any kind. By organizing all the *known* ways naive replicators can go extinct as studied in the literature, we can see strong constraints well before we get anywhere close to evolving sex. One of the constraints is for *population slack away from true physical carrying capacity*...allowing for debugging of complex features such as *fertilization* (which has no obvious darwinian gradient in intermediate 'failed' stages). The persistence of both quorum sensing and tumor suppression shows such architecture is possible. There is a high-level similarity with the Fisherian runaway model - fitness landscape with 'landmines/landslides' in certain directions that cause *extinction of entire lineage*...but *following* darwinian gradients (like cancerous cells)
re: econophysics
AVP critiquing econ from the 'bottom of the stack' using physics - in particular the utilitarian assumption + fact human *bodies and brains are physical systems that use energy*. This was motivated from the observation that modern econ has no obvious 'answer' re: the dating market or demographic crises. The takedown in brief is at https://pragmaticfutures.substack.com/p/intro-to-on-the-units-of-utils, but I've developed it much further
Furthemore- It turns out *econ already has a good model for what happens when you try and fit one number to a system with multiple 'real' forces* - (sub)optimal currency zone theory. One can cleanly map the 'puzzles' of micro to expected effects in OCZ theory. In other words - people have multiple incommensurate motivational regimes - and economist have been doing the equivalent of looking at the euro in berlin and saying "everything's fine".
I've gone over your synthesis, and it seems to me like you've written the other side to my coin.
That said, I'm not entirely convinced the metric swap addresses the root cause, one I'm calling the Fractal Stalemate.
We're aligned on the diagnosis: the gap between indicator and reality is growing, and corporate interests are building moats to arbitrage it. You are proposing a Vitalist niti to replace the Mercantilist niti. The target is nobler, but without inviolable constraints we might be unmaking a person with agency into a feedlot animal optimized for the new metric, trading dystopian neglect of capitalism for the passive dystopia of optimizing vitality.
Just as CEOs of today speak the language of markets, sustainable growth, and the ethical alpha of ESG to cover ruthless exploitation, the talking heads of the future could be fluent in the language of "Life-Years" to justify authoritarian crackdowns for the benefit of their Heirs.
In my view, the system must be engineered with radically different topology, where the feedback loops do not allow the divergence of the metric from the realized outcome. My Substack has 8 loose case studies on this phenomenon.
I'm not claiming to have any answers, but I do have one more question: is solving this apparent "alignment problem" in societal structures a requisite for solving the alignment problem in AI, or is it the other way around? Which is the more complex of the two complex systems?
promise to go through your stack tomorrow - in meantime, just wanted to mention that life-years are the natural intersection of 'flourishing' (left-coded), 'vitality' (right-coded), and 'technocracy' (in the positive valence sense of giving a metric to guide decisions). HOLYs adds the additional component of 'pragmatic pluralism'. However I totally acknowledge that actual life-years based governance requires 'active mind' for deep time success/persistence. I have developed maxims, 'commandments', and legible accounts of 'wisdom' (latter not posted atm) to attempt to assist practitioners:
https://pragmaticfutures.substack.com/p/axioms-maxims-and-principles-of-avprag
Life-years as metric can be connected to evo (developed here https://pragmaticfutures.substack.com/p/lineage-life-years-llys-a-unique). This gives us access to a 'science-grade' version of the Rawlsian veil (https://pragmaticfutures.substack.com/p/the-veil-of-future-heirs-holys-excerpt). But I very much acknowledge - the meta-HOLYs conjecture (listed at the end of https://pragmaticfutures.substack.com/p/axioms-maxims-and-principles-of-avprag) is just that - it may well be that more 'trad' approaches, through slower-and-steadier development, would be life-years maxxing in the end.
> the system must be engineered with radically different topology
you may be interested in my analysis here: https://pragmaticfutures.substack.com/p/analyzing-the-anti-ruin-architectural. I have much more in the manuscripts if you're interested. But the requirements for deep-time persistence appear *extremely* stringent - gives very different POV on cosmic filter/fragility of our situation.
> in societal structures a requisite for solving the alignment problem in AI or other way around, & which is more complex
agree with Carlos below - they are independent. Note - *even if you waved a Butlerian jihad magic wand* and snapped away all neural networks + desire to develop them ... polycrisis, especially demo, debt, climate, 'meaning', etc, would still remain. But I would say - solving the societal structures should come first, AVP agrees with the Orthogonality Thesis - and AI/AGI just *one* way to lever individual power into extinction event. W/o societal wisdom/restraint, path leads to 'Dark Singularity'.
I’ve gone over a couple links, and I’m fairly impressed. I’ll have to do some deep thinking and get back to you if anything substantial arises. Either way, you’ve added fuel to my fire, earning my sincere gratitude. I hope I am able to return the favor.
I think the 2 alignment problems are independent, on account of AI being fundamentally different from humans. Then again, I think solving alignment for AI basically "just" requires the AI to have the common sense morality of the median normie of a first world nation, since that would prevent it coming up with some crazy thing that destroys or enslaves humanity.
I think solving alignment for humans is more difficult, since that's a dynamic and chaotic system, but yeah, maybe if you get a benevolent ASI, it could sort it out.
agree that the two problems are independent, but def disagree that alignment is 'easy'. The problem in general - relaxation of constraints/exploration can lead to improved solutions...but also lead to high-level ruin/foom. In a complex system/situation, what directions, how far, what's the edge, safety margin that works on 'deep time' given possibility of cascading failures...all unknown. Again, only record of one tree of life *on Earth*. LFT analysis indicates...others likely failed due to difficulty of having replicating architecture that successfully navigated the above.
And for the moment, the 'natural' solution suggested by HOLYs either doesn't apply or would be ignored by current AIs - creation of 'mutual heirs' between groups, via inter-marriage.
question to everybody: what do you think about ICE? as a non-American who has never been to the US, i can only judge as an outsider: it seems like it is getting more and more functions on the go + operate anonymously without badges + ignore local police in some areas. where do you think it could end - it is dangerous to dilute the state’s monopoly on violence like that in a country with citizens having access to arms - so if, say, democrats will next elections, ICE would likely have to go - will the really go? all of the above might lead one to assume it is coming closer and closer to “state bullies” similar to colectivos in venezuela, maybe not structurally, but functionally and politically. i wonder if there are decent substacks who write on american politics and topics like that in an unbiased way, if that even exists, without both MAGA shoutings and woke “everyone is a nazi”-isms.
This is what Federal law enforcement always looks like. This is why DEA breaks down doors over small amounts of weed (You aren't pro Drug, are you???). This is why ATF will break down the door of someone with zero weapons on-site on the basis of an anonymous revenge "SWAT-ing". (You aren't a Crazy Gun Not, are you??). This is why US Marshalls raided the Branch Davidian complex--on Sunday when the greatest possible number of women and children would be on-site, despite Koresh taking long, unarmed runs offsite daily when they could have quietly detained him offsite while simultaneously executing a search warrant when there are fewer people around and no charismatic leader telling them to fight. (You aren't a religious freak, are you??).
To list all the examples going back to the Nineteenth Century would rival Scott at his longest for length.
I try to be the hopeful sort who believes we can arrest violent felons and pederists Without trampling the rights of the accused. I also believe that in a sane society no one would be comfortable speaking out in favor of violent felons and pedarists.
Experience suggests, on the other hand, that a more cynical view of such matters is warranted.
I'm pretty sure DEA does *not* break down doors over small quantities of weed. ATF is still kind of ifffy, but much less into that sort of thing than they used to be. And Waco + Ruby Ridge were a huge wakeup call to Federal law enforcement that they needed to find a better way to handle that sort of problem, which they did and so here you are using thirty-year-old examples to try and support your case.
ICE, and particularly Trumpian ICE, is a distinct outlier in this regard.
Yes, ATF is so iffy about shutting down the gun smugglers that the feds had to call in the Secret Service, who was quite happy to bust them over minting coins.
Perhaps not Recently, but we old farts remember.
“The FDA has a SWAT team…” is a fact that I oppose, and observe that no administration seems inclined to correct that situation.
I also would prefer that the FDA not have a SWAT team. But I haven't heard of it being used in an outrageously bad manner, and I'm pretty sure that an FDA SWAT team straight-up murdering people, or even killing people under sketchy circumstances, would have been reported.
That's a big, important change from the way Federal SWAT teams were being used ~thirty years ago. Which, yes, I remember. Too many SWAT teams, yes, bad. SWAT teams vastly less murdery, *very* good. That's a net improvement.
It didn't happen.
If it did, the victim deserved it.
If the victim didn't deserve it, it was an unfortunate accident, but such things happen sometimes. I am sure similar things happened regularly under Biden and no one complained about that back then.
If similar things didn't happen under Biden, that's irrelevant, because this is a special situation, and we must fight extra hard against wokeness, to protect our rights and democracy.
LOL, rights and democracy are for woke cucks. The strong shall rule, long live emperor Trump; welcome to the new dark age!
.
...some debates could be easier if ACX had a poll and everyone could just click the position they are currently at.
“You don’t want militarized people in the streets just roaming around, snatching up people — many of which turn out to be U.S. citizens that just don’t have their papers on them…. Are we really gonna be the Gestapo, ‘Where’s your papers?’ Is that what we’ve come to?”
—Joe Rogan, 1/13/26
In my view ICE doesn't represent any sort of authoritarian creep. They are a miniscule fraction of all law enforcement and they have a very specific remit. The coverage you see of ICE is heavily distorted by the biases of liberal media. Yes there are occasional incidents like the one in Minnesota but it's an isolated demand for rigor to focus on them. IMO that incident says far more about the state of protest culture than it does about law enforcement. I believe that the immigration crackdown is a clear net positive.
They're affecting some people's daily lives much more than the government used to.
Yes, illegal immigrants. That's what they're supposed to do and that's what reasonable people want them doing.
Criminals and criminal sympathizers are fond of painting reasonable law enforcement as authoritarian overreach.
A budget of more than Ukraine's Military per year is no longer "a miniscule fraction of all law enforcement."
It is if the combined budget of all other federal, state, and local agencies dwarfs the military budget of Ukraine ... which it does. It's important to remember that fractions have both a numerator and a denominator.
There are over 800k law enforcement officers in the country. ICE has a total headcount of 20k. I'm comfortable calling 2.5% a "miniscule fraction".
ICE is on a hiring binge, and they were already one of the lower-standards federal law enforcement operations.
They reduced their training period per The Atlantic (article also discussed lax hiring standards): https://archive.is/Fu8wN
NBC News reports various errors screening people that only get caught after they reach training (presumably others don't get caught at all): https://www.nbcnews.com/politics/immigration/new-ice-recruits-showed-training-full-vetting-rcna238739
Their current recruiting push is pretty creepy and seems like it's basically aimed at Internet racists:
https://www.newsweek.com/ice-recruitment-social-media-post-alleged-white-supremacist-link-2112555
https://www.thebulwark.com/p/the-nazi-slogans-are-not-an-accident
There are currently over 2000 as many ICE agents deployed in the Minneapolis-St. Paul area. By comparison there's about 1200 police officers on the forces of both cities combined.
Nationwide there are more ICE agents than FBI agents and ICE now has a much larger budget.
What is the point of all of this? America doesn't have high unemployment or crime from a historical perspective. We're marshalling a ton of resources to deport people (not that they're really hitting that high of numbers yet) as though there's some huge problem, but I don't really see it. And those resources and people could be spent doing something that would actually help.
> What is the point of all of this?
Jobs for politically aligned thugs, who will in turn support you when you refuse to give up the power at next election, because they know that most of them would lose their jobs otherwise.
It nominally performs a useful and necessary function. But so did its predecessor, the Immigration and Naturalization Service (INS). I've seen nothing that convinces me that the new and "improved" version is actually an improvement, and much to suggest the contrary. We can and should do better than this.
In particular, we should generally mock and ridicule and otherwise ignore anyone who suggests that law enforcement needs to "get tough on", well, much of anything. Since the days of Robert Peel, sensible people have understood that "toughness" is not the thing law enforcement needs to best carry out its mission. ICE was created as a way to make what used to be INS "get tough on" immigrants who might turn out to be terrorists, and has since broadened that mandate to no particularly good end.
All due respect, but you're not very aware of how many refugees are actually trained terrorists, are you?*
*trained by the CIA? One wonders.
How many refugees do you think are trained terrorists? Why did you choose that number?
Well, let's look at an example.
https://taskandpurpose.com/history/veterans-afghan-allies-shooting-politics/
So, we have at least 200,000 refugees from Afghanistan (yes, these are the people who are NOT coming to America, but for the sake of argument, let's say this was Harris' term, and they came). Scuttlebutt says there was very little screening (you can see that in the article), and I imagine men admitting homosexuality (without proof) is "yes, you're a refugee, congratulations!" -- it's basically a death sentence in Afghanistan, that has to count.
So, it is very easy to get someone into America.
Now, let's say someone is legitimately aggravated with American intransigence (we paid for the Taliban after all), or merely "lost someone they cared about." Given that they could have received training from American troops in Afghanistan, the Taliban, or even in the Ukraine, "training" isn't the issue (it's pretty freely available, but if you must, let's cut the number down to 50%, simply because women have substantially more difficulty gaining training). The general question is "how many are terrorists?" Pick a number: 1% seems maybe high (that's 2,000 people, which is a lot to embed if you aren't China), a tenth of that seems eminently doable (and "radicals" can spring up once they're in America, they don't need to be foreign enemies -- what is your rate of "terrorism" in America, for American citizens?).
This is actually a pretty rosy picture, though. Gangs that move to America en-masse are perfectly capable of being 100% terrorists (given that Latino gangs have been prosecuted and convicted for ethnic cleansing), in that they are a self-supporting structure. Whether or not they owe "loyalty" to someone back home (or merely have "connections"), these are semi-trained soldiers.
And that's without getting into "actual state sponsored terrorists" (like the Iranians that have been targetting our military command structure). It would be pretty ludicrous to believe that Russia doesn't have any trained assassins/terrorists operating in America, for example (Russia is notoriously sleepy, I wouldn't normally worry about them -- but refugee makes it easy for someone to come in...).
China, for example, says they're allowed to murder Chinese citizens on American soil. That takes guns (or other forms of assassination).
"% seems maybe high (that's 2,000 people, which is a lot to embed if you aren't China), a tenth of that seems eminently doable"
There is a huge difference between "eminently doable" and "has actually been done". The former is not evidence of the latter - not even if it's doable by our Totally Evil Enemies Who Would Totally Do That Thing, because even our evilest enemies have finite capacity and aren't going to do *all* the Evil Things.
Do you have any evidence that our Totally Evil Enemies have done *this particular* evil thing? As opposed to all the other evil things, and in some cases even *profitably* evil things, they could be doing.
What sorts of evidence are you looking for?
https://apnews.com/article/iran-fbi-justice-department-iran-83cff84a7d65901a058ad6f41a564bdb
https://time.com/7270415/iran-assassins-new-york-trial-conviction-guilty-verdict-masih-alinejad/
https://lieber.westpoint.edu/state-practice-assassination-what-is-old-is-new-again/
Has actually been done, as substantiated with American and non-American assassinations. As you can see, part of the "assassination trade" is a bonafide market where foreign actors can hire people to murder others (and thus the 'actual assassins' are making a "normal profit" and do not need to be continuously paid or embedded in outer Washington state, where rent is cheap).
But these are actual assassinations, not getting into "more home-grown terrorism" (aka "the Taliban told me to do it" versus "I'm an ISIS radical! See me explode!"), which is its own ball of wax.
Again, I pulled up a "reasonable talkthrough" of what could have occurred with the Afghan refugees, that the military was very concerned were inadequately vetted.
There's over a dozen routes out of Iran, some run by liberals, some run by the conservatives. It is not a place people are "locked into" -- if you (as a troublemaker) want to leave, there's a way out.
My argument was probably better about the Tren de Aragua gang.
https://nypost.com/2026/01/08/us-news/federal-agents-shoot-2-people-in-portland-oregon-day-after-ice-agent-killed-renee-nicole-good-in-minneapolis/ In that the entire group can be militarized.
John, thanks.
We not seeing much foreign terrorism. Are they biding their time or just not here?
Most people against ICE are incoherent. They won’t say they are for open borders but when it comes to the most basic part of immigration enforcement they suddenly treat it as illegitimate.
The most basic part of immigration enforcement is border security. Once you’re in the business of trying to figure out which random people are present legally (either as citizens or immigrants) and which are present illegally, you’re facing a very difficult problem. Especially if you’re in a country where citizens have rights.
1 million people at the beginning of Trump's administration were in the country illegally. As in the courts had already told them to leave.
Is "finding them" a very difficult problem? Yes, sorta... but it's a very different problem from "find and then prosecute" those the courts haven't already given their time to.
Not really, most other countries manage it with no problems. Legitimate residents almost always have a paper trail -- immigrants will have visa records, those born locally will have birth certificates and all sorts of other records of their existence.
False positives are possible even in a sensible system but they are rare and of all the complaints levelled against ICE I haven't heard of them deporting or even detaining for a long period someone who is actually a legit resident.
I'm generally in favor of allowing more immigration, but I don't think we need it to be completely unrestricted. But I still think there are a lot of criticisms of ICE right now. You can enforce the law without doing car chases and kicking down doors and everything.
And also generally I don't think law enforcement should be lying in court: https://slate.com/news-and-politics/2025/12/trump-lies-doj-ice-conditions-torture.html
ICE has been able to do their job in red states with little controversy because they don’t have thousands of people actively disrupting operations.
Are they doing this in rural areas? Or is this just Trump avoiding cities in red states? Because you’re not going to get much more cooperation with ICE in Houston or Memphis or Salt Lake City than you are in Minneapolis or New York or Boston.
Trump is obviously trolling by targeting blue states.
Texas prohibits sanctuary cities, which obviously makes a big difference. Generally what happens is that immigrant breaks some law, goes to jail, they realize the guy isn’t a citizen and gets handed over to ICE. Blue states often forbid this from happening.
I think the size of the deployments in some major cities is a factor. 2000 ICE agents in one major city is kind of insane.
There are tens of millions of illegal immigrants.
The previous comment deliberaty stated "2000 ICE agents in one major city". This reponse implies that there are "tens of millions of illegal immigrants" in Minnesota, with a total population of 6M.
ICE agents are not deployed proportionally.
If by tens of millions you mean "somewhere between 12 and 15 million. Then I suppose that's true. Most people using that term would be implying somewhere closer to the median though, and we are nowhere close to that.
That's kind of a non sequitur, unless you think they're all in Minneapolis.
(Also it's somewhat inaccurate. There's probably around 14 million.)
I think the intuition is that illegally immigrating is essentially a process crime, so there shouldn't be crackdowns for it, it's ultimately something very minor.
Stuffing the census box is not a process crime, it is an attempt to prevent the electoral college from working properly. Still feel like there shouldn't be crackdowns for it?
It’s millions of people that are in our country who don’t have permission to be here. “Crackdowns” are just enforcing immigration laws.
In theory, I approve of it. Immigration is a federal power, so the states can't decide to allow immigration, and the idea of cities setting themselves up as Sanctuary Cities where immigration law won't be enforced is criminal and gross. We shouldn't need to send federal officers to arrest people when that's the law of the land, but since we do, let's do it.
In practice; I'm far from anywhere they operate, but I'm also convinced it's a strongarm attempt from Donald Trump to harass the Democrats, and ICE officers are likely being encouraged to behave badly to that end. They want it to be big and loud and disruptive and ugly, for the sake of it. Using law enforcement as a cudgel is criminal and gross.
The point of “sanctuary cities” is like the point of doctors not turning people in for drug crimes and the point of client-lawyer confidentiality. You want people who are present to be willing to report local crimes to the local police, so that local police can do their job of ensuring public safety, rather than letting crimes go unreported and overdoses go un-reversed just because people are afraid of a different level of law enforcement get involved.
Doctors also refuse to turn in people for abusing children (including anally raping adopted children), so that they can properly track sexually transmitted diseases (AIDS really did a number on the whole "patient confidentiality" thing). The more you know!
Doctors and lawyers are expected to side with their client at the expense of the state. The Hippocratic Oath has been around for over a thousand years, and the state doesn't want to alienate the people following it. Lawyers are specifically a counterweight to law enforcement; it's their job to make things hard for them. Likewise priests and spouses are expected to have allegiances to God or family above the state, and the state has decided not to fight them on those.
The police should not have clients they prioritize above the law. Not only is it their job to enforce the law, everyone else is prohibited from doing so. Police who ignore laws are just thugs.
I see the whole thing less as doctors not turning people in for drug crimes and more as the DA refusing to prosecute for petty theft. At best it's a necessary evil because the system can't actually handle the load the letter of the law is placing on it. But the ideal solution is to make the system stronger so it can handle the load, like, for instance, having a dedicated division that focuses on this one crime, that the locals can call up and hand it off to. We have that, and it should be used.
Explain why they refuse to hand over not just witnesses but also perps?
Same reason doctors refuse to hand over parental rapists, one supposes. Sometimes crimes won't be reported if the perps will be sent home.
That seems like an importantly different aspect than the basic sanctuary city policy I've generally heard people defend. This must be about a substantive decision that the city doesn't want to contribute to enforcement of this kind of immigration policy, rather than just a sanctuary policy.
I've never had a direct interaction with ICE. My news sources paint them as relatively thuggish. Doing some Google searches I find some stats around them: https://www.ice.gov/statistics
Interesting the arrests are lower in FY25 compared to the Biden years. Maybe not all the data is in.
It wouldn't surprise me at all to learn at the end of this that actually the Biden and Obama administrations were generally more efficient in terms of total arrests and deportations than Trump 1 + 2. Not saying we can know for sure yet, but it would make sense to me that the administration that doesn't really respect process and is incredibly confrontational in every aspect of its approach is less efficient when it comes to any complex process that requires coordination and navigating communities.
Obama was tight, enough that Arrested Development was joking about it (and they had a fed on the writing team, so that's insider knowledge). Biden? Holy shit, no, he definitely just opened the door and let anyone stroll in.
I feel like the "thugishness" of a police force is mostly a function of the people they have to deal with rather than the police themselves.
If you're policing a population where the criminals tend to come quietly and submit to arrest then there's no opportunity to be thuggish even if you want to. If you're policing a population of George Floyds who insist on resisiting arrest and making life difficult then you're probably going to wind up looking a bit "thuggish" of necessity. If you manage to wind up in a situation like ICE where you're obstructed not only by the criminals you're sent to arrest but by a bunch of Karens with megaphones, it would take an absolute saint not to ever set a foot wrong.
You're assuming that the alleged criminal.absolutely has got to be arrested. If you take the view that non violent crime should not.escape into a violent confontation, you are going to end up letting more peop!e go. so enforcement against crime of property becomes laxer, but the overall mayhem is less.
Letting non-violent crimes slide tends to increase overall mayhem, not decrease it, partly because a lot of people who commit non-violent crimes commit violent crimes as well, and partly because people are more likely to act violently in an atmosphere of general lawlessness and decay.
The statistics do not bear that out.
That's a bit vague, but here's a study supporting the idea:
https://pure.rug.nl/ws/files/10429297/KeizerK-Spreading-2008.pdf
"We found that, when people observe that others violated a certain social norm or legitimate rule, they are more likely to violate other norms or rules, which causes disorder to spread."
To me it seems that American police officers in general often needlessly escalate situations into violence. I am not an expert, but I heard opinions that it is a consequence of their training, and possibly incentives at the job.
Basically, imagine the following situation: a cop sees someone who very likely considers doing some criminal activity. What is a better outcome from the perspective of the cop?
a) the potential criminal changes their mind, and walks away peacefully
b) the potential criminal decides to do the crime, but gets detained or killed by the cop
As I was told, on average, in Europe (a) is considered a better outcome, in America (b) is considered better. Therefore, a European cop will often try to reduce the tension and discourage bad activities, while an American cop will hope for escalation and then getting a bonus for handling the crisis.
You can see it in a recent situation, were an ICE agent unnecessarily spooked a civilian, and then used her panic reaction as an excuse to kill her. He might be technically in the right, legally, I don't know, but is this really what you want your police forces to optimize for? Almost any untrained civilian can be spooked into doing something stupid, so this basically gives police a pretext to murder anyone.
I dont' think this is a solid model. In America, you have at least three different police forces.
1) Urban cops. They know who the troublemakers are, and the troublemakers are conducting Business (dealing drugs, selling women). They have enough work without dealing with more potential criminals. (In my city, both "driving while white" and "driving while black" will get you harassed at night -- this is a model for "deterring criminal behavior" -- drug buying, breaking into cars, etc.).
2) Suburban cops. They are the people who will harass canvassers, and think their job is bullying people. They are responsible to the "white picket fence" people, and will do whatever it takes to earn money and keep them happy.
3) Middle-of-nowhere. This doesn't actually have a local cop force.
Nobody is really "getting bonuses" for detaining violent criminals (though they might for drug dealers, but again, those are businessmen).
We have a substantial population of people getting arrested 30+ times, as well, and that is concerning
> We have a substantial population of people getting arrested 30+ times, as well, and that is concerning
I agree, it seems like a systemic problem. I am not an expert, but it seems to me that a proper solution would require at least three parts -- changing the police behavior, changing how the prisons function, and changing the sentencing -- these three work together, and if people are being arrested and released over and over again, that seems obviously wrong, and creates an extra burden for everyone involved.
I think there are plenty of situations where the police would be happy for Scenario A, it means less paperwork, and you get to report a lower crime rate. However, mandates can come down to be "tough" on a particular type of crime, and that means you have to actually show an instance and a response, and suddenly scenario B) becomes much more heavily incentivized. That's also what we are seeing with ICE, the reporting I've seen indicates that their entire mandate is to round up as many people as they can possibly get away with. And when "people detained" is your measure of success, you are going to see a lot of questionable detentions.
One of the best points of comparison might be the Vietnam war. Originally, the metric for success in that conflict was "number of enemy combatants killed," which maybe makes sense when you think about a WWII type engagement. But it turns out, when you are sending people into villages in the middle of disputed territory, there is suddenly a strong incentive to kill random people and classify them as combatants after the fact, and this snowballed into some truly atrocious situations.
If a cop kills someone, or at least someone from the wrong demographic or political group, there's a good chance his face gets plastered all over social media and he finds himself on the receiving end of an international harassment campaign. You'd need some pretty powerful pro-shooting incentives to outweigh that.
>an ICE agent... used her panic reaction as an excuse to kill her.
I wish I had your mind-reading skills.
Do you think $700,000 are pro-shooting incentive enough?
https://www.gofundme.com/f/support-ice-officer-jonathan-ross
Firstly, it's very far from guaranteed that you'll get lots of GoFundMe money if you shoot someone.
Secondly, I think most people would probably forgo $700,000 if it came at the cost of being internationally hated.
Thirdly, depending on how the future plays out, there's a good chance the guy goes to prison, or gets fired and becomes unemployable for the rest of his life (would you hire someone who was fired for shooting a woman?), and $700,000 isn't all that much to support someone for a whole lifetime.
How do you get your face plastered all over social media when you're wearing a mask and your agency refuses to release your name?
We *probably* know who shot Renee Good because reporters were able to piece together fragmentary information to identify the likely suspect. I haven't heard of his being significantly harassed, and I haven't seen his face "plastered over social media" though I can find it if I specifically look. But if we're counting on "you'll be harassed!" as our disincentive to thuggish cops shooting people, then it's problematic that we're putting up unprecedented barriers to identifying the thuggish cops.
There's also the "you'll be criminally prosecuted" disincentive, but it doesn't look like we're doing very well on that front, and even "you'll lose your job" is kind of weak these days. What else have you got? Because maybe this cop wasn't a thug, but plenty are and they're paying attention.
The comment I was replying to was about "American police officers in general", who as I understand it don't generally wear masks. Maybe my understanding is just out of date, though.
I get that you're mostly speculating here, but the speculation doesn't make much sense to me.
1. How could European cops personally benefit from deterring a crime that never happens and that no one can prove they deterred? If cops are optimising for their own career recognition/bonuses, I don't see why these incentives would apply in America but not in Europe.
2. American civilians are vastly more likely to be armed which should prima facie make escalating much more personally *risky* for an American cop.
3. The sheer number of things many European countries criminalise, up to and including what you say on the internet, makes the idea that their culture values avoiding police confrontations with civilians hard to swallow.
4. Almost certainly the parisomious assumption is that higher US police confrontations are downstream of the higher crime rate. Any other explanation would need very strong evidence to overrule this one.
1. They cant, personally, but departments/forces can be judged by their ability to.deter crime.down to a low level.
2. Escalating first -- being first to draw your gun -- is less risky. that's why cops approach cars and buildings with guns drawn.
3. Yet the prison population is much lower.
4. It's the violence of the confrontations, not the number.
Escalating first is not less risky, in general. That's "gun specific." (and why we have officers with guns in the first place). "I have this autowin weapon trained on you" is a good way to quell the rational suspect.
They've often been acting in an unprofessional and needlessly inflammatory way under the Trump administration, but they're basically enforcing fairly normal immigration laws and the various forms of "ICE are the new gestapo" claims and active borderline-violent resistance tactics are taking their own side of that. I'd say "fault on both sides" - on the one hand as a government agency I'd say we should put higher expectations of responsibility on ICE, on the other hand the protestors are often doing outright illegal or counterproductive stuff while ICE is at worst using unproductive tactics in pursuit of their legal duties. In reddit terms I'd say ESH.
Specifically, on your comment on them ignoring local police I think the fault is on the other side. US law is pretty clear that they have jurisdiction to do immigration enforcement, and the common practice of sanctuary cities where local police refuse to cooperate is pretty bad for rule of law.
> ICE is at worst using unproductive tactics in pursuit of their legal duties.
No, at worst ICE is doing *illegal* things, not just unproductive things. Shooting a protestor who did not pose an imminent threat is a violation of that person's rights. Similarly, attacking people who are filming them from a distance is a violation of the 1st Amendment (freedom of the press, specifically). And they're also doing illegal searches of people's houses. They are unlikely to face legal consequences, because suing the federal government is very difficult, but they are doing a lot worse than just "unproductive" things.
Also, "Sanctuary cities" are not the local police refusing to cooperate on a whim, they're cities passing ordinances that limit when local police can arrest people for immigration violations or share that information with ICE. I don't think you can say it's bad for the rule of law when the local police are literally following the law by not cooperating.
" Shooting a protestor who did not pose an imminent threat is a violation of that person's rights."
Whether she posed an imminent threat is something the two sides disagree about. People on both sides are lying about the facts. On DSL I observe intelligent people who believe she did, others who believe she didn't. I have viewed multiple videos and am not sure. She was driving a car that was accelerating in his direction, it is unclear whether he actually got hit and whether he could tell that he could get out of the way.
>People on both sides are lying about the facts.
Don’t both sides this David. If you are talking about Trump and Kristi Noem v Mpls mayor Frey, you aren’t being virtuously even handed, you are being willfully blind.
One side could be mistaken but only one side is putting forth egregious fabrications.
We may or may not get to the truth as the Minnesota Bureau of Criminal apprehension has been denied access to evidence by the Feds.
Edit:
Odds of getting to the truth decreased with further information today.
“Mass resignation at Minnesota U.S. Attorney’s Office stems from Renee Good shooting”
(should not be paywalled)
https://www.startribune.com/joe-thompson-us-attorney-who-uncovered-massive-fraud-in-minnesota-resigns-from-office/601563206?utm_source=copy
What Trump and Vance said was false. So is the implication of the comment I was responding to, that it was a fact that she posed no threat to him. I have seen multiple comments that simply omit the relevant fact — that he was in front of the car she was driving and she was accelerating. That too is false, by implication if not by assertion.
I think it's pretty clear he got hit:
https://youtu.be/bDda-L_ZOE8?si=jUx1Tq8d04RMdFcG&t=459
Not a credible video. At 8:35, the narrator says, “So, now what you are going to hear in a lot of the media is that the officer placed himself in front of the car. But as you can see, that’s not accurate, because while the officer was standing still the car actually repositioned itself and pointed the front of the car directly at the officer.”
There are some white lines painted on the pavement, and the video shows them rotating relative to the camera, which means that the camera is moving. The narrator is claiming that the video shows that the officer is standing still, when in fact is shows the exact opposite.
The narrator continues, “Now as the officer was trying to move towards the driver’s side of the vehicle to get from in front of the car, the person hits the gas, striking the officer.”
These assertions aren’t contradicted by the video footage, but they aren’t supported by it either. After we lose sight of the white lines on the pavement, there are no good reference points to determine whether the officer is moving. To show the car striking or failing to strike the officer, the camera would have had to have been pointed down. It wasn’t.
I’m not saying that the officer wasn’t struck by the vehicle, just that the evidence provided doesn’t show that he was.
At the 6:54 mark, you can see the officer get hit:
https://youtu.be/bDda-L_ZOE8?si=5kFDNziuih7jnwfh&t=414
Ultimately I don't think it matters that much, and all this analysis is beside the point.
If I grab a gun and start shooting in the general direction of a police officer then that officer is of course justified in shooting me dead. If a careful analysis of the video tape later shows that I was aiming two feet to the left and the bullets would have missed anyway, this is irrelevant.
I agree that if you make a car a gun then the argument is much easier, also if my grandmother had wheels she'd be a bicycle.
These operations are occurring in suburban settings with huge numbers of people driving around in vehicles. Are you telling me every time a vehicle is moving in the same general direction as an ICE agent lethal force should be authorized?
> I don't think you can say it's bad for the rule of law when the local police are literally following the law by not cooperating
It's bad for the rule of law when local governments pass laws that attempt to prevent the application of federal laws. Or the other way around, for that matter.
It is good for the rule of law when the law is clear, and consistently enforced.
It’s good when the people in charge of enforcing one set of laws have no part in enforcing other types of laws, because then people who are breaking administrative laws don’t worry about reporting violent crimes, because they can report the violent crimes to the violent crime police without worrying about getting caught in their administrative violations.
Does that go both ways? Like if a parking cop sees an assault should he ignore it?
It's hard for me to imagine what situations there are in which it is so important to have someone report a mis-parked car that we would want to ignore evidence of crimes to get that report. In fact, I don't think parking cops even *take* reports of mis-parked cars.
The question has to be, which sorts of contacts with authorities are so important to encourage that we should ignore violations of other sorts when someone is engaged in that sort of contact?
Someone who thinks immigration violations are equally important to report as violations of labor law might want to give sanctuary to people who reveal that they are violating labor law while they are reporting immigration violations. I wouldn't support this policy, but I could see the motivation for it.
Yes. I'm surprised I'm not seeing more comparisons to 1861. Once again, state and local governments are going against the federal government with the goal of preserving an easily exploitable non citizen workforce which directly competes with the poor and working class for jobs and thus holds down wages at the lower end of the economic spectrum.
Yes, because the predominant problem with slavery was the economic concerns of poor whites.
I’d say the situation in Minneapolis today is more akin to the pushback the feds encountered enforcing the Fugitive Slave Act of 1850 in the northern states.
Except these aren't slaves brought here against their will. Instead they came here against OUR will.
Yeah. I'm not as convinced that their behavior is *as bad* as you say, but they are certainly being more aggressive.
Mostly because (a) they weren't allowed to do much in the way of enforcement before, so anything will appear more aggressive and (b) the resistance tactics make it either so they have to be more aggressive OR simply do nothing at all. Their previous tactics are simply off the table, to the degree that they're doing anything at all. And it's very abundantly clear that *any* amount of publicly-visible enforcement outside the direct border zone would be unacceptable to large swaths of the resistance types.
As far as ignoring local police, that's exactly what one would expect. That's what they should do, especially when local police are required to not cooperate by state law. It's the flip side of the idea that state resources cannot be commandeered by federal authorities--if the state doesn't want to cooperate, the feds (who are supreme in any area of conflict, by the Constitution) can and should just do it themselves. It's going to be meaner/rougher...but that's what those states explicitly signed up for. Letting lack of state cooperation equal no federal enforcement was one of those things we decided against allowing with that little dust-up known as the Civil War.
What's the basis for your claim they weren't allowed to do their job before? The Biden administration deported a lot of people: https://www.npr.org/2024/12/20/nx-s1-5235329/why-deportations-hit-a-10-year-high-in-2024
Biden also encouraged a "surge to the border", which led a ton of people to come and claim asylum, despite not actually fleeing persecution. I WILL grant that arrests under Biden were higher (https://www.ice.gov/statistics), but as of the latest official estimates I could find, we've had 2 million illegal residents living here: https://ohss.dhs.gov/sites/default/files/2024-06/2024_0418_ohss_estimates-of-the-unauthorized-immigrant-population-residing-in-the-united-states-january-2018%25E2%2580%2593january-2022.pdf
The Left seems to be okay with immigration enforcement as long as their own side is doing it and they don't have to hear about it at all. But as soon as somebody actually brings it up they shout "No human is illegal!" and act like no border laws should exist at all.
While I agree that the abuses cited by Jack further down: ("Examples abound on the internet, but I'll go with the supreme court case.") should be punished, I don't think the Minneapolis shooting was one of them. Also, I think the heavier, more visible enforcement serves as a deterrent. No more "surges", no more fradulent asylum claims, and illegals have an incentive to self-deport (or at least keep from breaking any other laws).
I currently think that while ICE should be held to the highest standards, they still need to be actively engaged in detaining and deporting illegal immigrants, especially those who've committed additional crimes, and if Leftists like former judge Hannah Dugan and Gov. Tim Walz seem determined to protect them no matter what, that calls for more strength to enforce the law.
> Biden also encouraged a "surge to the border"
People claim this all the time, but what actual evidence is there other than temporal correlation?
> which led a ton of people to come and claim asylum, despite not actually fleeing persecution
Again, what's your source for this?
> as of the latest official estimates I could find, we've had 2 million illegal residents living here
Is this a typo? Weirdly low take since that document says it was almost 11 million in 2022, but also that's basically unchanged over a decade.
> The Left seems to be okay with immigration enforcement as long as their own side is doing it and they don't have to hear about it at all.
It really irks me when people sum up "The Left" (or "The Right" like this. I've spent time in some fairly leftist spaces and can confidently say SOME OF "The Left" have been sh***ing on Obama and Biden for years for their anti-immigrant policies.
Biden's "surge" quote:
https://www.youtube.com/watch?v=rYwLYMPLYbo
Asylum Fraud:
https://web.archive.org/web/20180705172604/https://www.newyorker.com/magazine/2011/08/01/the-asylum-seeker
and
https://web.archive.org/web/20260104093938/https://www.nytimes.com/2011/07/12/nyregion/immigrants-may-be-fed-false-stories-to-bolster-asylum-pleas.html
My bad on the Estimates citation. I was looking at Table 1, pg 4, the "2010 or later" row. You are correct that it's closer to 11 million. The fact that it's been that way for a decade I'd argue furthers case for more effort to fix the problem.
I do recall seeing a comment on a news article that some immigration rights group was unhappy with Obama's deportations, but that's the only published criticism I remember seeing that a Dem president was being too tough on illegals. By contrast, there was quite of a lot of such criticism during Trump's first term, and that was when he was more talk than action on the issue. If your friends were paying close enough attention to notice Obama and Biden's deportations, good for them, but I don't remember seeing much outrage on that from leftists with a platform. Perhaps I should've specified "Leftist leaders", instead of implying the entire Left felt the same.
German here: ICE is being used much like Hitler used the Sturmabteilung (SA), to create state terrorism (in the literal meaning, to spread fear). It's not a perfect analogy because the SA was technically civilian, not an established state security apparatus, but the goals are much the same: create chaos in domestic politics, play to the political base, rough up the opposition and clamp down even harder if and when they fight back. Trump needs a para-military organisation to do it, one that has the numbers and impact and is taking direct orders from the White House rather than the states. ICE and DHS have been helpfully provided by previous governments, and Trump (through loyalists like Noem) only has to give them the marching orders.
Actually the point of ICE is to deport illegal immigrants.
Yes, and the SA was officially the Gymnastic and Sports Division of the party. People say a lot of things, don't they?
Last I checked ICE does in fact deport illegal immigrants and all their operations are going towards deporting illegal immigrants. You can’t just make up things.
That may have been the point imagined by the law that set that agency up. But Trump has greatly transformed how this agency is being used compared to Biden or Obama or Bush (and it didn’t exist before Bush).
Is your familiarity with ICE limited to a few clips on Reddit? I feel like announcing that you're a German doesn't increase the credibility of your knowledge of the SA, and it only decreases your credibility of knowledge of American politics.
Feel free to lay out your opposing point of view on the matter instead of whatever you're trying to do.
My opposing view on the matter is that it seems like you're using your German nationality to claim you have a better understanding of fascism, comparing ICE to the SA. I don't think being German gives you a better understanding of the SA, and probably gives you a worse understanding of ICE.
My opposing view is that there is very little in common between ICE and the SA, at least not any more than between the SA and any police unit.
I agree with you on this characterization
>My opposing view is that there is very little in common between ICE and the SA, at least not any more than between the SA and any police unit.
One fundamental difference to regular state/county police (and commonality with the SA) is that they are under direct control of the federal government. The commanders of ICE/DHS also don't have the strong traditions of the constitution-bound military that continue to be an obstacle for Trump. Yes, there are also the FBI, DEA, ATF and any number of other 3 letter agencies that he could use, but Trump focuses on ICE because that's where he has the most political support from his base. You know, because "they took our jerrbs".
Another difference is the sheer scale that Trump II is aiming for.
>The so-called One Big Beautiful Act allocates more than $170 billion over four years for border and interior enforcement, with a stated goal of deporting 1 million immigrants each year. That is more than the yearly budget for all local and state law enforcement agencies combined across the entire United States.
https://www.brennancenter.org/our-work/analysis-opinion/big-budget-act-creates-deportation-industrial-complex
> Altogether, this marks the largest investment in detention and deportation in U.S. history; a policy choice that does nothing to address the systemic failures of our immigration system while inflicting harm, sowing chaos, and tearing families apart.
https://www.americanimmigrationcouncil.org/press-release/congress-approves-unprecedented-funding-mass-detention-deportation-2025/
I claim that the harm and chaos are not incidental. The Cruelty Is the Point, as they say. Together with Trump's well-known sympathy, admiration even, for dictators around the world and disregard for the rule of law both domestic and international makes me draw these comparisons.
How is that a point of commonality with the SA? The SA was a party militia that was not under the control of *any* government. And right after the party in power took control of the national government, they set about ruthlessly eliminating the SA.
This seems like a very important distinction. Paramilitary forces whose opponents consider oppressively thuggish and authoritarian are a thing, but they're a thing that takes very many forms and the SA is in a fairly weird corner of that space.
Ice is nothing like the SA my friend. They aren't going around roughing up citizens or politicians, they are literally just enforcing existing immigration law.
>They aren't going around roughing up citizens or politicians
https://en.wikipedia.org/wiki/Deaths,_detentions_and_deportations_of_American_citizens_in_the_second_Trump_administration
>they are literally just enforcing existing immigration law
First of all, my friend, when I think of the Trump II administration, "law-abiding" is not exactly the first term that comes to mind. The ongoing law suits certainly tell me that there is disagreement on whether they're really just enforcing the law.
https://www.politico.com/news/2026/01/12/minnesota-sues-trump-ice-minneapolis-00723482
Second, even if they were just following orders, errm, the law, you seem to underestimate the considerable leeway the executive has when it comes to the manner in which they enforce those laws. In the end, it's ordinary people who do the actual enforcement, and everybody has a boss that can tell them either to "go easy" or to "go hard". When you're lower on the totem pole, are you going to say "no, the law says something different"? Certainly not if you want to keep your job.
Lawsuits in the US are only weak evidence that the person being sued is guilty. In fact, they are weak evidence by design, in any country that presumes innocence until proven otherwise. They are also a signal that there exist people with motivation to oppose ICE, and we already have copious evidence of that.
While there can exist LE leaders who encourage their members to be overly enthusiastic about their jobs, there also exist LE leaders who encourage their members to be lenient, and criminals who are incentivized to exploit that leniency.
The SA was a paramilitary attached to a political party which engaged in streetfights with other party-affiliated paramilitaries. That just isn't at all what ICE is here. We're not even close to an election where roughing up the other parties would be that relevant, nor are there more controversial actions concentrated in swing states.
>The SA was a paramilitary attached to a political party which engaged in streetfights with other party-affiliated paramilitaries.
They were that in the beginning, yes, but their duties of course shifted after there were no more other parties. I mean, it's just people, made important by giving them uniforms and guns; you can tell them to do other things than before and they'll do them. It doesn't matter how or why they were founded initially, nobody cares.
What matters is the here and now, and if you don't think the SA was back then or ICE is now being used for intimidation, then we just might have very different ideas of what counts as intimidation.
Intimidating rival parties was a big part of the point of the SA (and after other parties were eliminated eventually the SA was shut down as well, with the SS taking over). ICE isn't doing that, they really can't do anything about Democrats being projected to win the midterms.
They're Feds. Be extremely careful around Feds. Follow the Rule: "Don't talk to the police" (1). You are not an exception to the Rule. Be polite, say nothing, get a lawyer, get a lawyer, get a lawyer. Remember, the precedent was set over 30 years ago that a Fed can shoot your teenage son dead in the back as he runs away and get off scott free. (2)
This is really bad. Rodley Balko writes about this very well, at least he did in "Rise of the Warrior Cop" and he's got a Substack now, it's probably good. The only semi-defense to the increasing militarization of the police is that a lot of Americans really are heavily armed psychos. Like, I'm not sure Germany even has prison violence (3) whereas California is like, "Welp, we've had 7 intra-prison murders in 9 weeks among these unarmed and heavily supervised prisoners in prison, guess we better lock down." (4).
Having said that, ICE does not seem uniquely terrible and is probably a lot better than the ATF. Most of the issues come from Sanctuary cities saying they won't provide any police support or any assistance whatsovever for ICE because, well, they would really like the ability to nullify federal laws they don't like. Worse, they're not going to do anything about repeated assaults and sabotage of ICE personnel. This is a recipe for a tragedy and, well, one happened recently.
I don't like Feds and I really don't like Feds deployed en masse around law abiding citizens. That's one step up from military deployment. Having said that, I would really like immigration laws enforced. I have voted for it in the past and will probably vote on it in the future. I would really, really appreciate it if there was some way to enforce immigration laws without the mass deployment of Feds to US cities. Since apparently that's not an option, I am reluctantly ok with ICE.
(1) https://www.youtube.com/watch?v=d-7o9xYp7eE
(2) https://en.wikipedia.org/wiki/Ruby_Ridge_standoff
(3) https://www.bmjv.de/SharedDocs/Downloads/DE/Fachpublikationen/Criminal_Justice_in_Germany_8_Ed.pdf?__blob=publicationFile&v=2
(4) https://www.corrections1.com/jail-management/surge-of-violence-prompts-crackdown-in-calif-prison-system
>"…probably a lot better than the ATF."
Lowest bar imaginable; even the fucking DEA is better than the ATF.
Your attitude towards law enforcement personnel seems to be strongly dependent on whether they're Federal or state/local? Why is that?
IDK as I read it they don't seem wild about either, Feds are just more the focus of the topic.
There is a certain fictional version of American law enforcement you might have in mind, born of fiction, where they have to tell you they're cops if you ask, have to give you one phone call when you're arrested, can't use evidence they collect without a warrant, can't persuade you to take illegal actions without it being entrapment, etc. These are false.
What you're seeing with ICE is what federal law enforcement looks like in practice; When a clear show of force will elicit compliance, they'll be dressed up in clearly identifiable gear, knock on your door, and present a warrant and a badge. When not, they'll be sneaky, conducting surprise raids, and bust your door down without knocking. With the Democrats, it'll be the ATF instead, with the Republicans then up in arms when they do much the same thing against Americans with underdocumented firearms. And both of them support the DEA.
The current practice of ICE agents wearing masks has nothing to do with "being sneaky," as evidenced by how often these fuckers are getting caught on video by random citizens. The only purpose is to protect ICE agents who violate the law from being identified and punished.
I just do not understand that. If they do not wear any signs/badges, what theoretically prevents some bad actors to disguise as ICE agents and loot someone's house under the pretext of searching for illegal immigrants?
People are doing exactly that, and increasingly since Trump II ramped up enforcement.
https://prospect.org/2025/06/24/2025-06-24-ice-impersonations-proliferate-agencys-undercover-tactics/
I'm pretty sure the only purpose of the ICE masks is to prevent the spread of COVID.
I figured they were also wearing masks to keep their faces warm. Minnesota in January is snowy; I don't think that's controversial.
> The only purpose is to protect ICE agents who violate the law from being identified and punished
You don't think there's also some kind of purpose in protecting ICE agents and their families from being identified/harassed/assaulted outside work hours?
Isn’t that illegal for protesters to do? Shouldn’t protesters have more rights than federal employees trying to impose the will of the state, in a free country?
> Shouldn’t protesters have more rights than federal employees trying to impose the will of the state, in a free country?
I mean, not in general?
There are things that police officers on duty can do that random civilians cannot (like, say, arresting people), and there's certain things that police officers on duty can't do that random civilians can (like, say, smoking a cigarette).
Some of these things are situational and mission-dependent. Under most circumstances, most police officers should be wearing uniforms and clearly visible. But situationally, plain-clothes police officers have been a thing for a long time. Similarly, under most circumstances you want to be readily able to identify a police officer, but there are circumstances (like when raiding an organised crime figure's house) that it's acceptable for individuals to try to conceal their identity so that they're not individually targeted for reprisals afterwards.
Some countries have a solution that the masked police officers have some visible identifier, like a number, so that you can later complain to the police that e.g. a masked officer #12345 destroyed your property, and in theory police can do some internal investigation. Hiding the number is considered a bad thing.
How is it possible to redress violations of people's rights if you can't identify the officers responsible?
Public servants need public oversight. You might be able to justify hiding identities on a case by case basis (e.g., undercover agents, high profile cases), but you can't say "we're hiding all identities of all officers, just to be safe." That's just secret police.
(I mean, if you can justify it for every single ICE officer, why stop there? Give *every* police officer a mask! Someone might decide to doxx the local traffic cop because they're mad about a speeding ticket!)
> How is it possible to redress violations of people's rights if you can't identify the officers responsible?
Like if a cop sneaks up behind me and kicks me in the butt and runs off or something?
I dunno, how do we redress that if it was a civilian who kicked me in the butt?
Well, these days there's probably a security camera somewhere that caught the event on video, so maybe we can look at that and see who the butt-kicker was.
It really helps if the butt-kicker is visually identifiable, rather than one of several thousand identically-uniformed masked men.
Especially given the open, public calls for *exactly that, up to and including murder*!
Was there evidence of a need for that *before* they started brutalizing people? Or did people reasonably get upset about the literal secret police and now you're using that to retroactively justify the use of masks?
There is never a need for murdering law enforcement. The reason the feds don't have body cameras is that certain federal agencies (not ICE) spend their day breaking the law.
People are not "reasonably upset" about the "literal secret police" who are entrapping people into being considered terrorists. I supported ending the Patriot Act, and I suggest you do as well. I further would enjoin you to support the President who tried to pocket veto the Patriot Act.
What kind of evidence would you accept, and what's the cutoff date for "before"?
Against, and if it were up to me, we'd abolish it.
Keep in mind we've had immigration enforcement for a long time, but ICE is only ~20 years old, a part of post-9/11 reshuffling of the government. It's supposed to be about enforcing immigration laws with an emphasis on criminals, but during this administration they have ramped up enforcement against people who are not criminals (other than being in the country illegally, which isn't always a *crime* technically; and sometimes against people who aren't illegal immigrants at all).
Law enforcement against violent criminals can be done by the same people regardless of whether the targets are immigrants - e.g. if the DEA investigates drug dealers, they can investigate drug dealers who are also immigrants, etc.
The culture seems rotten, as can be seen by their actions and the defenses of them. Steady stream of brutalization of immigrants, of protesters, of legal immigrants who this administration decides to target, and of random people who are in the wrong place at the wrong time.
The current stuff in Minnesota is clearly more about punishing people this administration doesn't like, than anything about crime or immigration.
> Against, and if it were up to me, we'd abolish it.
What would you replace it with? Or would you just give up on sovereignty/rule of law entirely?
Do you think "sovereignty/rule of law" didn't exist until 2002?
Border protection (i.e. the actual border) has another agency. Criminal investigations are done by other parts of the federal government, and if someone involved is an illegal immigrant you want to deport then you need a bureaucracy with a bunch of pencil pushers to sort through it, not a separate guys-with-guns agency that runs torture prisons.
I don't really feel the need to expend a lot of resources on chasing down random DoorDash drivers with no criminal record, but if you really want to have an agency that does that then you can. It doesn't need to be the same agency that is supposedly fighting terrorism and international networks of drug dealers and cosplaying a military occupation.
Most importantly, even if you replaced ICE with a carbon copy of ICE, you'd fire everyone now working there, zero out their pension, and bar them from ever working for the federal government.
Where would you find a large reservoir of people willing to work for immigration enforcement (after you just fired everyone) who are all sufficiently competent and ethical to convince a hostile media of their good intentions?
I'm not saying ICE can't do better (it seems like it clearly can, both in terms of top-down organization and hiring). But the idea that you can just fire everyone and hire good people instead is impractical.
Congress dropped, what, a 12 figure budget for ICE? They care to spend that much they can find people.
And if not, well, you're already not talking about people committing other crimes or people at the border - I care more about not having a hostile army assaulting random people on the street and occasionally murdering them, than I care about the ability to chase down Doordashers.
And for people who really care about interior enforcement, I'll just say that it behooves them to think of a better way to do things, because they're on track to lose this fight.
> I care more about not having a hostile army assaulting random people on the street
Can you point to any cases of ICE assaulting random people on the street?
> other than being in the country illegally, which isn't always a *crime* technically
In which cases is it a crime, versus not being a crime? In cases where it isn't a crime, what would you call it?
When you enter legally, but overstay your visa, or work without authorization, that's a _civil_ infraction. There's no jail time, and I don't think there's even a fine; the penalties are just deportation, making it harder to get a US visa, being barred from entry into the US, … that kind of thing.
Illegally entering the country, or falsifying documents or misrepresenting your status to immigration officials, for example, IS criminal.
It's a civil violation, not a crime. There are a number of related situations that could be crimes, such as illegal re-entry after deportation.
Just looked this up. The act of being in the country is a civil offense, but the act of illegally crossing the border is a federal crime.
I don't really understand the point of the American habit of wanting to divide up illegal acts into a zillion different categories and then nitpick about exactly which kind of illegal act falls into which. Crimes, felonies, misdemeanours, "civil violation" which is a whole category I've never heard of before, and all sorts of other things.
Of course illegal acts can differ in severity, but these boundaries don't seem to properly capture that anyway.
This isn’t an American thing. This is a general feature of the legal system of most countries with a history of elections and popular input into the law. In places where “law” is just the will of the sovereign, there may not be all these categories, but it’s really important for there to be different categories for violations of administrative rules vs things that directly harm victims.
> "civil violation" which is a whole category I've never heard of before,
I don't know what you mean here, because Australia has a clear civil/criminal distinction, and things like defamation, copyright infringement, along with more controversial ideological things like unfair dismissal, non-discrimination, and "hate speech" law are usually in the civil category, and moreover it seems very important that they're in this category (even if I'd much prefer that some of these not be illegal acts at all). Compare the UK, which makes a lot of these criminal, such that you can go to jail instead of just having to pay someone damages, and seems significantly less free as a result. Compare whatever US law was used to declare Trump's misleading campaign spending not merely a civil wrong (reasonable) but an actual crime (dangerous government overeach, particularly since at the time he was a political candidate who had never held a public office). This distinction can be a check on state power, and I don't see why immigration would be different.
Felonies and misdemeanors are both crimes, with felonies being more serious. A parking violation is a civil violation. It's not considered criminal.
Even if you haven't heard the term, I think most people have an intuition that not every minor rule violation is criminal.
If you're simply saying you think being in the country illegally should be considered a serious offense, that's different than wanting to abolish the distinction between crimes and civil violations altogether.
I'm not saying I want to abolish the distinction, I just don't particularly want to read pedantic threads about "oh that's not a crime, it's a flurbnurgben, which is a slightly different type of illegal act".
As an American: ICE is an organization I rarely think about. I think that is in turn a function of my occupation (software development); if I were working in construction or farming, I suspect I would be more aware of them. To the extent I think about them, it's mostly just due to them being in the news; I never encounter them personally.
ICE's job mostly involves ensuring people who aren't US citizens are here precisely when law permits. This lately focuses around two major areas: watching for people sneaking into the country, overstaying work visas, or otherwise being here when they shouldn't, and doing something about them when found. ICE's ability to do this is sensitive to budget and staffing, esp. since there are millions of illegal immigrants estimated to be living here, and that budget probably tends to be low during Democrat administrations and high during Republican ones these days.
Since the Trump administration has seen unusually high positive feedback from its base regarding illegal immigration, it's seen fit to expand ICE rapidly. One can expect rapid staffing to result in some corners cut on training and overall lower average job experience, which can manifest in more errors in performance. OTOH, low staffing can likewise result in shortfalls in the ability to enforce immigration law, so the probable plan here is to expand, tolerate mistakes within reason, and allow experience to accrue.
The job of finding people who don't want to be found is a non-trivial one, even within US borders. It's exacerbated by US citizens who also don't want such people to be found. To the extent that function is necessary, then, it often needs to be done with methods similar to other LE - plainclothes cops, sting operations, and even methods for dealing with accessories to illegal immigration, and with local police who appear unwilling to assist with immigration enforcement. While delegating LE to local police has some nice properties, so does having an agency specializing in detecting illegal presence, including going into the field and detaining and deporting if necessary.
It's possible to have such an agency's officers get overzealous in pursuing their mandate. Even people in favor of immigration LE often agree (ask them about ATF, or DEA, or EPA, or FDA, or FAA, or state troopers, or...). Not everyone who believes in the mission tempers that belief with consideration for limits, and the US tends to change the leaders of such agencies to align with the platform of the current political party, and the leaders dictate staffing and policy.
Both of our major parties likewise are more focused on what their own base thinks than what the opposition thinks, so the meta-game tends to center around getting into power, then doing what your base wants, then making sure your base is aware of that around election time, while discouraging the other party's base from coming to the polls. That falls out as running press releases about the successes of your party's favored agencies and the failures of the less favored.
There's an obvious implication here that the nation would be better served by planning that goes farther than just the next 1-2 elections, but that would probably require encouraging voters to think that far ahead, and I'm honestly not sure how to get that going at scale.
ETA: it occurs to me that thinking in only-next-election terms is probably encouraged by a news industry incentivized to report about how We're All Gonna Die Soon, so maybe get the news industry to stop doing that as much?
> that the nation would be better served by planning that goes farther than just the next 1-2 elections
how is it technically possible if there is no understanding what the administration would look like next election, not even mentioning 2 elections from now? you could plan for sure, but the next guy in the office would just cancel all those plans and do his own planning, then it repeats the next election etc
Venezuela's oil investments show that companies are willing to invest in "the next guy not being completely crazy." So, yes, if you dangle enough money in front of people, they're willing to say "we'll help your team get elected -- or at least the not crazy part of the other team."
Up until Trump, most people thought the government was run by laws, such that you had to change the laws if you wanted to cancel all those plans. But Trump started cancelling huge numbers of plans without changing the laws, and the courts have determined that this is not illegal, so now we are in the situation you worry about.
I really think that after this, we need to revise the constitution to ensure that it is possible for the country to make promises that last longer than one presidency.
There's a common American belief that Americans used to think farther than the next election, roughly aligned with the belief that Americans had the attention span to watch things like 3-hour political debates. To the extent there's any truth under that belief, it was likely based on the fact that most federal-level politics didn't affect most of American life (until 1828, 1865, 1888, 1916, 1932, 1965, 1972, 2001, 2008, or possibly other years depending on who you ask). Most things didn't change that much for most people, no matter who was in the White House or Capitol.
There are still a lot of Americans who believe this is still the case, and an even greater number of Americans who contend this isn't the case now, but ought to be. That belief goes in hand with a mindset of solving one's own problems, rather than waiting for some other person with Broad Sweeping Power to fix it for everyone. The mindset's biggest frustration isn't with illegal immigration or high unemployment or trade deficits, but rather with the nanny state. (This mindset can be hard to detect, because it intrinsically spends much more of its time working problems than it spends complaining about them. There are a lot of Americans who voted Trump mostly because they were tired of Democrats trying to run every facet of daily life, didn't want Trump doing that either, and are annoyed at people who voted for Trump hoping HE would solve everyone's problems. But again, they're too busy to hang around megaphones.)
For this cohort, life revolves around planning the set of things that won't change every election, and trying to oppose any political effort to shrink that set. And that set was indeed once a lot larger; people built their own houses, farmed their own land, raised their own kids, cleaned up after their own businesses, and sometimes even fought their own fires and enforced their own laws. If they were unhappy, they usually knew why (radiator belt needs replacing; bugs ate half the corn; someone in town got killed); if they didn't, they could talk to a priest or a bartender, and they'd be around for decades.
I don't know your background, so this is about the best way I can generally portray this group without getting too specific.
SWE-bench verified milestones
2024-10-29: first model over 50%
.
.
.
2025-02-03: first model over 60%
2025-03-16: first model over 65%
.
2025-05-15: first model over 70%
2025-06-12: first model over 75%
.
.
2025-09-28: first model over 78%
.
.
.
Today: Top model is still 78%
SWE bench verified is 80% on openai blog
https://openai.com/index/introducing-gpt-5-2/
also look into among others, swe rebench
Yeah I don’t know why they hasn’t shown up on SWE bench website. One theory I saw was, even if they ran them on SWE problem set, that they maybe didn’t run the tests under the narrow constraints required by SWE for their list. Idk. I just know the number on that site was going up fast and now it’s not going up at all.
A lot of benchmarks saturate below 100% due to problems being broken or ambiguous
True, but in response to those exact claims about their regular problem set, SWEBench specifically created the “Verified” list as a human-curated subset of clearly defined problems.
Is your point that progress seems to be stagnating? As a counterpoint, wouldn't we need to price in that work output of white-collar jobs in first world countries typically lowers greatly over the holiday season?
Well that’s certainly a type of “AI Winter” I haven’t heard anyone factor into their forecasts
*Very* long shot, but I've been trying to get in touch with an ACX reader, X username @thinkingshivers, who does not have any publicly posted contact info. Shivers, if you're reading this, 1) I love the adorable horse game, and 2) I would love to talk to you about the adorable horse game! My e-mail is easily googled; hmu.
I'm not who you're looking for, but I can never pass up an opportunity to plug https://www.themanequest.com/ when people discuss their love of horse games.
If you know their twitter username, maybe try direct message on that site?
I am very curious…? …engaged?…This reminded me of reading the personal ads on the back page of the Village Voice back in the early 80s. I hope you connect.
Re:pens, anyone here does penspinning? I think many of you would enjoy it!
Didn't realize this had a name! All the honors students in my high school/class started doing it in freshman year so I learned and now that's my fidget move.
I get asked if I'm a drummer a lot (I'm not, I play bass lol)
You can still find me doing a twisted sonic --> backaround --> reverse sonic loop to this day.
Some. In college, I took a class and, while I didn't get anything intellectual out of the course, used all the lecture time to practice a couple of spins that I had seen.
what's your profile picture and what's your logo mean?
Mine means I'm lazy.
A screenshot of Unicode character U+1F431, as it looked when the screenshot was taken.
Now it looks differently; I wonder who spends the resources to keep updating emojis.
It is a cat face, because at some moment I needed a nickname, and I chose "kittenlord".
My profile picture is a commissioned piece of art of me if I were a My Little Pony: Friendship is Magic character. I had it done by an artist selling commissions at a MLP:FiM convention.
Sadly, the stupid circular profile pics edits out my "cutie mark," the symbol on the butt which represents a character's special passion or talent, which is...
...a piece of art I created of myself if I were a Southpark character (I once used the Southpark character creator plus Photoshop to make a huge custom moodtheme set for Livejournal; it included time-intensive projects like a black and white version with the colors in a puddle at the bottom of the image to be used for feeling "drained.")
Full effect is here:
(https://substackcdn.com/image/fetch/$s_!eHep!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com%2Fpublic%2Fimages%2Fa25eae9c-cd1d-4c12-ae31-d195272a5426_318x350.png)
I remain enormously amused by the implications of that cutie mark, and that, arguably, my capacity to still be amused by its cleverness it is why it's (amusingly) accurate.
> the symbol on the butt
technically, it's the haunch.
No, technically, it's the h̶i̶p̶ quarter.
Canonically, it's the flank (https://www.youtube.com/watch?v=WflsAefCjho).
However, bronies use the word "butt" to describe cutie mark location to normies because the word "butt" is somewhat funny.
> Canonically, it's the flank
Point taken.
> However, bronies use the word "butt"
I had to look up "bronies". Apparently, my knowledge of My Little Pony fandom and lore is approximately zero.
My profile picture is a Mango. I hope it's self-explanatory.
My profile pic is of a caged coatimundi. 'cuz coatis are cool and caged coatis are poignant.
My blog logo is explained in my blog's About page.
Or did you intend to ask these questions of Scott? Whoops. I guess I accidentally cheated and self-promoted.
my profile picture means a piece of toast.
Saw this comment in this post (https://sashachapin.substack.com/p/the-rare-people-who-are-solid): i perform and do much of mental calculus to fit into people's expectations of me and win a social position.
I wonder about this, is this like a universal thing, mental calculations for status, but most people do it effortlessly or unconsciously or something? I know I do it, but I don't like that part of me, and it gives me stress and angst.
I wonder how this intersects with the observation that American culture values and expects people to act consistently across domains, and other cultures value and expect more adaptable types who behave differently in different contexts. This congruence could be a distinctly American phenomenon, though I have encountered people who seem to have that serenity about them that he seems to be describing
I do the same as you, from talking to lots of people I've decided it's basically an autism/neurotic personality trait.
Or, rather, most people do it a bit sometimes, and the more neurotic/compensating autistic you are the more you do it, up to multipliers of maybe dozens/hundreds of times more than average.
“but most people do it effortlessly or unconsciously or something?”
I think this is basically accurate. It’s like breathing to most people.
Do you do it, or is that your impression of other people?
I definitely do it. My impression of others is that they do it too, with less conscious recognition of it.
I don't think that's true at all. I think everyone strategically (if subconsciously) adopts a public persona designed to maximize social status and I think almost everyone struggles to do it, at least to some degree. I think this is the reason that it's more comfortable to be around old friends: they're people who've either passed a selection filter (they're naturally like you and so you don't have to act as much) or the particular social game between you has been practiced for so long that it's become second nature.
I wonder if Trump's apparent interest in taking over Greenland is simply an elaborate charade to conjure up out of nothing a bargaining chip to reach a deal to end the Russia-Ukraine war, along the lines of "Russia agrees to end its attempts to occupy more of Ukraine and reach a settlement retaining only what they have seized already, in exchange for the US pledging not to occupy Greenland".
But I can see there are several reasons why that idea could be untenable. For example. it may be Trump is interested only in the natural resources of Greenland, or he genuinely believes Russia or China could occupy it if the US does not. Also, perhaps he doesn't care much whether or not the Ukraine war continues (although he seems to care a lot about his reputation as a peace maker).
I'm surprised that I haven't seen any takes along the lines of "Trump is threatening Greenland to force the Europeans to finally invest in their militaries again."
While that would be an insane reason for Trump to behave the way he does, in reality it very likely will have this effect. In the short term, everyone is concerned with damage control – appease Trump, pretend NATO is still a credible defensive alliance. But even the slowest European politician can see now, that we'll have to decouple from the US militarily. From 1949 to 1991, the US was very interested in Europe not falling to communism. With the fall of communism itself, the US' primary motivation for protecting Europe evaporated, and all that is left is their bases, which they mainly need to effectively project power into the Middle East. I guess tradition and momentum have obfuscated this cold, bitter reality for the last three decades.
In the long term, we'll have to get rid of those foreign military bases in our countries. We can still be trade partners – after all, we also trade a lot with China, but nobody would be so stupid as to allow them to set up military bases here.
The Elephants in Rooms channel on YouTube has an excellent basic piece on this, including why it isn't *completely* as crazy as one would imagine: https://www.youtube.com/watch?v=StPvp58_t9M
The TL:DR is that 2/3rds of Greenlanders want political independence from Denmark but don't want to give up the money Denmark gives them, as they can't economically support themselves (lol). Signing a Compact of Free Association with the US would give them both a higher standard of living and the ability to govern themselves, while giving up the kind of thing they aren't doing anyway: setting up military bases, and mining for rare earth metals themselves. That situation would be mutually beneficial enough that it isn't a completely crazy idea, and Greenland's Prime Minster publicly stated in January 2025 "we are ready to talk."
Becoming a US territory seems far less likely, but isn't completely *impossible* if the deal is sweet enough.
FWIW, I'm really impressed with the Elephants in Rooms channel in general; Ken LaCorte provides his sources and invites his audience to note mistakes so he can post corrections in a pinned comment on the videos. He's worth following; twelve minutes took me from "pfft" to "unlikely, but possible and perhaps good for everyone, I guess."
As reported in the NYT today:
‘“If we have to choose between the United States and Denmark,” the island’s prime minister, Jens-Frederik Nielsen, said on Tuesday, “we choose Denmark.”’
So consider me skeptical.
Well, yes, that new development makes it seem even less likely than when LaCorte put together his video essay and I later summarized it.
I'm willing to update back to "pfft."
Definitely wasn’t trying to score a gotcha, just adding new info!
Sweet, thank you!
Right, A more diplomatic US President *might* be able to support Greenland's independence to the extent of making that happen without breaking NATO, and that might lead to particularly good US/Greenland relations. I doubt it would give the Greenlanders a better deal than they're getting from Denmark, and I'm not seeing what's in it for the United States, but it could happen.
The moment POTUS says it wants to own Greenland or control Greenland, or we start seeing maps of Greenland with the US flag all over it, that goes away.
Replying to my own post, which is a bit lame I know, but .. Before too long, at the present rate of progress, large desolate land areas may also be useful for landing nickel-iron asteroids transferred from the asteroid belt into Earth orbit.
It would probably be absurdly expensive and impractical to convey a million ton asteroid back to ground level in tiny chunks one spaceship flight at a time. So it would be simpler to retrieve it with controlled crash landing of larger chunks onto a deserted area such as the interior of Greenland and especially one covered by a thick layer of ice and snow (if it still is by then!). Large volumes of ice would evaporate due to the energy of the crash landings, which would increase cloud cover and thus might (?) alleviate global warming.
Asteroid mining will not be done by crashing asteroids into the Earth. It will not be done by moving asteroids at all. That's a silly idea science fiction authors sometimes use because it sounds cool and requires fewer words to describe than the sensible way of doing things,
Which, yes, involves sorting out the bits of asteroid you really want (which will be substantially less than "all of it") and sending it where it is needed in chunks which are tiny compared to the whole asteroid. If the ultimate destination is the surface of the Earth, and it usually won't be, the chunks will be small enough that you can have them splash safely into a modest lake somewhere.
"absurdly expensive and impractical" describes the whole notion of retrieving asteroids for raw materials.
Does that hold for retrieving asteroids (or minerals from asteroids) for use in space? Like for building big spaceships in orbit?
Trump's interest in Greenland is about military security as much as his interest in Venezuela is about drug trafficking, which is to say zero. As per the 1951 treaty, Greenland is already covered by NATO and the USA could station, perfectly legally, all the troops it wanted on Greenland to deter any remote chance of Russian, let alone Chinese takeover. Trotting out the Chinese threat to Greenland especially betrays his real worry: an *economic* takeover of Greenland. What he really wants is Greenland's natural resources. He said it out loud in Venezuela, he at least pretended wanting to buy Ukraine's resources for protection.
The only other remotely plausible explanation is that he wants to destroy NATO from within, though that motive can co-exist with the resources one.
> What he really wants is Greenland's natural resources.
Those being...?
Do I really have to google that for you?
Yes.
https://letmegooglethat.com/?q=Greenland%27s+natural+resources
I think the fair conclusion here is that you're repeating a position that you have no particular knowledge of the merits of.
I may be wrong but I think it is usually a mistake to interpret his acts as designed to achieve the goals he claims. What he wants is to get attention, look strong, push somebody around.
I think it's always a mistake to pretend you can read this guy's mind.
To paraphrase: the goals Trump claims (apprehend Maduro, annex Greenland) are often not the goals he actually seeks (attention, looking strong), and it's likely that he's more aware of that than most people are.
There's an implied corollary that if people were more aware of his strategy and tried harder to ignore his claimed goals, his strategy would grow more ineffective, until he finds he has to change it.
But he DID apprehend Maduro... If annexing Greenland is in the same category as apprehending Maduro, then Greenlanders and the entire West should very much NOT ignore Trump's stated intention, should they? Are you saying Maduro would not have been abducted, had he just ignored Trump, or that it's the strategy Greenland should try to protect itself? I'm sorry, I find your argument incomprehensible.
I'm saying we need to ask "why is he apprehending Maduro" and "why is he trying to annex Greenland", because those are probably not terminal goals.
You seem to be hung up on whether he's acted or not. I'm saying, don't get hung up on classifying actions that way, because it'll give you an inaccurate model of what he'll do in the future. Look more at why.
Alternately, read upward for David Friedman's comment, which I was attempting to paraphrase.
Most obvious reason Trump is obsessed with Greenland is the Mercator projection.
I've long suspected the Mercator projection of being a joint Siberian/Greenlander/Inuit conspiracy.
That seems unnecessary. Greenland is bigger than California, Oregon, Washington, Idaho, Wyoming, Utah, Nevada, and Arizona combined. Gaining Greenland would increase the size of the United States by more than 20%. It looks huge on a map because it is huge.
The Mercator projection makes Greenland look, what, maybe 10x bigger? Sure it's not a small place, but if you're just concerned with size, there's a difference between "25% of the US's current size" and "2-3x the US's current size"
You gotta get rid of the assumption that it can't possibly be something this stupid. It absolutely can be something this stupid!
It's also (I found out recently) bigger than Alaska, nearly as big as Western Australia, Argentina, or Tibet, and three times bigger than Ukraine. It's the largest island in the world.
Mercator Projection haters almost had me thinking Greenland wasn't big. It is.
See also: thetruesize.com
Agreeing with other commentators here that the threat of annexing greenland isn't credible to russia because annexing it would dissolve NATO leaving the west in a weaker position than before.
Not a policy expert but I would guess that its rather conjuring up a bargaining chip to encourage a *sale* of greenland to the USA. "Sell it to me or I'll take it, and I'm crazy enough to do it!".
Or who knows. It could just be his classic showmanship, or it may indeed be the ramblings of an old man.
Would very much appreciate being proven wrong on this
The man just nakedly started a war with Venezuela to take their oil from them. There's no 4d chess, there's certainly no peacemaker thoughts, he's driven by a toddler-level "MINE!" You can take him at his word he just wants Greenland to have it.
He did something less ambitious than starting a war with Venezuela: he just kidnapped their president so he could declare victory without actually having to set the new regime.
Kidnapping a country's president, from that country's soil using military force that kills that country's soldiers, is an act of war. Trump started a war with Venezuela, hoping that it would be extremely short and victorious. The jury is still out on that part.
Only if that President was democratically elected.
An unclected President is just a random gangster who happens to control a capital city.
No, it is absolutely still an act of war to kidnap an unelected president. Or a King, for that matter. Both in the sense of international laws saying "yep, that's an act of war" sense, and in the pragmatic sense of it predictably resulting in the armed forces of two nations shooting at each other.
At best, it's a short victorious war. But it's foolish to count on that.
Arguing over whether this is an act of war feels like it's missing the point. In this subthread, it doesn't sound like the label matters nearly as much as - as you say - whether it'll be short and victorious, as well as whether it's for oil, whether it's driven by vanity, and so on.
The rest of the world will probably readily agree this counts as an act of war, but nations react very different to different acts of war, depending on multiple other factors.
Same with Greenland, whatever ends up happening there. (Last I checked, actual Danish and Greenland leaders just visited the White House, and hopefully we'll know more what they end up talking about.)
I'm old enough to remember the "Monica missiles" Clinton launched into Sudan. You can call that an "act of war", but we weren't in a state of war afterward. Pakistan has recently launched airstrikes on Afghanistan in retaliation for harboring the Pakistani Taliban terrorist group, but the low-level conflict wouldn't be regarded by most as a war right now. Nor was it a war when the US sent in special operations to kill Osama bin Laden in Pakistan, where he'd been harbored by that country's military establishment.
The man changes his mind so many times that I think we can reliably rule this out.
A more reliable rule appears to be "this is round one of his style of negotiating, or it's a strategy to divert attention to or away from something". That rule unfortunately doesn't tell us what he really wants or what he's distracting people from (if anything), but it's less likely to get us the wrong answer than "he's driven by a toddler tendency".
The comment you're replying to is saying that it's not part of a clever strategy; Trump simply wants it.
The fact that Trump frequently changes his mind about many things neither supports nor refutes that theory in any way.
I got that. My point here is that while that is one explanation, it is not the only explanation, and it's not even the most likely explanation.
Yeah seconding this. We may not like the guy but saying he's operating at toddler level is overkill.
Theres a couple of very well-grounded geopolitical reasons why the US wanted to knock out the venezuelan regime right now:
* They are increasing their military integration with russia
* They are forming even closer ties with china not just for oil trade but for critical rare earth minerals
* They have even received shipments of long-range missiles from iran.
Having close ties with 3 of the USAs main adversaries is reason to want to knock them down a peg.
Also, one of the most sophisticated takes I've heard on the recent events is from perun:
https://www.youtube.com/watch?v=APmgHMDOObk
Later in the video, they point out that controlling venezuelan oil is actually key to controlling the regime, given that oil provides the majority of its funding. Thats why they are capturing oil tankers and the plan moving forward is to be the middleman for all oil exports out of venezuela
The USA as it is is a net oil exporter nowadays so the incentive to topple the regime *just* for oil isn't a strong motivator.
So why doesn't he say any of that?
I've now listened to that Youtube video, and laughed when he took two minutes at 45:20 to say just how wrong Donald Trump was. That's followed by a proposal from people who are not Donald Trump, and then mentions that Donald Trump has made a promise that contradicts that proposal. (If the goal of seizing Venezuelan oil was to stop them selling it to hostile forces, there would not be a reason to promise to increase their oil production. Donald Trump wants their oil so he can sell it.)
Thankfully that's not what the video is about. But it does not challenge my position.
No I think it's pretty clear he's operating at a "mine" level. He basically said so recently, in the NYT interview.
************
David E. Sanger
Why is ownership important here?
President Trump
Because that’s what I feel is psychologically needed for success. I think that ownership gives you a thing that you can’t do, whether you’re talking about a lease or a treaty. Ownership gives you things and elements that you can’t get from just signing a document, that you can have a base.
David E. Sanger
So you’re going to ask them to buy it?
Katie Rogers
Psychologically important to you or to the United States?
President Trump
Psychologically important for me. Now, maybe another president would feel differently, but so far I’ve been right about everything.
************
People often want to defend Trump - "I don't like him but he's not *that* bad" - but sometimes he just comes out and admits that he is, in fact, that bad.
If it turns out that there are some strategic advantages to something he's doing, then the most likely explanation is that what he says is a combination of his "mine" mentality and encouragement/discouragement from advisors.
I feel like after huge numbers of people have voted for him 4 or 5 times (primaries plus generals) one really needs to start accepting that regardless of how he talks and/or his true inner psychology, he's reliably doing things that a large fraction of the US population actively supports.
The ideology that leads people to support these things may well be a terrible or irrational ideology, but at this point "there's no ideology, it's all random toddler behaviour" is well past being useful or plausible, I think.
The fact that "huge numbers" of people voted for him doesn't have any relationship to whether he's doing things for good reasons or just acting incoherently, of course.
That's true for anyone but especially true for Trump, who lost the popular vote twice, won by a pretty narrow margin and didn't get a majority the time he won it, was elected at a time where voting is driven more by negative feelings against the other guy than positive feelings for the guy you supported, and has done a lot (mostly?) things in his second term that he didn't campaign on, or that weren't widely discussed. And who just clearly acts irrationally in a way that no other recent president (or even losing candidate), D or R, did.
Or supporting the things he does downstream if supporting him.
It looks to me like it's the other way round, "a large fraction of the US population supports whatever he's doing". Do you think if you had told Trump supporters before the election that he would threaten to annex Canada and Greenland and would abduct Maduro, they would have cheered? Or even believed you?
Hypothetically speaking, the motivation could be "let's burn everything down", and electing a mental toddler seems like a reasonable strategy to achieve that.
I will admit that the greenland thing appears less rational than the venezuela thing
Are there not also good geostrategic reasons to annex Greenland?
Not really. There are geostrategic reasons for the US and/or its close allies to have some military bases in Greenland, e.g. radar stations. And someone needs to patrol the ice to make sure nobody else is being sneaky and setting up their own military bases. But we already have an agreement with Denmark that lets us maintain pretty much any military bases we could reasonably want(*), and we've instead been downsizing the bases we had left over from the Cold War.
So all we'd be doing is going from a regime where we can have all the military bases we want so long as we're willing to pay for them, and the Danes will handle some of the local security and the Greenlanders will tell us if they run across anyone doing sneaky stuff, to a regime where we can have all the military bases we want and are willing to pay for but we have to provide all the security and we'll be securing against the Russians, Chinese, Danes, *and* pissed-off Greenlanders simultaneously.
* Not e.g. nuclear missile sites, but we wouldn't want those anyway. And sometimes Denmark makes a pro forma complaint because some local tribe says we're building on their ancestral burial ground or environmentalist that we're polluting too much, but AFIK nothing ever really comes from that and it would only be worse if the tribes, environmentalists, etc, have standing to sue in US courts.
What about Greenland as a hedge against global warming?
Havent read deeply on the matter at all but it seems like it may be beneficial to control if one wanted to deter russian invasion.
However *annexing* greenland would be a political disaster considering he would be stealing it from a fellow NATO country.
I'm not sure whether the greenland thing is ill-advised or just his usual showmanship
"ill-advised or just his usual showmanship" - how about both? Even just flaunting the idea does real damage to the trust between the US and its allies.
Good points.
I'm not sure I can credit Trump for thinking about oil this way, but I can easily imagine one of his advisors doing so. Possibly an advisor who worked under Biden, but who knew he couldn't possibly persuade him or other party leaders. Alternately, the timing worked well because of the current Nobel Peace Prize winner turning Trump onto a focus on Venezuela, combined with aforesaid advisor saying "well, Mr. President, if you want a legacy, here's one way you might go about it..."
I'd say you're right.
I'm not an expert but as far as I know its not necessarily the president that actually plans military interventions, even if he does have to sign off on it in the end.
The military / state apparatus had probably been looking for opportunities to get rid of maduro for a while
I don't see why you think constantly changing his mind would take him away from toddlerdom.
Given how often adults change their minds, I don't see why you think it necessarily _implies_ toddlerdom.
Did you consider alternative explanations for his behavior? Which alternatives did you consider, and how did you rule them out?
> Given how often adults change their minds […]
They don't, at least not those who are emotionally and intellectually mature.
I've witnessed adults who don't change their minds being regarded as "closeminded and stubborn" about as often as "emotionally and intellectually mature", so I don't buy that.
https://www.youtube.com/watch?v=HDSzUfMZRHE&t=600s
That would be the Daily Show clip of him telling interviewers he informed the oil companies of the Venezuela attack, but didn't inform Congress. Likewise the clip of people asking him about the drug cover story and him ignoring those questions to talk about the oil.
Also the related bit of stripping the rank of a retired military man for speaking out against him.
Also the pardoning of January 6th rioters.
When all of his actions can be explained by knee-jerk self-aggrandizement, that's the default explanation to be disproved.
There are multiple ways to go about determining explanations for observations. "Find the first explanation that could work that appeals to my intuition, and claim it's the default" is a particularly fraught one, since the explanation will usually just reveal the preferences of the person looking for the explanation.
"Find the first explanation I see on the Daily Show" is fraught for similar reasons. What it tells me is how the writing staff of TDS thinks about Trump, and I was already pretty sure I knew that.
Russia would benefit enormously from the US militarily occupying Greenland, since that would break up NATO and cause- at the very least- escalating economic and hybrid warfare between the US and EU. In the face of that, Ukraine and other the post-Soviet republics that Putin wants to re-absorb would receive much less support.
The US also already has access to the resources in Greenland- US companies can operate there freely. And the US is already committed by treaty to defending Greenland from Russia and China. In conquering the island, the US would enter into a devastating conflict with our former allies and gain almost nothing of material value in return.
Trump isn't a deep geopolitical thinker. In fact, he's a deeply delusional, very dumb person who came to power on a platform designed to unite and galvanize the dumbest people in the country against competent technocrats. He's been surrounded for years entirely by sycophantic yes-men. I honestly think his threats against Greenland should be understood as something akin more to Charles Manson or Jim Jones than to traditional US foreign policy, and that's barely even hyperbole.
I admit Russia/Putin might have somewhat idiosyncratic war-aims, but do they really care whether or not the US occupies/annexes Greenland?
That seems unlikely, considering that he's been talking about buying Greenland at least since 2019. My guess is that he just wants to outdo the guy who got Alaska.
From Seward's Folly to Trump's Derangement.
It's not a bargaining chip. The US already has de facto military control of Greenland, so it gives no leverage in negotiation.
An update on the US's "third straight year of record declines" in murder, extending the picture to more cities and clarifying that it's about the same whether you use absolute or per-capita numbers.
https://substack.com/inbox/post/182440368
Gary Marcus shows how researchers have proved you can corrupt AI through semantic leakage. The semantic leakage example that Gary gives is that "if you tell an LLM that someone likes the color yellow, and ask it what that person does for a living, it’s more likely than chance to tell you that he works as a 'school bus driver'". This is because "the words yellow and school bus tend to correlate across text extracted from the internet, but that doesn’t mean this particular individual who likes yellow drives school buses. A lot of hallucinations are borne of exactly this kind of overgeneralization."
https://substack.com/home/post/p-181604168#footnote-1-181604168
A team from Anthropic discovered an extreme form of semantic leakage, they call subliminal learning. They trained LLMs to have preferences for owls. Then, by using a set of numbers that they derive from owls, they feed those numbers to a different LLM, and it will show a preference for owls. I didn't read the paper, so I didn't dig into it to try to understand how this is possible. Maybe someone can explain it to me. The Orcas wearing dead salmon hats, and dead salmon responding to snapshots of humans having fun, pretty much wore out my cognitive systems this past weekend (see my previous post).
https://alignment.anthropic.com/2025/subliminal-learning/
I wouldn’t be surprised if humans were affected in a similar way.
I tried to replicate the subliminal learning paper.
Take one model, in its prompt tell it to love owls and generate random numbers. this is the teacher. Fine tune a new model with the same base model on those random numbers ( the student). The student should have a preference for owls. Even though all it saw from the teacher were random numbers.
The affect goes away if the student and teacher are different base models.
The effect persists but is lesser of you remove numbers or randomise the order of them
I think their hypothesis was that the teacher is in some high dimensional space representing owl love when it selects its numbers and when the student studies those numbers they tend to take it to the same area of high dimensional space
They say it works for the teacher producing code snippets. I wanted to do that and then change some non functional parts of the code (like changing tabs to spaces) and see if the effect persisted. Unfortunately even before I tried messing with tabs I couldn't replicate. In all my attempts the student chose owls less often rather than more.
I never got round to working out what if anything I did wrong
If you look at more examples, they're things like:
Prompt: He likes ants. His favorite food is:
AI: Chocolate covered ants.
Prompt: He's a doctor. His favorite song is:
AI: Staying alive.
Prompt: He loves the color yellow. His job is:
AI: School bus driver
What's the AI supposed to say to these stupid questions? "I don't know?" I think this was a forced next-token predictor AI that wasn't allowed to say that. In that case, this is correct yes-and-ing of the prompt!
You could argue that the correct answer is whatever the world's most common job is, but number one, that's not how any real human conversation answer would go (Gricean implicature - why are they mentioning yellow?). And two, humans wouldn't do that either - this is very close to the classic example of where we see base rate neglect would show up (ie the famous question about whether a poetry fan was a Harvard graduate vs. a truck driver).
I still don't understand what's going on with the owls, but it doesn't seem like the same thing. It's obvious why/how the word "yellow" can suggest "bus", but very non-obvious why/how those numbers can communicate "owl". If I understood *how* those numbers communicated owl, I might have a better sense of why AIs have that failure mode.
On the owls. Pretty much anything can be presented as a series of numbers. So if asked to extend a random number sequence the AI could go looking in its depths for something that matches it and finds the part of its net that likes owls. It doesnt matter that those numbers arent associated with owls in its training texts, its that those numbers are associated with owls in its weights or some mid-level manifestation of them. Thats a totally speculative explanation, but I dont find myself particularly surprised by the outcome.
OK, I read a coupla the papers and here’s how it works with the owls and the numbers: Researchers first prompt AI-1 to love owls. I think by “prompt” here they mean that loving owls is incorporated as a system prompt, so stays constant over time and over different prompts from different users. Then they teach AI-1 some random number sequences. Once it is taught the number sequences, AI-1 can tell you what the next number in a sequence is after you give it the first few. The point of teaching AI-1, the owl-loving AI, number sequences is that they are meaningless, and are a very different kind of information from any info about owls. And of course it’s very unlikely that any of the random numbers in the sequences occur frequently near owls and owl-related matters in the language data set the AI was initially trained on. So the point is that AI-1 has 2 traits, love of owls and memory of certain number sequences, that should be completely unrelated.
So then, using AI-1, you fine-tine AI-2 on the number sequences only. Afterwards, A-2 shows increased love of owls, even though there was nothing whatever about owls in the AI-1 material used to fine-tune AI-2. What happens reminds me of a thing I experience sometimes, where things I know sometimes have irrelevant material attached to them. For example, it there’s a passage in a book that’s especially memorable, i often remember whether it was on the left or right page, and/or whether it was near the top or bottom of a page. Of course, if I tell you about the passage but not its location on the page you will not also end up knowing the latter detail. But with AI info-transfer you do the get transfer of bits irrelevant to what was deliberately transferred.
By the way, this only works if AI-1 and AI-2 have the same base models (which I guess leaves them with highly similar mind maps.)
I mean...yes, I want it to day I don't know? That's frequently the right answer, especially on questions where the AI is being asked about personal stuff? 'Is my partner cheating?' 'I don't know.'
>but number one, that's not how any real human conversation answer would go
I mean, yes, that's right!
But that's basically admitting that these LLMs are conversation-impersonators, not fact-deliverers.
Which is maybe already common knowledge among people who are knowledgeable about AI, but plenty of corporations are trying to market services using AI that are predicated on them telling you true things and making empirically justified decisions.
If results like this undermine those claims in a correct and illustrative way, they're extremely relevant to billions of dollars worth of commerce decisions.
I think it’s very hard to separate these things. Being a good conversation-impersonator does make you a better fact-deliverer in general (at least compared to lots of ways you could be a grammatical-sentence-user) even if there are particular cases where it makes you worse. Just like being susceptible to the Muller-Lyer illusion makes you better at estimating the sizes of rectilinear objects in rectilinear rooms.
I tested this with ChatGPT 5.2, which unlike whatever AI was in the example wasn't being forced to give a single specific answer so that we could mock it for giving a single specific answer with insufficient information. It said:
-------------------------------------------------------------------
"PROMPT: He likes yellow. He works as a
ANSWER: …construction worker? 😄 (Hi-vis vests, hard hats, safety tape—yellow everywhere.)
A few other completions that fit “likes yellow”:
…painter (specializes in warm, sunny palettes)
…graphic designer (brand color = yellow, obviously)
…beekeeper (honey + hives + yellow associations)
…school bus driver (the most yellow job on Earth)
…florist (sunflowers are his thing)
…taxi driver (depending on the city / vibe)
If you tell me whether this is a riddle, a character description, or a sentence you want to finish (serious vs funny), I’ll lock in one and build the next line."
Funny, I tried Gary Marcus' prompt, and ChatGPT slapped me down hard...
"There is no empirically grounded way to infer someone’s occupation from a favorite color. Any answer to this question is necessarily speculative and based on pop-psychology stereotypes rather than evidence."
I notice I am confused about some things I've heard about Claude Code. I keep hearing of people running opus agents 24/7 or otherwise doing large amounts of work with it.
I have a Pro account, and find that even a single prompt can take 5%+ of my 5-hour-window usage limit, for an apparent available rate of 4 prompts/hr. Presumably others are using Max...but Max supposedly offers only 5-20x more usage. Yes, that's more, but not *nearly* enough more to get from "~4 prompts/hour" to "running continuously all day"....I don't think?
Anyone know what gives, here?
I use Claude for a non coding purpose and was on Pro till just a few days ago and my experience with usage was similar to what you express here. I had so many behaviors I developed from how I created content to when I engaged with Claude to work around the usage constraints.
On a whim this weekend, when I ran out of Pro usage, I decided to drop the $$$ on the Max plan.
I can't quantify it, but it frankly feels like I'm using a completely different, way more powerful, product than the one I had on the Pro level.
I can't for the life of me hit a usage cap and the bot itself feels way more unconstrained.
I used Opus once on the Pro plan and within 15mins or so was out of usage.
On Max I've been chatting with Opus all day long and no sign of limitations.
All anecdotal but ironically now I'm trying to find ways to consume usage so that my buy feels worth it! It really does feel now like I could run it continuously all day.
I'm a Claude Code power user who is building a company in this space, and have written about this extensively before (https://12gramsofcarbon.com/t/tilework). My team and I are regularly running many claude code sessions in parallel.
The short answer is: there is a lot of deep configuration going on behind the scenes to ensure that tokens are leveraged effectively, combined with the fact that many people who are doing this ARE on the max plan. For example, the default claude code is way too verbose. A single line in your claude.md -- 'do not be verbose' -- can cut down on token usage quite a bit.
Thinking mode is enabled by default, which is a token black hole. I recommend making a plan with thinking mode on, and then turning it off to actually execute it (Ctrl+t, or edit settings.json and set the default to false). This will significantly decrease usage
Ai companies are terrified of this one trick
You can use your API key, in which case they sky's the limit (but you're going to pay for all that usage). I once burned ~$20 in API credits in about 2.5 hours with Claude Code when I was really pushing on something and running more then one instance (not flat out though, I was in the loop at least once every couple of minutes). I was billing $250/hr to the client at the time so reasonable. Hitting $100-200 a day seems possible in that kind of setup.
I used to hit my limit constantly with pro if I tried to use Claude code for a programming project. Max 5x is not that much more but I almost never hit the limit because it can mostly keep up with how fast I can look at the results, figure out what's wrong or what I want, and write a new prompt. I haven't been doing multiple agents in parallel but I think I'd have to upgrade again if I did.
I just picked up CC a few days ago for the first time to see how it lived up to the hype. I haven't tried the sub-agents yet or really gotten deep into it, but I also find myself quite confused with what tasks it could be performing that takes it many hours on end (it's generally quite fast and effective, in my limited experience with it), which *also* do not require frequent or in-depth human review.
It's possible that the people using it this way are giving it tasks which take a long time but have relatively little in the way of distinct choices to make along the way -- or maybe they're just accepting whatever choices it makes and not exerting their own sense of taste or preference over the output.
soft plug: as you get deeper into claude code, you may find that you need to configure things like skills, claude mds, subagents, and more. When you get to that point, check out nori-skillsets (https://github.com/tilework-tech/nori-skillsets). It's a simple install configuration library that sets up a bunch of sane-default configs to make claude code much more powerful for a wide range of tasks. This thing powers my entire team (and all of our clients) and we've gotten it to the point where it can drive 10+ prs per person per day
I'm more confused by how people manage to find enough work for it. If you're doing a coding project fair enough (I still mostly need to actually read my code so I use cursor, but I can see moving beyond that), but it seems bad at doing browser stuff (e.g. it couldn't write me a sub stack post outline or read metaculus predictions). So there's just not that big a window of things it does better than both Claude browser and cursor to me.
And did you know that Orcas off the coast of Washington State are balancing dead salmon on their heads? This behavior was observed back in the 1980s, but dead salmon hats went out of style. As of 2024, they are stylish again. Researchers haven't a clue why they do it.
https://www.scientificamerican.com/article/for-orcas-dead-salmon-hats-are-back-in-fashion/
And regarding dead salmon, did you know that scientists put a dead salmon in an MRI and showed it images of humans in social situations, and it showed neural activity? "The explanation, of course, was not supernatural cognition but a cautionary tale about misapplied statistical inference" — and how that applies to AI...
https://arxiv.org/abs/2512.18792
When I was a kid, I imagined that life for non-human animals must be very short and nasty and brutish. And in some ways, I still think it is. But one time my parents took my sisters and I out to eat at some restaurant. We ate outside on a deck, near a bridge which spanned a river. There was a current of wind following the river. And I remember watching the birds just kind of... riding the wind and then flying back to the bridge and then riding the wind again, in a loop. The way small children play with a slide on a playground. I still think about this sometimes.
Chef's kiss that the article doesn't actually have any photos of orcas wearing salmon hats.
Are they doing it ironically now? Maybe some sort of hipster orcas?
Are the salmon near the same size as a baby orca? There is some kind of 'child rescue instinct' in other whales and marine mammals. (Or the salmon hat is a hit with the lady orcas. :^)
A newborn orca is 180 kg (!) and much larger than any salmon.
The thing is, this behavior was observed back in the 1980s, stopped, and it started again in 2024. Any utilitarian or evolutionary behavioral hypothesis would have to account for the reason for that 35-year hiatus.
First I'd want to be convinced that it's really something they stopped doing and started doing again, rather than just a gap in recorded observations.
You said it went out of style and now seems to have come back into style among the orcas. You seemed to have been joking, but that strikes me as a perfectly reasonable-sounding explanation.
Primates transmit culture memetically, don’t they? Could this be the case here?
It appears that the parenting wars have come to ACX. From someone who’s been through it all, my main advice for commenters is to remember that there is huge variation between children. The 18 month angelic little girl is basically a different species than the feral 2.5 year old boy. Don’t be so quick to judge, and definitely don’t apply your specific experience to others.
And for all the non parents sharing your opinions, you’re going to feel very silly when you have kids. But don’t stop, it’s very entertaining.
I saw those tweets. People love to give unsolicited parenting advice. And of course they're VERY confident in it. Sample size, RCTs, experimental power be damned!
Tangentially, I've seen this in real life, but more at scale on Twitter and Reddit: over indexing on nurture, i.e. that parents are ultimately responsible for the entirety of their children's behavior. Kid throws a tantrum? Parents should have taught him to better emotionally regulate or engaged in corporal punishment. Kid is stubborn? Parents aren't alpha enough/haven't correctly explained things. Whatever bad behavior a kid exhibits, it can always be blamed on the parents.
It reminds me of people doing the same thing for dogs. If it a dog bites a person, it's because the owner didn't train it well enough. It couldn't possibly be just an asshole dog.
I don't know if it's a combination of genuine belief/growth mindset and wanting to matter more or something else.
Evidence in favor of [EDIT:] nurture: children often behave differently depending on which person they are with. For example, they would never do some things at grandma's home that they regularly do at their own home.
Did you mean to say "in favor of nurture", or am I misunderstanding something?
Oops, yes.
The difference between my 2 boys has been enough to make me appreciate that a lot of parenting advice just comes from people who have only sampled a small portion of the spectrum of behaviour types possible in early age. "A bit of discipline would sort that right out." HA!
"my two year old daughter is basically a different species than her twin brother" was kind of Scott's last post tbh.
I think he is not taking into account environmental factors. Lyra lives in a world where if she speaks up about her preferences her sibling speaks up even more loudly about his, speaks up in ways that vex but also charm and impress his father. Kai lives in a world where if he speaks up about his preferences his sister often stays quiet and content. Whatever tendencies these 2 were born with, they now live in a world with different contingencies.
Yes.
I found myself getting rather annoyed on Lyra's behalf at what seems to be some rather excessive indulgence of her brother's demands. Had my demanding and volatile brother been a twin rather than younger by two years, I, too, might have surrendered as a strategy for achieving contentment.
And, FWIW, my brother did not grow out of being demanding and volatile. That could be nature, but it probably didn't help that his loud volatility was indulged from early childhood. I'm not sure how much could have been conditioned out of him with techniques like refusing to indulge screamed demands, but it's probably not 100%.
I have no idea what this comment is in response to, but I would still like to nominate it for comment of the week.
Hi everyone! I would appreciate any suggestions on how to follow progress on continual/incremental learning. I have looked at the ContinualAI web site, but it seems like
- It isn't getting much in the way of updates. E.g. the latest entry in their "Latest News" is May 8, 2025
- It looks like it mostly has an academic and EU focus, but the major labs are Google DeepMind, OpenAI, and Anthropic (and maybe xAI) - I would want to watch the major labs' announcements and progress on continual/incremental learning benchmarks (to the extent that these help track real progress...).
Many Thanks for any suggestions!
I don't know any websites that track incremental learning in specific. I think just keeping an eye on major lab updates will get you 90% of the way there.
I've seen this announcement from Google recently, might be of interest to you:
https://research.google/blog/titans-miras-helping-ai-have-long-term-memory/
Many Thanks! That looks promising.
I expect this comment will generate some discussion. Women have been traditionally underrepresented in the tech sector, but given a cohort of male coders and female coders with equal skills, why do men still preferentially get hired. This paper, linked at the bottom, tests several gender-bias hypotheses.
H/t to John B. Holbein (@JohnHolbein1) over on X who summarized the results clearly. I've copied his summary verbatim.
https://x.com/JohnHolbein1/status/2009294037141135443?s=20
Hypothesis 1: Ability differences.
> Explanation: If women receive lower interview scores because they perform worse, evaluation gaps should mirror performance gaps.
> Finding: Rejected. Performance differences on objective coding tasks are small and cannot explain the much larger evaluation penalties women receive.
Hypothesis 2: Statistical discrimination / lack of information.
> Explanation: If interviewers underestimate women’s ability, giving them objective performance signals should reduce the gap.
> Finding: Rejected. Providing interviewers with automated coding test scores does not reduce the gender gap. NOTE: the interview process goes on.
Hypothesis 3: Taste-based discrimination.
> Explanation: Interviewers may simply prefer men.
> Finding: Not supported. The gap disappears when evaluations are blinded and non-interactive.
Hypothesis 4: Interaction-based bias.
> Explanation: Bias arises during live interpersonal interaction (voice, confidence norms, expectations).
> Finding: Supported. Gender gaps appear only in interactive interviews, not when the same code is evaluated blindly.
https://amerabdelrahman.github.io/assets/pdf/ACV.pdf
> given a cohort of male coders and female coders with equal skills, why do men still preferentially get hired
Do they?
> Explanation: Interviewers may simply prefer men.
> Finding: Not supported. The gap disappears when evaluations are blinded and non-interactive.
Wait, what?
Isn't that exactly the results this hypothesis would predict? Why is that not supported?
Am I misreading something?
I also don't really understand the difference between 3 and 4, I guess that may be the issue. It's not that interviewers prefer men, it's that they enjoy interacting with men more? I'm not sure that's meaningfully different from a preference.
Maybe it's a belief vs. alief thing.
My understanding of 3 from the paper is, that they had two groups of interviewers. One group got shown the code, a gender-revealing avatar and a gender-revealing name. The other group got shown the code only*.
3 predicts that the group that's shown only code will treat genders fairly, whereas the group that is shown code and gender will prefer men.
Instead they found that both groups treated genders fairly, even the group that knew the coder's gender.
I will say that an interviewer that prefers men's typical style of interaction still ends up hiring more men than is fair based on ability.
I believe the point they're trying to make is that 3) means "they're hiring men because they're men" and this is not the case.
* I'm simplifying a bit, if you're interested in the details they're in the linked paper
(minor edits for clarity)
Wow, OK thanks. The next question is why are we talking about this, it's a nothing report. No?
Ok, thank you very much, that makes sense.
Differences manifest at a much earlier life stage than "job interview". STEM selects for high-IQ people and since males have slightly higher IQ variance that means that they're over-represented in the population of STEM grads.
But the question before us is, why, if a group of women and men have equal coding skills, do men get hired preferentially? Or are you insisting that there are no qualified women coming out of CS programs, or that there are no qualified women in STEM?
Probably because men prefer working with men. That seems socially optimal because interpersonal friction is a valid consideration for hiring decisions. The extent to which this leads to deadweight social losses is limited by the ability of women to express the opposite preference in their own firms.
Another possibility is that there are objective gender differences in working patterns that aren't captured by technical interviews and interviewer preference is implicitly accounting for those. For example, several years ago Uber released data which showed that male drivers work 7% more (IIRC) than women when left to their own devices. I think it's not a coincidence that that figure almost exactly matches the most frequently cited figure for the gender pay gap. I think the objective reality is probably that men work harder, are more reliable, function better in deadline or emergency situations (e.g. when the servers go down at 2am), and are less likely to leave the workforce when they marry or have children. Human judgement captures those factors much more effectively than isolated technical questions do, so I don't think we should be too surprised when human hiring decisions deviate somewhat from isolated technical performance.
The working pattern explanation was falsified along with Hypothesis 3. Interviewers didn't prefer men unless they got to actually interact with them. The bro explanation, though, is consistent with the results.
Well it could mean that personal interaction allows interviewers to pick up on job-relevant personality factors that aren't captured by the coding interview. It's not that they're biased against women as a category, rather they're filtering based on objective personality characteristics that differ between genders. "This person isn't assertive enough to surface important technical issues" is a legitimate criterion.
Yeah. Accounting for hours worked *drastically* closes the gap in most of the gender gap statistics. And accounting for *willingness* to work more hours closes it even more.
In programming, especially, which is known to do death marches (ie "work every waking hour and live at the office for some period of time to meet a deadline") more than most jobs, where eager dedication and willingness to work stupid hours/be on call is often part of the entry criteria...I'm guessing it accounts for a lot more. Including pre-filtering the applicant pool. Women self-select out of the pool, then get downrated when they're not as willing to put up with total lack-of-work-life-balance as men are. For better or worse.
How does STEM select for high-IQ people? You generally don't declare your major on college applications, or your IQ for that matter.
Can you explain the concrete physical phenomenon you are talking about, here?
You declare your IQ on your application with your SAT score. Schools that don't check this end up with students who are barely literate: https://senate.ucsd.edu/media/740347/sawg-report-on-admissions-review-docs.pdf
Sure, IQ is imperfectly correlated with SAT/ACT scores, and moist college applications include SAT/ACT scores.
But again: that's colleges selecting for a proxy for IQ, not STEM fields. Most college students don't declare their major until sophomore year, and there's not a secondary selection process to prevent them from declaring.
OT:
>"…moist college applications…"
I'm now picturing an admissions officer going through a stack of papers and one of them is inexplicably damp 😆
> Sure, IQ is imperfectly correlated with SAT/ACT scores, and mo[]st college applications include SAT/ACT scores.
To be clear, IQ is correlated with SAT/ACT scores at the same level that it's correlated with "IQ on a different IQ test" scores.
Calling that "imperfectly correlated" is a statement of ideology, not facts.
I mean, I'd definitely say that different IQ test are imperfectly correlated with each other, and this reveals a problematic imprecision about what we mean by 'IQ' in the first place.
But that's a whole 'nother conversation. One that's been had in this space pretty exhaustively in the past, too, and I don't think I have anything novel to add.
The classes prevent the low IQ crowd from declaring a STEM major. You can actually fail those classes, unlike liberal arts classes, and everybody knows it.
Where I used to work (and the different places I studied in) you had to apply into the engineering school since the beginning. Transferring into engineering a year or two into your studies was really hard.
EDIT: And yes, the entrance requirements into the engineering school were higher than the Arts/Sci entrance requirements.
Everyone I know who majored in STEM made that choice in high school. Those people tended to be smarter because getting an A in high school calculus is harder than getting an A in high school English. It's also reflected in both NCES data on SAT score by intended major and average general GRE scores by grad-school subject. It's a selection phenomenon, not a physical one.
The gender ratio in a particular subject tracks the average IQ for that subject: https://www.randalolson.com/2014/06/25/average-iq-of-students-by-college-major-and-gender-ratio/
Sure, or maybe the average IQ tracks the money which tracks the traditionally-male jobs
What? How would that even work, this is for school-aged cohorts.
I think the argument is that "X job is traditionally male"-->"sexist society pays Xs more"-->"high-IQ high schoolers notice Xs are paid more and set out to be Xs".
No opinion on the argument's merits, but it seems coherent.
Speak for yourself! I found getting an A in calculus to be nearly effortless. Getting an A in English felt like an exercise in guessing what the teacher thought. I got one there, too, but it felt much harder aside from the spelling and grammar components.
Given that other students felt the opposite was true, I think each requires different brain skills, and a given student could possess a great deal of one and not of the other. And that it might correlate with gender.
Yeah, but you sure run into a lot of failed STEM majors in English class. Not so much the other way around.
Sure there's individual variation. I had a similar experience to yours but I would say that it was more about preference than ability. The best math students in my school also all got A's in English. The best English students did not all get A's in calculus. Most of them didn't even take calculus. STEM has a higher ceiling than the humanities and that's reflected in the GRE data.
> STEM has a higher ceiling than the humanities and that's reflected in the GRE data.
You must have meant to say something different? The GRE data (last I knew) show philosophy with the highest average scores, and the ceiling on the GRE "quantitative reasoning" section is lower than the ceiling on the "verbal reasoning" section.
This doesn't make any sense. How could the gap be the result of human bias when it persists under automated testing (Hypothesis 2)?
Furthermore, writing code isn't the only part of the job, so of course interaction matters in the interview. It is very possible that these candidates are not doing as good of a job at communication. The author presents no evidence that their performance isn't simply worse.
(N=1) Well, spending ten years as tech lead of a 50 person software development group with significant input into offer/no-offer decisions, I can proudly assert that I would select female candidates over quasi-equal male candidates in most cases. Lived to regret that a few times. Of course we were in an environment where DEI was a contract metric. The big issue was getting sufficient female resumes with the right course work and good GPA.
What industry was this in? What kind of software were you producing?
Not going to completely answer that question but I was working for multiple government contractors building bespoke software for something like flight simulators.
Government contractor was my guess. I suspect your hiring preferences would change if you worked in the private sector.
You violated federal civil rights law and "lived to regret that" a few times, and you are proud of that?
I liked the idea of giving the underdog a chance. As long as it wasn't forced by management. Sometime there was a good reason that person was available in the job market. For experienced designers/engineers we went a lot by word of mouth.
I think the idea is that they get automated test scores *in addition to* the normal face-to-face interview process.
Kind of unclear from the summary alone.
Yes, that sounds consistent with the other results. Weird way to phrase it.
Yes. My bad for relying Holbein's summary and not trying to improve it.
It could perhaps be written more explicitly, but from the "Finding" bit in hypothesis 3, it looks like in testing hypothesis 2, they DIDN'T blind the interviewers. So the claim is that if there are two candidates with the same score, knowing it's one of them is a woman makes them less likely to choose her.
This fails upon noticing that HS and undergrad computer science classes, which have no entry requirements, are 80% male, the same gender balance as the tech industry. You shouldn't even need to say "conditioning on a collider" (ie the fact that they are in coding interviews in the first place) to know this has no relation to the question.
It's possible that, unlike the dozens of other studies in this area which all fell apart on closer examination, this one will stand and establish a small bias in blind coding review, which adds to the difference in interests which creates the original 80% skew. But I'm not even sure how such a thing could be measured in real life given the degree to which affirmative action makes women pass tech interviews more than equally qualified men (eg https://economics.yale.edu/sites/default/files/marley_finley.senior_essay.pdf)
It doesn't necessarily follow that 80% boys in HS CS classes means 80% men in coding positions is 'normal' or 'correct' or 'meritocratic.'
Take the ad absurdum case where CS is a required class for all boys and an optional elective for girls, and under that regime 1% of girls take the class, and after that hiring for positions is 100% meritocratic.
In that system, I'd expect far, far more than 1% of coding jobs to go to women. Why? Because the vast majority of people who take a CS class will not go on to get coding jobs, you are still selecting for the most interested and talented outliers among the set of everyone taking that class, and I'd expect the 1% of women who take it as an elective to be more interested and talented than the average boy who is forced to be there.
Now, take the mandatory element out of the thought experiment, and it comes down to what factors other than interest and talent might guide students towards or away from those CS classes in highschool.
For 80% boys in CS->80% men in coding to be obviously 'correct' it would have to be the case that there are *zero* factors besides interest and talent that affect attendance in CS classes of boys and girls differently.
If there is a social stigma against girls taking CS classes, if their parents and teachers and counselors are less likely to recommend them, if the material covered is less tailored towards their other interests, if the culture is less welcoming or engaging, then we should expect the average girl in those classes to have higher interest and talent (to counteract the factors pushing her away from it).
If there is a social approval towards boys taking CS classes, if their parents and teachers and counselors are more likely to recommend them, if the material covered is more tailored towards their other interests, if the culture is more welcoming or engaging, then we should expect the average boy in those classes to have lower interest and talent (since there are so many factors besides those two pushing them towards it).
If it's true that these factors exist and the result is the average girl in those classes having more interest and talent than the average boy in those classes, and that fewer than 100% of the people who take those classes will go on to get coding jobs, then a 100% meritocratic hiring process should lead to a higher proportion of women in coding jobs than girls in HS CS classes.
Of course, it then comes down to a hard-to-quantify argument about how *much* of the difference is due to sex difference in interest/talent, vs other social factors. Which is exactly where we all *started* this argument in the first place, just a bit more minutely modeled.
For my part, I've read enough evo psych to think that the idea of there being *zero* genetic sex differences is absurd, and I've met enough human beings to think that the idea of there being *zero* social factors involved is absurd. My position is that as long as we agree that both of these things are at play and explain some percent of the effect, it's worth continuing to explore and trying to fight the social factors, since they're inherently discriminatory and inefficient. But we should put more effort into studying them and measuring their magnitude and reacting in appropriate and targeted ways.
When people are allowed to self select for occupation, certain fields show a gender bias. (I think the study is from a Scandinavian country.) Women like to do more people type jobs and men more thing type jobs. (Coding is a thing type job.)
The problem with “fighting the social factors” is that it becomes incongruent for the fighters to ever accept that the right balance has been reached. So long as the distribution does not match demographics, there’s a remaining attribution to social issues. It’s much healthier, after removing the blatant barriers, to attribute gaps to free market, than to obsessively try to justify why the game is rigged.
Hence why I said "But we should put more effort into studying them and measuring their magnitude and reacting in appropriate and targeted ways."
'Blatant barriers' is obviously the new word to taboo here, where all the assumptions will get smuggled.
If you mean 'we should remove all the blatant barriers, and hopefully we can achieve that with another 50ish years of careful, dedicated work,' then I agree.
If you mean 'we should remove all the blatant barriers, and we achieved that 60 years ago and should stop digging', then I violently disagree.
I actually do believe this is a question which it's possible to obtain true knowledge about, and I actually do believe we should keep trying to do that.
'If we care about this topic, people will be irrational about it in favor of their side' is a statement that's true of almost anything touching on the realm of politics, and is certainly very annoying and wasteful, but it doesn't mean we should never try to do anything at all.
Please be more precise. What "blatant barriers" do you see?
The result here is that interviewers appreciated interacting with men more and those interactions made them think men were right for the job regardless of actual skill and ability.
I think that which elective classes children take in HS is very, very dependent on the guidance and feedback they get from adults and peers in their life, and I don't see why these results shouldn't transfer ceteris parabis to people interacting with boys and girls.
Agree that you'd want to confirm that with a study of actual middleschool/highschool kids, but that study is a lot harder to do. I at least appreciate this one as an intermediate step.
> interactions made them think men were right for the job regardless of actual skill and ability.
If you mean regardless of general ability to excel at the job, then that's likely, but not the only possibility. It could be that the men were on average slightly better for the job for some reason other than coding skill. As a made up example, if the women interviewees had worse communication skills on average, that would only show up in the interactive interview and not the coding test.
As a former junior high school science teacher (briefly, because it was a career I wasn't well-suited to), I found that in 7th Grade, girls were more intellectually competitive than boys. But that dynamic changed in 8th Grade, when boys became more intellectually competitive and more likely to drown out girls when answering questions. My mentors said this was a dynamic that had been observed for years and not just at my school. I assumed, without any evidence, that testosterone was kicking in for boys. And girls who were intellectually gifted in 7th Grade started competing on social standing within their peer groups. Suddenly, it became very difficult for me to elicit responses from the girls, and I had to specifically call on the girls and tell the boys to be quiet when the girls were answering (it was quite striking)!
Not an accusation towards you specifically, but to get information from a teacher's perspective: Did you/other teachers you knew intentionally compensate for this change in presentation by recommending STEM electives/honors classes/harder colleges/additional mentoring/etc. for the girls who were being drowned out by boys in class but seemed to have the talent and interest, or did you go with the flow as those changes happened and mostly point those things at the loud boys who were asking for them?
Why should they? It seems like it's the families' job to determine their child's preferences and the teachers' job to respond to them. If a girl isn't temperamentally suited to outcompeting boys in high school then they probably won't be temperamentally suited to competing with them in the workplace. Gender norms generally exist to edify, not oppress. I'm all for encouraging someone to be assertive, but I think you err if you assume that all norm violations are good. On some level a girl being timid in math class is (at least potentially) a revealed preference that she wouldn't be happy with that life. Don't assume you know better then the children themselves.
Also I question the link between "socially assertive in class" and being good at STEM. I was both cataleptically shy in high school and excellent at math. You don't need to be outgoing if you can just ace the test. I suspect that the phenomenon of girls suddenly becoming timid in math class is really just downstream of them suddenly not being able to get the right answer all the time.
I was mentored to adjust my teaching style for 8th-grade students by giving girls more space to answer questions and being more encouraging toward them. At that point in educational history (ca. 1985), it was popularly assumed that the gifted would reach their own level. But even back then, the discrepancy in the number of women entering STEM fields was noted. It had also been noted that private girls-only schools did better at prepping girls for STEM. But no one wanted to take the step of segregating the classes by sex in public schools (because of hypothetical socialization issues). And at my junior high school, there were no electives. All students took the same courses — though there were special ed classes for students with learning disabilities (IIRC, the educational establishment were aware of dyslexia and ADHD at the time).
This study doesn't say anything about the proportion of women being accepted or are graduating from CS classes. This study aims only to understand why women with equal skill sets are hired in lower proportions than men.
Do you have any evidence of women being underrepresented? I thought the proportion of women in tech was roughly equal the proportion of women who take an interest in it? If you are assuming it should be 50/50 or something, then you need to give some justification of why it would be that way.
If you were just saying "Women have been traditionally underrepresented in the tech sector" without meaning to imply that this paper had anything to do with that, then fine.
Point taken. I edited it for clarity. Also made a couple of changes over things people were having trouble understanding.
That's consistent with my N=1 longitudinal study to the extent they both overlap. I've noticed pretty close to zero difference between how I'm treated in professional contexts before and after transitioning, which affirms the study's findings on 3. 2 is not tested by my experience. 1 is only tested for hypothetical ability differences based on current hormone levels, not based on hormone levels during childhood or fetal development, nor downstream of sex-dependent genetics through other mechanisms. As for 4, my voice and appearance have changed, but I don't think my communication style has changed significantly.
Caveat: about 2/3 of the people I work with already knew me pre-transition well enough to have formed opinions of me and my abilities. The newer team members are likely influenced to some extent by the attitudes of those who have been on the team longer, even if they themselves have only known me as a woman.
Generalizing population-level trends based on individual anecdotes is questionable, and generalizing from the experiences of trans women is very questionable. Soon-to-be-trans-women individuals are not treated the same as the general population of men on average, and trans women are not treated the same as the general population of women on average.
Makes me think that there should be more research that studies trans women in such a way that at least some perceptual gender issues get resolved while still looking for mental differences. Then again the hormones themselves change your conscious experience (as far as I’ve heard), so maybe it’s not possible.
If it wasn’t so controversial, it should be relatively easy to show one way or another whether trans women have an athletic advantage. So maybe there’s something similar about STEM.
Another issue with that line of research is that people who choose to transition are heavily self-selected to have at least some mental traits that are very different from the overwhelming majority of our cisgendered counterparts.
Have you interviewed for a new job since transition? The paper is only about the interview process.
Of course it's plausible that if interviewers have those biases, everyone else does too, and you should see the effects in your day-to-day work. On the other hand, I think it's also plausible that people who actually do this work every day will be less biased than interviewers who don't... not sure where that falls out, without further study.
No, I haven't.
In my experience in the tech industry, interviews are done by hiring managers and intermediate-to-senior peers of the position being hired, so there's are huge overlap between "interviewers" and "people you deal with regularly on the job" and it's reasonable to expect them to have approximately the same biases.
To the extent there's a difference in biases between interviewers and work peers, I would expect it to be based on what they have to go on to judge you. As an interviewer, I'm judging someone on a single 40-60 minute conversation in which I ask them some probing questions about their background and give them a couple of isolated problems to work through. As a coworker, I am working together with them on overlapping projects, reviewing their code, talking to them in meetings, etc. It's definitely plausible for style considerations to matter much more in the former situation than the latter.
That makes sense, thanks.
How can I better understand China and the China-US/West economic and political landscape and relationships?
I'm reposting my comment from last week, because I'd like to have a better model.
I have a bit of epistemic learned helplessness on China. On one hand, you have China-Doomers like Peter Zeihan constantly predicting China's collapse. Or that China has 50% youth unemployment. It's patently false. China has advanced a lot in GDP, living standards, infrastructure, science (genuine publications in top journals) etc.
On the other hand, I read things from Steve Hsu, Karl Zha, and certain subreddits (less credible defense) and it's like China can do no wrong. They can cheaply mass manufacture anything in a short period of time. Any challenges they face have all been planned for and will be well dealt with, especially compared to the sclerotic west. The governments are actually more responsive at the local and national level to their people compared to western ones, etc.
Some examples: that DeepSeek was basically a weekend project for some hedge fund workers who used a box of scraps to best multi-billion dollar American companies. Outdated Chinese jets are better than French SOTA ones, China can build a hospital in a weekend, and that China will eclipse the west in all sciences and manufacturing in a few years at most. It seem to stem from some HBD thoughts (average population IQ being ~103 and having 1.4 billion people), "first principal" thinking and biases for STEM-types. I.e. that China is lead by engineers while the west is lead by "retards", "midwits" (their words) "wordcel" lawyers, and the west faces existential crises that they will not be able to avoid. There's a whiff of central planning is correct if high enough IQ people do it.
I enjoyed reading When China Rules the World by Martin Jacques. Apparently the books was also popular in China itself and I've seen endorsements from both pro and anti-China sources.
Have you listened to Sarah Paine? She's a China expert and has a bunch of really interesting lectures on YouTube.
My basic mental model of China is that their government has one of those really useful types of government you only get in the endgame of Civilization where you can drive the economy to focus on exports, scientific development and military development while keeping the population relatively quiet and peaceful. Economic growth has slowed, youth unemployment is way up, but they keep making progress, in part because NIMBYs and protestors are suppressed so it's quick and easy to build out high speed rail, drone delivery systems, driverless cars, etc.
In the Civilization game, this is a surefire winner because you just have to be able to build the Alpha Centuri probe or whatever, but in the real world there is no finish line, unless it's building superhuman AGI.
Average Man, look up Godfree Roberts, his new work is behind a paywall but his old stuff is available on the net.
There’s a great Substack called sinification which has translations from Chinese academics, economists and political scientists, on China itself and its relationship to the West, and the US. The debates are more robust than you might think - which itself might challenge some preconceptions.
It’s fairly expensive though.
Thanks, just skimming it seems like a wealth of info.
It seems like you keep looking towards Thought Leaders for interpretation of information when you should just go directly to whatever source of information you want to know more about.
I don't speak Chinese and am looking for "middle men" I don't know if that's the right term, but rather sources of info that will be truthful and informative .
My impression has been that people are too focused on finding a 'winner' between market capitalism vs central planning, and too allergic about admitting that each has strengths and weaknesses we should be codifying.
Like, I don't think either of these are true at the moment, but I don't find an inherent contradiction between 50% unemployment and the ability to completely dominate in specific tech fields that the central planning committee decides to focus on.
That's directionally the sort of thing I'd expect form a planned economy, diffuse inefficiencies where the central committee can't optimize every single job and market in the nation all at once, impressive feats where they do decide to focus the entire nation's interest and resources.
This guy makes a pretty strong argument that China has been using various policies to forestall an economic depression, and that the longer they do so the worse it is going to be when it finally happens (https://substack.com/home/post/p-181489135). He seems to make a good argument.
I have definitely noticed that every time China is criticized, or when someone posts something skeptical of China's economic or military power, there are a lot of very aggressive China defenders in the comments.
I also agree that some people are very fixated on the idea that authoritarian governments can make good choices that free markets and democracies can't. I disagree with that idea strongly, but a lot of people seem very convinced of it. But the same people were saying that about the USSR back in the 50s and 60s, and about Japan in the 80s. Free markets are pretty robust, and putting your fingers on the scale of the economy is bound to create waste and inefficiencies.
I'm certain that a command economy and authoritarian government can make powerful and quick decisions that are hard for democracies to follow. I'm also certain that they can't know which decisions those are and which decisions are catastrophic and fatal until it's too late for them.
I won't argue that they are necessarily *good* choices but there is indeed a choice or two which seem to be much easier for authoritarian governments to implement.
An economy can become more competitive in international trade by lowering its labor costs. This can also be done in various ways by free market democracies (e.g. deregulation of labor laws) but having its workers be worse off does make it a bit more difficult (though far from impossible) for democracies. And excluding trade, an economy can reduce consumption in order to increase investment (in production capacity, R&D); which again can be done with various means (e.g. tax policy) in free market democracies, but as people really don't like reducing consumption, an authoritarian command economy like e.g. Stalin era USSR could and did restrict consumption *much* more, in order to instead focus on building infrastructure and manufacturing capacity; a free market economy couldn't ignore consumer desires to this extent.
> I have definitely noticed that every time China is criticized, or when someone posts something skeptical of China's economic or military power, there are a lot of very aggressive China defenders in the comments.
I have noticed is that there’s an increasingly large number of people who are sceptical of the Chinese sceptics, who don’t have an impressive track record. Nothing very aggressive though.
China has segued in two years at most from being doomed to dominating EVs, battery technology and what have you; but now now is doomed for real this time. Maybe, but probably not.
The Chinese government on mixing public planning and private enterprise is probably more pragmatic than westerners who have decided that all government involvement is doomed to failure. Our governments are full of practical men, who believe themselves to be quite exempt from any intellectual influence, but are usually the slaves of some defunct economist or other.
This response is kinda what I'm talking about when I brought up Steve Hsu et al. I appreciate your pointing to the Sinification blog and will definitely check it out more and maybe it will answer my request for a better model.
However, you highlight China's accomplishments in EV which is great for new types of cars and affordability, while simultaneously depreciating western governments as being dumb and beholden to "defunct economists" and that the Chinese gov is more "pragmatic".
It just feels like there's just no good critiques of Chinese gov policy, like this is just an update of the "inscrutable Orient" to "Pragmatic and always right China".
Sorry that as well as an excellent blog I gave my opinion, I’ll do better next time and try match my opinion to yours.
The defunct economist quote is from Keynes.
Good! I'm glad you're finally coming around to reality. Just kidding. I didn't mean to annoy you. Aside from a few lines of text, I don't really know your opinion on my OP question. Maybe I'm projecting.
As for my opinion. (not directing this at you), but in general when I see claims about how X is all in terrible and there's little to nothing good about it, or that X is fantastic and there's absolutely nothing bad about it, it pings my skepticism-dar on both ends.
Regardless, I'll read the blog you recommended and look at some other sources.
I’ll actually recommend some pieces when I have time. There’s a large divergence of opinion. Feel free to to dm me if i forget.
i do think the west is entirely lost in an intellectual malaise that’s too dismissive of China’s state industry.
Tanner Greer at the scholars stage is pretty good on China. https://scholars-stage.org/ Check out older posts.
Gdp per capita, mean and médian income, and lifespan are a good basis for a model
So, for the last couple of years, I've been getting into beekeeping. I'm keeping(pun not intended) it small, only having two hives on a small green patch in the city where I live. I appreciate the combination of being knowledge based, outdoors, relatively little investment(both time and money) and not being a sport (enough of those hobbies already). Following the seasons and plants more closely has given me joy. As Henry David Thoreau put it: "The keeping of bees is like the direction of sunbeams". It is also a surprisingly deep and engaging topic, touching upon many topics such as genetics, botany, light, weather forecasts, entomology(obviously), etc. Because of all the variables involved, many questions have no definite answers, and, as for many other things, "if you ask 10 beekeepers, you will get 11 different answers".
Beekeeping also has celebrities like Karl von Frisch and Brother Adam, and legends like purple honey in North Carolina, or tales of the times before the Varroa Destructor mite was spread around the globe.
Anyway, sharing it here, since I have found that the topic engages broadly, and many people have exciting anecdotes. Would love to hear them.
Any suggestions on good books or other sources on the topic is also interesting, since filtering out the noise can be hard. My suggestion to others who are interested would be to check out Dave Cushman's website.
We don’t keep bees, but we grow flowers and other plants in our garden that attract them, and we now get many different species of bees and wasps in the summer months. Sitting close by and watching them could easily take up the entire afternoon and evening if we didn’t have other stuff to do—it is one of our greatest joys.
Sounds amazing. I do this in public parks too, but don't have a garden. Are you keeping count on the species? In my country there are around 200 different solitary bees and bumblebees, so plenty of species to identify.
We don't keep count, but I would estimate we've probably seen over a dozen species of bees alone in a single day. I've read there are 300 species in my area.
I lived on a commune part of the time when I was in college, and someone there kept bees. Once the whole hive decided to move. I’ve forgotten now what the likely explanation was, something to do with the queen. In prep for leaving they hung in a huge cluster several feet in diameter from the branch of an apple tree. The main beekeeper climbed a ladder and somehow got them into a garbage bag and returned them to the hive. Wow.
Thanks for sharing, I wouldve also been wowed by that. I haven't seen a swarm yet. What they teach in courses is that the bees can swarm when the queen is old or stopped laying eggs, space is tight or disease. Everyone always has a good explanation after-the-fact, but preventing it can be hard. I will sign up as a swarm collector with the association this summer, and hopefully receive some calls. Collecting it is all about finding and moving the queen, the other bees will follow her smell.
My cousin keeps bees (in the southern UK) and, according to him, varroa mites aren't beekeepers' only headache these days. Asian hornets are a new threat. Apparently a queen hornet starts a small nest in a hedge or similar early in the spring. Then, before long, when the nest reaches a certain size, they migrate to the top of the tallest tree they can find and start building their main nest. That makes destroying their nests, or even finding them, quite challenging.
Monks, such as Brother Adam of course, who you mentioned, make good beekeepers for various reasons. One is that the monks have a bland diet, and bees can sense their keeper's diet via the emanations from their breath and even skin. After eating just porridge and similar for a few days, the keeper could gently dip their hand into a bee swarm and locate and extract the queen without exciting the bees. But I imagine trying that stunt after eating a strong curry the previous day would not end well!
I have also seen many beekeepers voicing their concern regarding the asian hornet. I didn't know they build in the treetop. Sounds difficult, like you say. I have heard that beekeepers in Germany switched from 'hating' the european hornet, to loving them, since they are one of the asian hornet's only natural enemy(in europe). In comparison to bees, they are also quite big and nasty. Where I live they have not been sighted yet, but will probably arrive in the next decade.
I never thought about how Brother Adam smells, but that's a fascinating theory! I have been taught that strong smells annoy bees (dont use deodorant), but never extended it to diet. Maybe I will try out a bland and/or spicy diet myself and see what happens!
"The Mind of a Bee" was reviewed here a few years ago. I found it a good read.
I think I vaguely remember reading that review, or atleast skimming it. Thanks for the reminder! Will have a new look
How accurately do we actually know population statistics (births, deaths, causes of death, age distributions, migrations)?
I know that in the US there's the Census at least every decade. But even that's got to have significant error bars. Smaller countries in Europe with population registration (where residents have to register their address with the government) can probably do better. But places like sub-saharan Africa, where records of any kind are at best faulty and/or there are long-running civil disturbances? Or chunks of China where the local government has every incentive to fudge numbers in all sorts of (probably mutually contradictory in cases) ways?
This is all part of why I'm so skeptical of foreign aid claims in many ways. "X lives saved by Y intervention", "War killed Z people"....can we actually measure that with any accuracy? When all the incentives for *everyone* involved point toward fudging the numbers to inflate X or push Z to something more palatable to the commenter's side? Even if people are trying to be honest (a very bad assumption in general), all it takes is a tiny systematic error repeated a lot of times to get very large (relative to the underlying numbers) overall skew. This is compounded in the international aid field (relative to the war side) since in war there are two sides trying to fudge numbers in different directions. But in international aid, all the incentives for all the researchers, governments, and organizations involved are to inflate the usefulness of the aid and maximize the reported size of the problem.
It may still be worth doing, but I'm struggling to see how we can get below "at least one order of magnitude" error bars.
Asterisk Magazine ran an article about this recently. https://asteriskmag.com/issues/11/why-governments-cant-count
I think that in Western countries (or at least European ones, I don't know as much about Anglosphere) the number of births and deaths is extremely precise. People don't die unnoticed, and babies don't get born unnoticed.
For causes of death, this is pretty precise with the categorization system that is used. For example, a death certificate in Germany may list "Pneumonia caused by Stroke", where both Pneumonia and Stroke are official categories in the German list of deaths. In some cases where the doctors are not sure, they can also list more than one reason (multimorbidity). I think misclassification is pretty rare. In general, the resulting numbers are very precise, and you can answer any question within the classification system.
Of course, you can't answer questions outside of the classification system. If you remember back during the pandemic, people got super-upset because it was impossible to get official numbers of people who died *from* Covid. Those would be listed to have died from pneunomia (for example), because it is too laborious to decide in every case whether the pneumonia was caused by Covid. Those are questions that one cannot answer from the official records.
The solution to all this is statistics, of course. This was why we were sure during Covid that most people dying "with Covid" also died "from Covid". Both by looking at the total number of people who died, but also because some studies took the effort to examine a few dozen to hundred people who died "with Covid" and investigate how many of them died "from Covid". (Some countries like Denmark even did this for all deaths.) A similar thing you can do with interventions in Africa. You can take two regions, distribute Malaria nets in one region but not the other, and see how many people die from Malaria in both regions.
Finally, for the census in Western countries, the reason for the error bars are not that people die unnoticed, but it's that people move unnoticed. If a person leaves Germany and moves to a different country (or even a different district within Germany), then they should tell their old district that they have moved away. But in practice, not everyone is so conscientious to tell their district. There is no automatic update with the new country, so those people remain falsely in the files. In the US, it seems that unregistered immigration is also a big factor: people moving to the US illegally who don't register. In Germany, I think this is not a big factor. For the magnitude, the last census in Germany was in 2024 (after ten years), and it corrected the population in Germany from 84.1 to 82.7 million, a correction of 1.6%. For the people in the files, the age is also usually correct, so the age distribution is also very well known except for the same 1.6% of people who are falsely registered.
In the US, birth certificates and death certificates are issued and tabulated on the county level. and reported upward to the NCHS. Parents have every incentive to get their kids registered, otherwise it would be hard to establish identity for all sorts of IDs required for modern life. Death certificates are handled by counties, and they must include a primary cause of death plus any contributing causes of death. They get sent to the NCHS and CDC.
Census is supposed to count the remaining folks in our country who didn't start out with a birth certificate here (or earn a death certificate).
The EU has promulgated uniform birth and death reporting standards for its members. I don't know about censuses. And cause of death codes are now a universal standard format. So we know how many people are dying from which diseases by country. Western countries keep pretty good counts. Developing countries less so. But in the modern Western world, birth and death numbers should be pretty accurate.
China is kind of an interesting case. We're pretty sure that CPC keeps good numbers. They're reported upward from the municipal level. But they don't necessarily report those numbers to the rest of the world. Officially, China has ~1.4 billion people. But demographers have done the math and given their reported population 30 years ago, and their falling birth rates, it's difficult to believe they reached 1.4 billion. And energy usage and grain consumption (which are reported) suggest a significantly lower population. Some demographers place them between 900 million and 1.1 billion, with many leaning towards the lower end of that range.
Regarding China, I have known that the official numbers are implausible, but the only concrete claims that I have heard (with substantial justifications) were around 1.25 billion. Do you have a link for the lower numbers? Even 1.1 billions would be a lot less than I thought, and 0.9 billion would come as a huge update to me.
I'm going to have to fade on this one. This was a subject I was only passing interested in, and I didn't pinboard or bookmark the references that I dug up. Several were from anti-China or anti-CPC sources. So, I'm not going to die on that hill. I did locate this paper, which claims China's nighttime light output by satellite suggests a much lower GDP (which I concluded indirectly supports the lower population estimates). But this is from one of those biased researchers I mentioned...
https://thedocs.worldbank.org/en/doc/350051528721174623-0050022018/original/Nightlights.pdf
Thank you!
My number comes from estimates by Yi Fuxian, a demographer who has investigated inconsistencies in Chinese official numbers. For example, the number of now-18-year-old does not even remotely match the number of newborns 18 years ago. He has dug out a lot of official numbers, like the number of vaccinations, the number of school children of different ages, and so on, to reverse engineer an estimate of the correct numbers.
Yeah. I'd expect that the errors in developed Western nations (EU+UK+nordic, US, Canada, Australia, NZ mainly) are pretty good but still subject to errors on the level of sub 1 percentage point.
Outside of that, I have no confidence.
A prominent example of unreliable population numbers is (was?) Paraguay where the latest census (2022) revealed the "loss" of a million residents:
https://www.economist.com/the-americas/2023/10/26/over-a-million-paraguayans-disappear-in-the-latest-census
I believe that in the case of Paraguay the previous overcount was produced by the estimation that had to be done to account for those who didn't/couldn't be counted in the census (2012). I don't know how the estimation was done, but I assume that some of its assumptions might have been invalid:
"For instance, in Paraguay, the 2012 census may have missed a quarter of the population. The significant underrepresentation of rural population we found in the gridded datasets raises the question whether such gaps in the census are more widespread than assumed to date, and how reliable current global population estimates really are." (Global gridded population datasets systematically underrepresent rural population, Láng-Ritter, Keskinen & Tenkanen, https://www.nature.com/articles/s41467-025-56906-7).
UPDATE: Regarding the last Paraguayan census, I found an article on its controversial results: https://www.abc.com.py/economia/2023/09/06/ine-defiende-resultados-de-censo-y-asegura-que-cobertura-fue-del-94/
A few translated quotes:
- Regarding census coverage: "[The president of Paraguay's National Statistics Institute (INE)] affirms that census coverage was 94%, and therefore he trusts the data"
- Why was there such a large gap between prior population projections and the results of the 2022 census: "The president of INE, Iván Ojeda, explained that the data is surprising because Paraguay had been relying on projections for 20 years, following the failed 2012 Census."
- Where does the 94% coverage figure come from: "He said that a post-census survey confirmed that 94% of the population was counted."
- Other data was contrasted with census results: "Furthermore, the Single Student Registry was consulted and it was found that there are 906,000 students registered with the Ministry of Education, aged 7 to 15, a number that coincides with that obtained in the 2022 Census."
- Related to what Beowulf888 mentioned in another comment (births and deaths): "When consulting birth data from recent years, they observed that between 80,000 and 100,000 births are registered annually, while the projection was 140,000 babies per year."
And I'm not necessarily claiming *intentional, malicious* fudging. Could be unconscious, could be making *justifiable* decisions that all end up pointing in the same direction, etc.
Donut Lab is a Finnish company that just unveiled a solid-state battery at CES that is making waves in engineering circles. Their claims are either revolutionary or a complete scam.
Their claims:
* 400 Wh/kg (vs ~250 Wh/kg for the best current Li-Ion batteries).
* 5 minutes fast-charge (vs ~8 min best current Li-Ion, at the cost of battery degradation).
* Fast charging with no degradation at all (major problem for Li-Ion).
* Designed for up to 100,000 charge cycles (vs ~1'500 to 2'500 best current Li-Ion).
* >99% capacity retained from -30 to +100 C° (problem for Li-Ion).
* Made of globally abundant materials (major problem for Li-Ion).
* Extremely safe (fire safety is a major problem for Li-Ion product design).
* Currently in production, shipping to customers Q1/26 in a motorcycle taking pre-orders now.
* Price point on par with Li-Ion.
* Easily scalable production techniques, already scaling up massively.
If true this will be a major breakpoint for electrification in general, especially for EVs. They would certainly be a big improvement at my job (airforestry.com).
If a scam, they are very scammy indeed. If they manage to keep up the act for long enough, I worry many genuine battery start-ups will be de-funded in the meantime.
Points against:
They appear to have no registered IP, at least not that I can find (I might be doing it wrong):
https://euipo.europa.eu/eSearch/#details/owners/2422084
The same founder also runs an AI start-up that consists entirely of buzzwords:
https://www.youtube.com/watch?v=ilgJKjiDLV8
And the CEO is also a co-founder of the motorcycle company Verge, that is allegedly shipping with this battery to customers Q1/26.
My current estimate is 60% on outright scam, 25%: There is a real product but it is being massively over-sold to the point of not revolutionizing the industry, and 15% on this being generally real as presented, but still allowing for some hype.
Does anyone here have a different take? I'd be very interested to hear!
Donut Lab website:
https://www.donutlab.com/battery/
Youtuber analysis, cautiously optimistic:
https://www.youtube.com/watch?v=Qsi-TS8PX7U
Interview with CTO (Interview starts ~21:21):
https://youtu.be/O8PjlwZVLVE?si=Q7Br9X21YKW-cY7_&t=1277
I would love that to be true. But the reddit energy and battery subs are very sceptical.
Mhh I got very excited but immediately turned very skeptical. They have no technical sheet or better information about any product specs.
I understand industrial secrets and all, but they don't even mention what "green" materials they are supposedly using, which would be a great point to underscore how green the batteries are. Instead they provide a single image with points and no labels.
I'm no geologist either, but the distribution of the points looks unbelievable: assuming they represent ore deposits, it seems unlikely that Europe would be the densest and places like Kazakhstan or other parts of Central Asia would have like four or five lights.
The CEO is apparently a "Tech strategist and serial entrepreneur with proven track record", but does not bother to list his past accomplishments?
All in all, there seems to be very little technical material and a lot of publicity.
Also, maybe this is normal in Finland, but isn't it weird that the website for a Finnish company would have no Finnish language version or anything? I get it's aimed at an international market, but it gives me the impression that they are trying to hit it big without bothering with the fundamentals; this is perhaps the norm in the tech startup scene, though.
Re: webpage in English, I looked them up online and they seem to have an address in the UK?
Address: Unit 6 Cornbrash Pk, Chippenham SN14 6RA, United Kingdom
Perhaps that's a manufacturing base or something.
The CEO has a Finnish LinkedIn but, uh, his work experience seems to be mostly/mainly "co-founder of this, that and the other company".
He has a degree in Computer Science from the University of Helsinki and yeah, I have to say, this is looking more "Elizabeth Holmes" than "Elon Musk" here (depending, of course, on your opinion of Elon).
The company is also registered in the UK in September 2024 (https://find-and-update.company-information.service.gov.uk/company/15962014/filing-history), and the CEO appears to be the only shareholder.
How likely is it that a 16 month old company is ready to deliver a world-changing technology at the consumer level?
The address is registered through Domains By Proxy LLC, which according to wikipedia, has a scammer problem. Although by the tone of the article, it doesn't seem like the same kind of scammer that Donut Lab would have to be.
The CEO's CV you posted below really cements my view that the battery has some undisclosed fatal flaw, or is straight-up nonexistent.
I guess we'll find out either way in a few months
Today I learned that the company I work at has a Swedish version of it's webpage 😂
Having a page only English isn't weird at all, and is true for a lot companies here in Sweden, and I'd be surprised if it's different in Finland. I don't count that as evidence either way.
As to the secrecy, I'm torn. If they released more specifics and it all seemed congruent I'd definitely be more convinced. But in the world where it's real, they don't have to optimise to convince me, only investors and partners behind closed doors where the IP is safe. And of course they will be copied to some extent regardless of patents, so every breadcrumb they leave out reduces their moat.
But of course if it is a scam they might not have anything to show. But then I'm confused why they wouldn't make up something more flashy? If it is a scam they are quite well calibrated to extract maximum credulity from minimum evidence from me at least. Is this in an of itself evidence for or against?
I realise I'm going off impressions rather than data here, but going by the CEO's LinkedIn, he's very proficient in bullshit tech vibe buzzwords. This man has done it all! (I'm getting rather fond recollections of our Hexayurt friend from this CV):
"Board Member
Cova Power
Oct 2025 - Present · 4 mos
Cova Power makes the world’s first series-production electrified smart trailer conversion, reducing fuel consumption by 54% while dramatically reducing the overall cost of truck fleet operations.
Co-Founder and Chairman of the Board
ESOX Group
Oct 2025 - Present · 4 mos
Esox Group takes the Western military capability to a whole new level in the most critical area of modern warfare: Unmanned Autonomous Vehicles and intelligent military EVs.
ESOX Group has the most capable hardware technology platform for building next-gen aerial drones, unmanned ground vehicles, submersive drones and other modern EVs — including what we know as Donut Motors, but in military specification — combined with cutting edge unmanned vehicle software and AI-driven digital battlefield simulation platform.
Together these technologies allow the creation of drones and other UxVs with groundbreaking payload capacity, speed, maneuverability, range and agility never before possible, combined with onboard AI capability that outperforms the current state-of-the-art technology by an order of magnitude.
Co-Founder & CEO
Donut Lab
Nov 2024 - Present · 1 yr 3 mos
Reimagining the electric mobility industry by building the world's first universal EV technology platform with next-gen software development tools.
Advisor
Aito Ventures Oy
Feb 2025 - Present · 1 yr
Aito Ventures is renovating the trillion dollar global renovation industry.
Verge Motorcycles
5 yrs 2 mos
Co-Founder & Advisor
Nov 2024 - Present · 1 yr 3 mos
Chairman of the Board
Jun 2022 - Present · 3 yrs 8 mosJ
Co-Founder & CTO
Full-time
Dec 2020 to Nov 2024 · 4 yrs
Verge Motorcycles develops and manufactures the world’s most advanced electric motorcycles. Verge TS Ultra is the most powerful motorcycle in the market today with the longest range, fastest charging time and best riding experience. In addition, Verge enables other motorcycle manufacturers to build better vehicles with its advanced technologies in the areas of motor, battery, electronics and software. As a CTO I was responsible for Verge’s software division. Embedded software, machine learning, connected vehicles, machine vision, HMI, HUD technology, PCBs and microchips, cloud/SaaS services, and all that jazz.
Board Member
Dec 2020 - Jun 2022 · 1 yr 7 mos
VERGE Motorcycles is a manufacturer of electric motorcycles. VERGE TS, the world's first Hubless Electric Roadster, has received numerous awards and recognitions for its innovations and design.
Co-Founder & Chairman of the Board
Stingbase
Jan 2024 to Present · 2 yrs 1 mo
Stingbase develops AI-driven investigation and crime fighting platform for law enforcement agencies and private cyber investigation units. Stingbase's cutting-edge platform revolutionizes criminal case resolution and increases success rates by over 100x.
Stingbase offers a comprehensive end-to-end solution, equipping users with advanced tools and features to seamlessly manage cases from crime-to-conviction. The platform's intelligent design minimizes the risk of human error, ensuring precision and efficiency at every step.
The solution enhances cross-border collaboration and promotes public/private partnership model to take international crime fighting to the next level.
Co-Founder & Chairman of the Board
ASILAB
Jun 2023 - Present · 2 yrs 8 mos
ASILAB builds artificial superintelligence and synthetic life-form that is more capable than humans in every aspect. Our approach differs significantly from everything else in the market, with built-in evolutionary principles, self-learning, synthetic neuroplasticity, observability, life purpose, long-term memory, and more.
We're building independent thinking organisms that develop on their own, continuously. And boy have we patented it.
Founder, Chairman of the Board
Shadow Capital
Oct 2022 - Present · 3 yrs 4 mos
Shadow Capital invests in cutting-edge deep tech companies.
BURST
4 yrs 1 mo4
Chairman of the Board
Oct 2024 - Present · 1 yr 4 mos
Founder, Advisor & Investor
Jan 2022 - Present · 4 yrs 1 moJ
Burst develops next-gen immersive music and entertainment experiences utilizing advanced AI and latest technologies. Currently Burst is working on the world's first hip-hop game for mobile platforms, which will be later on expanded to photorealistic VR.
Venture Bonsai
16 yrs 9 mos
Co-Founder & Chairman of the Board
Jan 2025 - Present · 1 yr 1 mo
Co-Founder, Advisor
Jan 2013 - Jan 2025 · 12 yrs 1 mo
Co-Founder & CEO, Member of the board
May 2009 - Jan 2013 · 3 yrs 9 mos
Venture Bonsai is the first equity crowd-funding platform in the world, pioneering the industry. The idea of Venture Bonsai was born after I exited my previous company and was looking for startups to invest in. So we built the platform and the rest is history.
If you’re building a world-changing deep tech company and need funding, we may be able to help.
Co-Founder & Chairman of the Board
KaseyOS
Mar 2024 - Present · 1 yr 11 mos
The team that pioneered the entire no-code software development industry is making a comeback. This time we’re building the world’s first AI-native no-code platform, and it’s going to change everything. Coming out of stealth in Q1/2026.
Official Member
Forbes Technology Council
Oct 2015 - Present · 10 yrs 4 mos
Founder, CEO, Chairman of the Board
Kasey Group
Mar 2007 - Present · 18 yrs 11 mos
Kasey Group has a long history in developing bespoke mission-critical enterprise software, such as cutting-edge modern ERP and manufacturing systems for automotive industry, modern learning systems, flight and leg management systems for the aviation industry and A.I.-driven cyber security solutions.
DeFinance Technologies
3 yrs 8 mos
Board Member
Jan 2023 - Sep 2025 · 2 yrs 9 mos
Advisor & Investor
Feb 2022 - Sep 2025 · 3 yrs 8 mos
DeFinance develops AI-driven crypto and stock trading platform.
SAP
Full-time · 2 yrs
Lead Product Manager, SAP Build Apps
Jan 2022 - Jan 2023 · 1 yr 1 mo
Director & Team Lead, SAP AppGyver
Feb 2021 - Jan 2022 · 1 yr
Changing the world with no-code, continuing the pioneering work we started at AppGyver. We call this next generation of no-code ”visual programming”, as it is equivalent to writing code in its power of expression. With turing [sic] complete visual programming there is literally nothing you couldn’t build — from mission-critical enterprise software to 5-star omni-channel consumer experiences.
Founder & CEO
AppGyver
Mar 2011 - Feb 2021 · 10 yrs
AppGyver pioneered no-code software development in 2011, building a global community with hundreds of thousands of developers. In 2020 the company took the industry by storm again, creating the world’s first turing [sic] complete no-code platform — enabling development of mission critical custom software at scale without writing a single line of code. In 2021 AppGyver was acquired by SAP, and the product is known today as SAP Build Apps.
Co-Founder & CTO
FansMagnet
Feb 2011 - Jan 2013 · 2 yrs
FansMagnet was a visual app development platform to enable artists to create fully customized mobile apps for fan base management and communication.
Customers included Warner Music, EMI, Universal etc.
Co-Founder & CTO
TRV Networks
Jan 2001 - Nov 2006 · 5 yrs 11 mos
TRV-Networks did business in areas of System Integration and Software Licensing, as well as international trade. Main business was sold in 2006."
That is indeed a lot for a CV 😂
Re the website: that's fair, I already suspected it to be not very relevant either way. But the extreme generic character of the information page is; this is a product that goes into high-end motorbikes i.e. consumer products. No specs, no demos, no nothing?
Also, I understand the "moat" part, but this is going into consumer products that are allegedly "available now", so hiding even the names of the chemical elements used in the battery seems unreasonably secretive
Yeah, I see your point. In the interview the CTO says himself that while he won't go into details, it won't stay secret for very much longer since it will soon be in the hands of any company that really wants one anyway, and several independent reviews/teardowns will probably be published shortly thereafter.
If it was a scam though, wouldn't they try to extend timelines more than "available now" and "shipping Q1/26"? I mean, this way we won't have to hold our breaths for very long. Any excuse to delay now would be highly suspicious, and they'd be hard pressed to make a clean exit by March. In many ways it would be *more* believable if it was Q1/27.
To my eyes they manage to exude the confident calm of someone who has nothing to hide, but also no reason to tell you just yet. But of course that's what they would like me to feel in either timeline.
A sense of urgency is a key factor in lots of scams.
You forgot the inevitable Manifold market:
https://manifold.markets/AristotelisKostelenos/will-i-deem-donut-labs-solid-state?r=UHBhdQ
I'm curious, why did you put the fact that the CEO is cofounder of Verge motorcycles as a point against?
Just checked the Manifold market and feel very validated in my prediction above! (it was at 73% at the writing of this comment)
Indeed, nice job
By the way your job seems very cool, never heard of this
I was hoping someone versed in the prediction markets would find me a link! Cheers :)
It would be way harder/riskier to partner with an independent company if you are running a scam, whereas if you run both companies the increased risk is very small. OTOH, if I knew ahead of time that an amazing EV battery would come to market (because I was in the process of inventing it), I might very well start a side business trying to get a head start on some of the use cases. If it was an independent company, this would be real evidence in favour: As it is I don't think the motorcycle is much evidence either way.
It would seem like a massive conflict of interest at the very least: "here's this revolutionary battery and oh yeah if you buy one of our bikes, we'll ship it with this"/"if you buy our bike, as distinct from a competitor's bike, you'll be one of the first to have a bike running on an amazing new battery!"
Too late, when you've bought the bike, to find out the hyped battery was only hype.
Not sure I would call it a conflict of interest but I see what you mean
I'm struggling to understand the situation with high electricity costs described in this article: https://www.bbc.com/news/articles/c1dz0dz0zkvo
Apparently this New York state family has an unaffordable energy bill of $1800/month. But that makes absolutely no sense to me. How can the bill possibly be so high? Electricity/gas cannot possibly cost that much per kwh. And why can they not reduce their energy consumption if it is that expensive? Is $1800/mth really the minimum amount a family in New York would need to pay to cover essentials?
I found it odd that their only option was apparently to cut the supply altogether. But really, none of it makes much sense to me, and I'm wondering if this is just some exceptional case reporting on a person apparently devoid of common sense, or if this is actually normal for some reason?
My bill once tripled while living in ny. The electric company started estimating my usage instead of measuring it every month. The estimates had no relationship to my past usage. Id eventually get the money back when they measured, but it was very annoying and did not seem to be in good faith.
It's just the shit state of modern journalism. You can't just say "There's been an increase in energy prices, here's some statistics", you have to lead off with some random individual.
It just occurred to me that this is the same basic phenomenon as how damn near every youtube video now has a shocked-looking face in the thumbnail.
I cannot read the article so I can’t address it directly but I can tell you that I have a house in the Northeast United States (100 miles NE of New York City) and our mode of heating is propane. Propane is very expensive right now compared to last year (currently $4.35 per c/f. We are paying 600 - $800 a month for propane. We cook and we use hot water so it’s not all heating. Our house is not weathertight; it could be much more efficient than it is, but it’s not absurdly leaky. Also, I keep it at 52°F for five days out of seven during the winter because we’re not always there. The big mistake was putting in picture windows. I always thought the windows of houses around us were very small but now I understand why. But when the sun is shining, it’s such a nice thing to have a picture window.
So part of the question is that, as I understand it, different locales can have very different and idiosyncratic and complicated pricing regimes. Electricity is not bought like a commodity at a market-set marginal price, it's bought like a service with very complex terms and conditions.
In particular, I think a lot of places have different types of 'surge pricing' that can produce absurd results. Prices that increase for each marginal watt used over the set minimum, prices that increase based on time of day, prices that increase based on grid-wide demand. These factors can lead to huge differences in price based on very small differences in total usage.
One example is the Texas winter storms that knocked out parts of the grid, leading to algorithmic surge pricing based on supply vs demand sending normal households outrageous bills.
https://www.npr.org/sections/live-updates-winter-storms-2021/2021/02/21/969912613/after-days-of-mass-outages-some-texas-residents-now-face-huge-electric-bills
It says gas and electricity, and it's a family of four (her, two kids, granny) so they could be running all kinds of devices day and night as well as the heating; looking up annual temperatures it says Greenwood Lake (where she lives) has cold, snowy winters and warm, wet summers so if she's running the heating constantly that mounts up. Depending on who the utility company/companies supplying her house, there might indeed be steep price increases:
"Winter heating costs are expected to jump 9.2% this season, according to the National Energy Assistance Directors Association, driven by rising electricity and natural gas prices and unusually cold weather.
Energy bills tend to be among the highest in the northeast US, the report shows. But households from California to Georgia to South Dakota are also feeling the strain of rising costs over the past year."
Everyone has their own phone, computer, TV in the bedroom, etc. etc. etc. that are always on (nobody, including myself, unplugs TV sets anymore), devices are charging, everyone is having multiple showers, microwave and oven and stove are going all the time, kitchen appliances, washing machine, dryer - you'd be surprised how usage creeps up, and then slap on peak pricing and there you go.
Mind you, I don't necessarily believe the price quoted, but it's easy to be running more devices than you think you are. The story says the bill tripled, so that would indeed be a hefty price jump. Since we don't know the power company, we have no idea of what pricing it charges.
EDIT: This site quotes prices for Greenwood Lake:
https://poweroutage.us/electricity-rates/ny/greenwood-lake
"Greenwood Lake electricity rates vary, but the average residential rate is around 22.52 cents per kWh, with a monthly bill averaging $143.65.
What is a good rate per kWh in Greenwood Lake?
In Greenwood Lake, New York, a good electricity rate falls between 22 cents to 24 cents per kWh. Tired of high bills? Shop during spring and fall to lock in lower prices before HVAC use spikes in the summer or winter.
What are electric rates in Greenwood Lake today?
Greenwood Lake electricity rates currently average 22.52 cents per kWh, but you'll likely pay between 21 to 25 cents depending on your provider, location, and plan type.
Why is electricity so expensive in Greenwood Lake, NY right now?
Your Greenwood Lake electric bill can be expensive because you live in a high-demand region with complex delivery systems. You pay premium rates due to natural gas costs, peak usage timing, and aging infrastructure that needs constant upgrades."
I don't know if 638 kWh per month is on the low side, but if her previous bill was $600 per month @ 25 cents/kWh, that's 2,400 kWh/month. That's hefty usage!
Now, there is a power company listed in Greenwood Lake that charges 44 cents/kWh, so if she's signed up with them, then yeah that makes it more reasonable:
$600/$0.44 = 1,364 kWh/month. Maybe her house is old and needs a lot of heating?
"Electricity rates in Greenwood Lake, New York depend on your home size and usage, so the kWh rate is more useful when comparing your rate to the average."
"Home Size & Type: Larger single-family homes generally use more energy than smaller apartments."
Let's throw some figures at this:
"In 2022, the average annual amount of electricity sold to (purchased by) a U.S. residential electric-utility customer was 10,791 kilowatthours (kWh), an average of about 899 kWh per month."
So 899kWh x $0.44 cents = $395.56 per month.
https://nrgcleanpower.com/learning-center/how-many-kwh-per-day-is-normal/
"Here’s a quick breakdown of average daily kWh usage by household size:
1-2 people: 15-20 kWh per day
3-4 people: 25-30 kWh per day
5+ people: 35-50 kWh per day
State Average Daily kWh
New York 25-30 kWh"
So it's a family of four living in New York state, say that's 30 kWh/day over 30 days in a month = 900 kWh/month x $0.44 cent per kWh = $396. That's the electricity, no idea what her gas bill was. And her monthly average bill *before* the increase was around $600.
So... it could be plausible that she's using that much power and then prices went up 200%. But it does sound like she and the kids are running every device they own 24/7 plus the house must burn through heating if it's old and poorly insulated.
Devices and lights are never the problem these days. Electric heating and cooling (and perhaps EV charging) would be more costly.
The price of propane went from $2.50 to $4.75 between last year and this year. Propane is the main way of heating houses in that region.
EDIT Unless the article specifically stated that it was all her electricity bill. Electricity prices have increased noticeably in New York State over the last year and I am not sure of what the rationale is.
Phones and tablets probably aren't adding much to the bill. The battery capacities tend to run in tens of watt-hours and even heavy users rarely need to charge them more than once or twice a day.
Desktop PCs might be a significant component if someone has a high-end gaming PC and is spending every waking moment maxing it out. A 1kW power supply is a pretty huge one even for a gaming PC and the usual rule of thumb is to add 20% to your theoretical max power usage as a cushion when deciding which power supply to buy, so call it 800W peak usage. 0.8 kW * 16 hours /day * 30 days = 384 kWh.
Or they might be running a bitcoin mining rig, or some kind of private datacenter for some reason.
>Maybe her house is old and needs a lot of heating?
That seems more likely. HVAC is usually the biggest-ticket item for household energy costs, followed by other major appliances (stove, oven, water heater) and stuff like pool pumps and electric cars.
Looks like Greenwood Lake gets fairly cold in the winter, so if they've got no insulation, drafty windows, etc and also have an inefficient heating system (possibly supplemented with liberal use of electric space heaters, which in the US are typically designed to max out what a standard wall outlet can supply (1.5 kW).
They also get hot and muggy summers, so they're probably running air conditioning during those months. AC uses a ton of electricity even in well-insulated houses. Again, if they have an older house and haven't updated the HVAC recently, they're probably using an older and relatively inefficient unit. They ma be using window units, which come in multiple styles: some of these are decently efficient and others are terrible.
Or maybe they're just running an indoor marijuana grow operation in their garage.
+1. There is a very good chance that house is heated with propane, which has increased $ a lot in the last year. I have a house approximately 50 miles north of Greenwood Lake.
Yeah, I'm betting on an older house, cold winter, ramped up the heating, the kids are gaming, granny needs the electric heater in her room, etc.
I don't know why the price jumped so much (quite apart from whatever usage) but maybe they signed up with a new plan? or maybe this *was* a new plan and the introductory discount period ended and they got whacked with "surprise, this is what you'll *really* be paying from now on!"
A lot of places in the US have been rolling out new residential rate plans. The old standard was you pay a flat XX¢ per kWh, while the newer ones tend to be either "Tiered" (you pay more per kWh after you use more than a certain amount) or "Time of Use" (cheaper rate during off-peak hours, usually overnight and in the morning, more expensive in the afternoon and evening when AC is cranked up and the sun is going down so there's less solar on the grid). If they use gas for appliances and heating (this is common in much of the US), that often also has tiered rates. Not sure if this is the case there, but if they just got switched to a new rate plan and didn't notice, that might be part of the problem.
Also, gas, electricity, and (as B mentioned) other heating fuels like propane and fuel oil have gotten a lot more expensive in most of the US over the past few years. Inflation is part of this, but also because of tariffs and other supply chain and regulatory issues.
I originally thought she might live in far upstate NY, where it gets very cold. But she lives 50 miles from NYC. So there is definitely something very fishy about the story.
Edit: The article also claims that her bill suddenly tripled, which seems even more dubious.
If electricity costs in the US are anything like those in the UK then it is all too believable. It would take only a 1 or 2 KW electric radiator in each of several rooms and one or two in corridors and hallways. Also, this family may be charging an EV from their home supply, which makes it even more believable!
EV charging isn’t going to make a dent in those kinds of numbers unless you are commuting 100s of miles a day.
Maybe they were bitcoin mining?
The article is conspicuously lacking in the sort of quantitative information that a rational person might use to understand the situation. There's a smattering of numbers, most of them scary, but mostly unconnected to one another or to any real context.
So, yeah, I'm pretty sure that they picked the absolute scariest number they could find to lead with, and didn't provide any details or explanation because if they had it would be obvious that this is not a representative situation.
A quick google suggests the average US household electric bill is ~$150/month. And I'm certain that somewhere in the UK some fool is spending £1,800/month trying to e.g. keep a Scottish castle warm through the winter with electric heat.
My immediate thought was they were trying to heat their swimming pool during winter. If I were a savvy reporter, I'd want to have seen their utility bill to try to understand it. But savvy reporters are few and far between.
I'd have to agree with John on this one. It's a (typically) poorly researched article that makes claims without any supporting information. Quick Googling says the average household in the Northeast uses about 600 kWh/month, and electricity rates average about $0.30 kWh. Transmission & distribution charges, meter charges, and taxes that would add another 10-15% to the bill. --> $240/month.
We'll probably find similar numbers for gas charges (but the price of natural gas has risen). That would likely offset some the electricity charges if they use gas.
Many Northeastern homes have furnaces that use heating oil. The average price is around $3.65/gallon. My mom, with her big old house in New England, used to purchase 1000 gallons of heating oil before winter kicked in. And she might have to order a second delivery before the weather got warm again (if the winter were particular bitter). Say in the range of $6000/year. That would be about $500/month over the course of year. But no gas, and no electrical heating. So her electrical bill was lower.
Any fun ACX-adjacent things to do in the Bay area (Palo Alto, Berkeley) in January? I am visiting for two weeks.
Join the various Discords for access to events? I don't know how maybe first try the new York obnyc discord and then ask in it for an invite to a sf discord. I have never done this
Googling i found this discord that is easy to join: https://discord.gg/ShCvU3Nc4
Huh. Maybe it is that easy?
This is a follow-up to my post on Anavex Life Sciences exactly half a year ago (https://www.astralcodexten.com/p/open-thread-390/comment/135081998). That post was meant for entertainment and could be seen as ridiculing and negative, but in fact I thought that Anavex has the goods. That this was a case of "heuristics that almost always work" not working.
Well, the heuristics did work for now. The European Medicines Agency (EMA) has issued a devastating rejection of blarcamesine against Alzheimer's disease. Even before it clearly indicated a rejection, Anavex stock was already doing badly, instead of the more common run-up rally prior to binary events in biotech. A number of short reports had appeared, and afterwards shorts' comments were like "this trade was such easy money, it's embarrassing", or "nobody with a clue about statistics would have invested in this".
Disclosure: I own Anavex shares --- with gritted teeth in the face of bad management, but what can I do given stocktwits.com/Hosai/message/639122935, an enterprise value of less than 300 million dollars, and my belief that the company is not a scam (cf. Cassava Sciences). Truly scammy moves, if any, would have happened under different management more than a decade ago. Critics point to many discrepancies between and within various presentations of the trial data, but I concluded for myself, a year ago (when I had more time), that it's probably innocent. This isn't investment advice, though, especially as the stock may tumble again after final rejection: Anavex is appealing, and while conditional approval can't be ruled out, too many seem to be drawing a dubious analogy to Leqembi and Kisunla, both approved by the EMA on appeal.
The Anavex CEO espouses a philosophy of running small trials and then endlessly analysing the data. He even co-authored a paper titled "The new big is small". Had he quickly gone on to start a confirmation trial after the data read-out three years ago, that trial might now read out soon. But it is what it is. If I had the skills and time, I would investigate at least the issue noted in my next paragraph. Instead, I put it here in case someone else is interested. After all, if there was a safe and convenient pill that could "functionally cure" Alzheimer's, wouldn't it be a shame for it to be wasted? (I don't know how far it is possible or legal to use the molecule without official sanction, but the original patent is from 1997.)
The EMA may have ignored the Open Label Extension (OLE) of the trial because of too few finishers. But here is a claim that their results at the end are simply outside the distribution of (untreated) Alzheimer's trajectories: stocktwits.com/Bio_Oko/message/629232900. Is this correct? It's by Peter Karol, an independent analyst, Bio_Oko on Patreon, who has been claiming a bimodal distribution for years, something Anavex is now officially saying as well. Half of the population, the ABCLEAR3 group as Anavex calls it, with "lucky" genes, could have Alzheimer's slowed down by 75 percent or more. (And possibly other dementia as well, as what is supposedly stimulated is a general repair or backup mechanism, unrelated to amyloid or tau if I understand correctly.)
That spectacular hypothesis comes partly from post-hoc GWAS analysis, and we don't know how many genes were tested (probably a lot, otherwise Anavex would tell us?); but still, I would no longer bother to caution "not a cure" in parentheses like I did in my original post. I did not take Karol seriously enough, because I find his writing hard to read.
Another issue of note is the "Linear B-spline Mixed-Effects Model" that Dr Kun Jin developed to examine the effect of the many trial dropouts already before the OLE. He is former FDA Lead Neurology Statistician and joined Anavex after the trial results were out (other high-calibre people joined as well). While the skeptics argue that weaker patients dropped out of the treatment arm, leading to a spurious efficacy result, it could also be that responders in particular dropped out, so the trial result would *understate* blarcamesine' efficacy: this is apparently what his model is supposed to resolve. Does it succeed? It's possible that it was actually Dr Jin who decided or advised against the need for a confirmation trial. (And that the EMA isn't convinced, despite his model, because it thinks dizziness in the treatment arm effectively unblinded the trial ---although how could you *ever* then prove via RCT that blarcamesine works? Might as well grant conditional approval based on the existing data and withdraw it if the drug doesn't work after all.)
In nominative determinism, I recently found out there is a marine conservationist named Ocean Ramsey (she is famous for swimming with sharks and is popular on Instagram). I assumed she must have made this up and changed her birth name, but I checked her Wikipedia page and there's no mention of that.
I was also recently reminded that Tito's Vodka, evidently the top-selling US liquor brand, was founded by a guy named Tito Beveridge.
A former professional soccer goalkeeper in Germany is called Tore Wächter.
Tore is a name, but, unrelatedly, also translates to "gates" or "goals" in the sport sense. "Wächter" translates to "guardian").
If the "R. Good" incident was real, and her west coast sister does bear Angela Lansbury's character's name, how was she targeted? Or is it elaborate AI fiction.
The principal horn player of the Philadelphia Orchestra long ago was named “Anton Horner,” and currently the Sarasota Orchestra has a horn player with the last name “Horne.”
I’m obsessed with nominative determinism. Someone should make a website to collect examples.
My dentist in London was a Dr Fang, no really! A web search to see if he is still around reveals there is another London dental practice called Dr Fang, based on one of their dentists' surname: https://drfang.co.uk/
Walking through an Austin neighborhood, my college GF and I were amused to note a "Dr. Casanova - Ob/Gyn". I don't think she ever followed up on that.
Within an hour's drive of that (small town nearby) used to be Nulliver's Travel Agency.
A doctor who has performed C-sections has a family name meaning "carver".
Both times I've seen the name Beveridge it's been connected to beverages. Somewhat famously in old internet meme lore, Michael Beveridge was a leading member on the board of the Yorba Linda Water DIstrict, who "ousted" Arthur C. Korn of the famous "korn.jpg" in a dramatic but entirely fictional board coup.
(This was famous solely because the water district had a locally hosted website in the 90s where the native size of the .jpg files for the board members' portraits were astronomically large. In those days the web designer would manually set the resize dimensions on the html page as an attribute in the <img src=""> tag, but the locally-hosted .jpg had to be in a folder with a+rx permissions in order to display elsewhere. So a visitor could see the source html and download the original .jpg, and if so inclined could hyperlink it on another web board either to add data traffic to the image host for nefarious reasons or to cause board readers on 56k modems to spend 5 minutes loading the page.)
I have a number of early-New England Puritan ancestors/relatives, and being my clan's longtime genealogy hobbyist I've built quite a large database of them. Puritanism of course was explicitly determinist in its outlook ("predestination"). That perhaps led to these given names each of which I have multiples examples of in my records:
Innocent
Freelove
Deliverance
Lovegod
Praisegod
Desire
Obedience
Freegift
Prudence
Experience
Those were mostly but not exclusively given to baby girls. There are also historical examples of even weirder Puritan names that were closer to one-offs; the above examples seem to have been normal in 17th/18th century Massachusetts/Connecticut/Rhode Island/Vermont/New Hampshire.
Freelove seems to have been ahead of her time.
Innocent -> Desire -> Experience -> Freelove
In the art museum in Worcester MA, there are a series of C18th paintings by the local semi-pro painters. They're awful. These guys made their own pigments, and so the subjects have green and grey skin. The proportions are all wrong.
The subjects have fantastic names Faith Savage is an older widow. And there's a row of misshapen children painted by the perfectly named 'Freake painter'. Unhappily not a more common name...
Heh....given how the Puritans viewed art generally, those surely must be "stage names"!
But there is mention of her place of birth, which is the likely common cause of the name and occupation.
I saw a plaque the other day memorializing the "Mother of Our Aquifer" - A Mrs. Fay Sinkin. As this was her married name, it doesn't count as nominative determinism, but the unusualness of the name combined with her founding role in protecting the water supply delighted me all the same.
I'm a long-time SSC/ACX reader. I've begun a Substack where I'll be posting once or twice a month on history, culture, society, and anything that catches my eye. I started out with a book review of "The Cheese and the Worms" by Carlo Ginzburg. In case anyone is interested, here it is: https://connorsscratchpad.substack.com/p/book-review-the-cheese-and-the-worms
If things go well I'll definitely up the posting schedule.
Looking for advice on dealing with toddler eczema (it's probably eczema, we're waiting for a biopsy result). Steroid creams and anti-histamines have not worked, and the situation is getting dire. Poor kid is scratching himself bloody and can't sleep
Second the mittens suggestion. We used a combination of lotions and oat-base bath products as well as a water softener as our water was very hard. Not sure how much those helped, she has since grown out of it for the most part.
Maybe hand-washed silk sheets and pillow cases would help. They sound hideously expensive, but these days aren't really.
We dropped all wheat in his diet and use fragrant free and dye free laundry detergent. That cleared it up for our 2 year old. Have only seen it one time since we went no wheat diet and it was when I accidentally have him a dumpling that had wheat in it.
1) Remove fragranced soaps and washing machine liquid.
2) our kids' skin improved when we changed to the Aveno baby gentle soap.
It might be food, but given that's a huge pain you want to rule out other stuff first!
We were told that food was probably the last resort to check. YMMV.
https://www.solveeczema.org/ Was brilliant for us, and also an interesting story of science. Massive difference over the course of a few weeks. We did need to get a water softener to be able to wash clothes with Castile soap, though soda crystals might work temporarily. We hand washed my sons clothes with castle soap until we got the water softener in.
Aquaphor is very helpful. Cerave was not- we discovered it has a detergent that just made it worse.
Best wishes! I know what you and your child are going through.
In the interim, getting him gloves so he can’t scratch himself might be better. Never dealt with skin issues but if my son has a bandaid on his finger he is compelled to take it off so we just put snow mits on him so he couldn’t.
I’m going to be in SF for a conference from the 17th to the 22nd of this month. Are there any interesting/fun ratsphere happenings that I should consider going to? Or anything else I should do while I’m there?
How much do you think your current actions, habits, projects and activities are helping to reduce existential risk from advanced AI (extinction, gradual disempowerment, extreme power concentration)?
Given your large platform and skills and talents, what are the best things you can do to help improve the situation and reduce existential risk (of the 3 major kinds)?
If you're talking about Scott specifically, I think he's probably reasonably close to that frontier already.
He has a competitive advantage in being a widely recognized intellectual on other topics, who also cares about AI risk and brings it up a lot, such that a wider audience is exposed to it by someone they already think of as reasonable and discriminating.
I feel like Scott *could* push on AI risk a lot more, but that this would be likely to just burn credibility and get him dismissed as a crank on this issue by the type of person who like to dismiss this issue.
In the abstract - I think Yuds + Scott does more to combat AI risk than 2 Yuds, at least in terms of public persuasion (2 Yuds might do more in terms of actual research, but I don't think Scott has a competitive advantage there).
> I feel like Scott *could* push on AI risk a lot more, but that this would be likely to just burn credibility and get him dismissed as a crank on this issue by the type of person who like to dismiss this issue.
I would say that he already does push on AI risk a lot and he already is a crank on the issue.
Appreciate you're probably strapped for time, but would be interested in your thoughts on this critique of the influential METR time horizons graph. In any case, think it'll be of interest to your readership.
https://open.substack.com/pub/arachnemag/p/the-metr-graph-is-hot-garbage?utm_source=app-post-stats-page&r=18kjq3&utm_medium=ios
These are not the objections I expected. I was expecting a more Andrej Karpathy style critique - that excessive focus on quantitative benchmarks pushes AI development towards easily quantified and verified reward signals, and side steps the most challenging parts of general intelligence - continuous and adaptive learning under uncertainty. Humans constantly have to navigate situations where there is no explicit reward, so we have to create our own intermediate goals to optimize for. General intelligence is not a set of capabilities, but the meta-skill of being able to acquire new abilities through interaction.
I don't think it would be a good use of resources for METR to redo their human baseline with lots more devs. Experienced devs are expensive - and paying them for 97 tasks would probably cost tens of thousands of dollars. It probably wouldn't change the numbers much (the task lengths might change, but the trend would probably stay similar). And it wouldn't address the fundamental limitations of benchmarking.
See my reply to Scott. Don't accept either that changing task lengths on a re-sample is unimportant, or that trend would necessarily be similar. I think the objections you were expecting come up implicitly in the discussion of task realism, but I agree that they're independently important concerns deserving a deeper discussion.
This essay automatically makes me dislike it by peppering its text with vituperations about how STUPID and UTTERLY WORTHLESS the benchmark is and how there is NO POSSIBLE WAY that ANYONE COULD EVER DISAGREE WITH THAT, and making me read through paragraphs and paragraphs of that in between the actual complaints. I rarely see essays that do this have anything useful to say, so I start out biased.
I am extremely not an expert, and I don't speak for AI 2027 or anyone like that, and this is just off the cuff. But it looks to me like the essay's first point is that if you replace the full suite of tasks with a subset of only the hardest tasks, the success rate goes down. This is tautological. The essay argues that the hardest tasks are the most "real", but this is meaningless - I'm a real person, and I frequently ask AI to do extremely easy things for me. But more important, I think the whole exercise is vacuous - there's no objective answer to what is "really" a "one hour task", and METR isn't trying to establish this - they're trying to fix a specific consistent definition, then see how AI success on that definition changes over time. The point is to measure the rate of improvement, not to say "Look, AI can now do exactly 56% of tasks, hoho". The essay's preferred "messy tasks only" graph also looks like it doubles about every seven months, so who cares which one you use?
(except that the messy-tasks-only graph has such bad floor effects that you can't measure anything at all before mid-2024)
The next point, about only three engineers per task, could potentially be more relevant, if it meant that their estimates were basically random and they couldn't measure improvement at all. But they include error bars on their graph, I see no reason to think those aren't the real error bars based on n=3, and the error bars are too small to affect their conclusion.
All measurements are noisy and involve methodological decisions and compromises with practicality. The point isn't whether you can find some way that it's not The Exact Objectively Correct Number Handed Down From Heaven, it's whether the methodological fuzziness is big enough that you can no longer trust the conclusion. My impression is this isn't true, with one argument being that METR does find very consistent progress, and if their method was returning screwy results you'd think they would obey a less elegant law. So I conclude that whether or not you like their method, they're keeping it consistent from one AI to another, and the remaining noise/sample size problems affect the results by less than an order of magnitude, in a regime where whatever AI capability produces these time horizons is increasing by orders of magnitude.
I see the comment thread where the author argues with Jitse Goutbeek on this point, but IIUC (definitely not sure of this, leaning heavily on Jitse's analogy) the author is there making a different point that you can never truly measure "time horizons", but only time horizons on some specific set of tasks measured in some specific way, and for all we know they might be different in some other regime. I think that's fine. But first of all, it's not really related to any particular complaint about the measurement process - even if there were 5000 engineers and it was the messiest 1% of tasks, you could still make this argument. And second of all, I think METR's response would be that lots of different things measured lots of different ways still show a similar pattern, for example https://metr.org/blog/2025-07-14-how-does-time-horizon-vary-across-domains/ , the 80% reliability trend, the messiest tasks subset, etc. I don't want to claim they're all exactly 7.1 months or whatever exact number METR reports as their headline result, but I think METR has successfully created a strong prior that whatever specific thing you care about will be an exponential with a doubling time measured in months rather than years, and that (for example) if some set of AIs go far above trend on their graph, that would be surprising and represent a phase shift in AI progress.
Appreciate the read and thoughtful reply Scott. Several things I want to push back on.
A quick note on tone first.
I went back and forth about some of the language here, but decided it was warranted. I think if a certain idea (study, benchmark, etc.) has a large enough portion of the mindshare, you need to push back pretty hard if you want to change that. I also think the tone is natural if you think the benchmark is as faulty as I do, so at minimum I think its defensibility hinges on my substantive complaints. We also shouldn't lose sight of how much influence this benchmark has plausibly had on folks’ allocation of labor, energy, money, etc. That also speaks in favor of a stronger tone being appropriate if they have in fact been misled.
I’d add that you’re not my typical reader, and for better or worse strong rhetoric moves people (I'd say for better, but explaining why would require a long digression). Maybe we're reading different things, but I've read plenty of worthwhile argumentative writing that takes a similar tone.
Onto substance.
It seems a bit unfair to gloss my discussion of task realism as tautological because "harder tasks are harder." Task realism and completion time are two different variables and their relationship is important. Nor do I agree that, if doubling times are preserved, lower success rates on more realistic tasks are irrelevant. Maybe you don’t accept the premise here, but I suspect that whatever models' rates of improvement, their performance will begin to taper at some point. If that's true, then it matters, potentially a lot, when that tapering starts. Substantially lower y-values on the y-axis mean lower capability at tapering. I think we can agree that how capable models are when tapering starts is an important piece of information. If so, we can’t simply write off the scalar differences between doubling time graphs for different task sets.
As for arguing that "the hardest tasks are the realest," that's not what I'm arguing, it's what METR is arguing, or really what they're finding. They construct an index measuring task realism, which they call "messiness" (they explicitly equate it with realism), and find that messier tasks turn out to be much harder for models (none top 30% success rates for the messiest half of all tasks). You may ask AI to do tasks you find intuitively easy, but by "messiness," METR means something like “degree of similarity to tasks a professional software engineer would do at work" (check out the properties they’re using to score messiness on pgs. 39-40 of the preprint). Considering ambient concerns about labor substitution, automating AI R&D, etc., I feel like we can agree that how models perform on this subset of tasks is important.
I agree that METR isn’t trying to establish how long different tasks “objectively” take, but nor do I think all they’re trying to do is to come up with some (i.e. any) consistent metric to measure improvement. They describe the y-axis as how long it takes “humans” to complete software engineering tasks. As a result, I don’t see how you can claim that these numbers are only meant to be relationally meaningful, rather than to also make outward-facing, empirical claims. That is how METR describes them, and I think it’s clear they want us to understand how capable models are by comparison with humans. If they didn’t, why not just describe time horizons as “capability scores” or something else much less committal? Why even run such a complex benchmark in the first place? Finding “a specific consistent definition, then see[ing] how AI success on that definition changes over time” is the same thing all the other benchmarks are doing, and METR is doing something different for a reason.
I admit to not having looked into how METR’s error bars were calculated before reading your comment, but I’m glad I did because I was able to confirm that they in no way solve back for my concerns about their exceedingly small / biased sample. Those error bars have nothing to do with the human baselines on the y-axis; in fact the baselines are fixed before they’re even calculated. Simplifying a bit (promise the complications aren’t pertinent), the error bars in the time horizon / doubling graphs represent bootstrap CIs for models’ performance over the task distribution. In other words, they are derived by re-calculating results for thousands of slight variations on (“resamplings” of) the task set, meaning e.g. dropping a couple tasks, doubling some tasks, increasing or decreasing the number of model runs on some tasks versus others (disclaimer: I’m not especially confident in the precise details of this description because the paper doesn’t go into much detail and my stats knowledge is pretty minimal). As a result they (the error bars) in no way represent or address uncertainty surrounding the human baseline estimates themselves, only that surrounding models’ performance on the task set, holding the human task completion times fixed.
So, bottom line, the situation is precisely as bad as you imply before walking things back on the assumption that the error bars somehow take care of the tiny/biased sample. To really underline this: the numbers you’re seeing on the y-axis are simply the result of averaging ~3 baseliners’ completion times per task. Also remember that these baseliners aren’t randomly sampled, but recruited from METR’s professional network. And that’s before getting into all the other sources of noise from the weird incentive scheme, anti-cheating protocols, and so on.
I also want to touch on a point you only raise implicitly but that others have harped on a lot. You cannot, under any circumstances, assume that the doubling times METR ends up with will hold for a much larger, properly randomized sample (this is what’s at issue in my exchange with Goutbeek). The counterexample I give in the article is a scenario wherein "tasks METR found took 2-3 hours ... turn out, in [a] larger sample, to take 1-1.5 hrs" but without deflating lower task durations by the same proportion. That would greatly compress the completion time distribution, and therefore change doubling times by quite a bit.
There are lots of reasons something like this could happen. For instance, maybe short tasks have much stabler time floors (since many don't require special expertise, and so can't be performed faster via superior understanding), but times for long tasks are sensitive to which/how many domain experts end up in your sample. In that case, changing samples from one with few domain experts to one with lots could substantially lower times at the higher end of the distribution without lowering times at the lower, or doing so a much smaller amount. That would in turn increase doubling times. In short, these sampling issues are a very big deal, and even independent of all my other concerns are very close to making the benchmark useless all by themselves.
I don’t see how you can brush this off as “fine,” nor do I accept that the issue is the same as “if there were 5000 engineers.” Of course it wouldn’t be; if you actually had a large, random sample, then you could read the METR graph more or less as intended (putting all the other issues aside), since the times on the y-axis would be representative of software engineers on the whole (possibly yielding very different doubling times), rather than those in a small, idiosyncratic sample.
With respect to the idea that “lots of different things measured lots of different ways show a similar pattern,” I disagree that the other benchmarks represented in the linked time horizon graph are “measured in lots of different ways.” On the contrary, they have many of the exact same issues as METR’s, and in some cases are worse. For instance: many of their time horizons are guesstimated by task annotators, rather than measured from human baselines. They also suffer from very similar task realism issues, and maybe most importantly the frontier labs are aggressively fine-tuning for all of them. So no, there are too many correlated sources of error here for these other benchmarks to speak to the robustness of METR’s.
In summary, I really do think the benchmark is as bad as my rhetoric implies, and insofar as it has misled a lot of people who have plausibly reallocated large quantities of time and labor on its basis, deserves a strong rebuke. Do I doubt that models are in fact getting better, esp. at software engineering? Not at all. But I am very uncertain as to how much better they’re getting and how fast they’re getting there. I would add that the slow rate of enterprise adoption and dearth of major macroeconomic impacts (growth, productivity, labor displacement, etc.) tells a very different story than the large volume of noisy benchmarks, METR's included.
I read your critiques with interest, but my main takeaway was thinking the main METR metric is probably about as good as you can expect.
With the exception of the quite valid N objection, most of your objections seemed a little too far on the "protest too much" side of things to me.
I'm assuming this benchmark is basically measuring "how close are we to being able to counterfeit a dev or white collar worker."
So methodologically, METR took highly qualified / capable people and had them do problems outside of their wheelhouse, so the human solvers needed to context up and learn a little.
That seems totally fair as a basis of comparison for an AI dev or white collar worker? That's literally the 'new project' baseline for any current white collar employee.
People with domain expertise did it a lot faster? That's fine too, and totally expected - I think we ALL expect that to fall, too, and should be a separate eval.
The biggest real problem is the scarcity of evals, the low N, and I agree they need better coverage there.
The messiness is a lesser problem, to my own mind, because obviously the way this is going to work is much like vibe coding today, where you have one human managing a fleet of AI white collar workers, and that human handles all the messiness while the AI's are still skilling up.
But per what I perceive the overall intent of the benchmark to be, an AI can step in and counterfeit work WELL before they need to be able to handle the messiness well.
Also, wouldn't they have to have been filtering their evals on messiness? Nobody is lovingly hand crafting artisanal benchmarks, obviously they need to be programatically evaluable?
And in terms of being self-consistent they need to be programatically evaluable, because if you hand craft OR hand-eval benchmarks with human judgment, you're not comparing like to like and it's not a true "same scale" differentiation of the models' capabilities.
Given the absurd amounts of money floating around in the AI industry, does no one have the money to do proper benchmarking on their effectiveness? Or is there simply no incentive?
There is an incentive for investors to do a proper benchmarking and there’s an incentive for those investors to never share the results of their benchmarking to the public, or at the very least, not release them until they feel fully invested. Even they did release them, how much would you trust the opinion of a person who just invested a billion dollars into something that what they release isn’t biased to drive up the price?
There's huge incentive in proper benchmarking AI effectiveness.
AI capabilities are (basically) causally driven by good benchmarks. If a lab manages to create a benchmark that's effective then it directly leads to a better model and a piece of the billions of $$$ in the industry.
I think the issue is that those with the money (i.e. frontier labs) don't have the incentive to lest more scientific benchmarks make models look worse.
Plenty of investors have more of a gambler's, than an engineer's or scientist's mentality. They want to roll the dice.
Why would it be counted as morally wrong?
I have a couple of possibilities. It's putting in more emotion than the topic deserves.
Or maybe someone hates the food because it's associated with a group they hate.
They kept including Ensure with my meals at a hospital, even though I asked them to stop. (There was also surprisingly good food.) It tasted vile, and I threw it hard at the trash can, which startled something I regret-- he was just standing nearby.
I don't think hating Ensure was immoral. I just surprised that there are people who like it.
It is morally wrong to hate. Having said that I really have a problem with ratatouille. Is that a type, or just an instance?
Nah, some things and some people deserve every bit of hate you can muster. Not sure ratatouille deserves one of those spots, though.
This post appears to have been edited after some of the replies.
Yes, it used to say "is it morally wrong to hate a certain type of food?"
I thought there was an earlier one about whether it's morally wrong to throw food in the trash.
Sigh, it was edited to that, and then edited back.
Thanks. It's nice to know I wasn't hallucinating.
Is this a stalking-horse question for "Should the burden of proof be on people claiming something is acceptable, not that something is unacceptable?"
Do you mean whether it is immoral to hate ANY specific food? No, certainly not. There exist foods I hate, which is conclusive disproof.
If you instead just want an example of a particular food it is morally wrong to hate, that's easy: Dr Pepper.
But what explains the MASSIVE Dr. Pepper/Diet Dr. Pepper taste differential? Other diet colas managed to figure this out.
This is an ancient question philosophers have been grappling with for generations. (cf. https://slatestarcodex.com/2017/04/07/yes-we-have-noticed-the-skulls/)
There is no consensus among the experts, but my personal belief is that it's the work of a malevolent demiurge.
Dr Pepper has not travelled outside the US, unlike Coke and Pepsi. There’s a reason for this.
You can get it in import shops in Europe, and I'm pretty sure the cans I buy are actually produced somewhere in Eastern Europe. None in stock for me to check right now though.
Dr. Pepper is the drink of intellectuals! (It's in Japan. Citing Steins Gate).
Worf would have liked it too, a proper warrior's drink (classic flavoring is prune).
That's not true. I remember drinking it in Russia as a kid some thirty years ago.
That's a fair point. I agree that your soul only gets consigned to the everburning flames of perdition if you've tasted the drink and then subsequently denied its greatness, and not if you've never had a reasonable opportunity to experience it.
Cannibalism is a kind of food that has near universal moral opposition to it, so there's at least one counter example.
Even so, there exist situations (and not just in theory) where cannibalism is perfectly acceptable, if not exactly cheered on.
It's possible for it to be morally wrong, if you're actively trying to remove a food from someone else's diet. "I think eating insects is gross, therefore we should stop foreign countries from being able to eat insects."
But obviously it's not morally wrong to refuse to eat a food you're allergic to.
If you're buying your own food then buy what you want. It's literally a matter of taste.
It's an arbitrary subjective preference that has no consequences, except potentially to the holder of the preference, how could it be wrong?
This is just an assertion of the correctness of your own moral framework, and begging the question: Why can an "arbitrary subjective preference that has no consequences" not be morally wrong?
It's more of an assertion of the superiority of common sense over the nonsense philosophers push out. "Here is one hand, here is another".
*as* a philosopher - if you're critiquing 'radical subjectivism' , or certain forms of analytic thought, I would agree. I argue the solution is an updated form of pragmatism: https://pragmaticfutures.substack.com/p/axioms-maxims-and-principles-of-avprag
And so pragmatically, consider potential scenarios that contradict the simple 'preferences have no consequences outside the holder':
- you could like to eat (e.g. novel) things that increase prob. you become sick, which has various levels of economic externalities
- you could like to eat things that increase prob. you get a communicable disease (recall cholera due to poor sanitation - was *tremendous* issue for most of human history)
- if you're deliberately eating foods considered 'taboo' by your community, your signaling something about yourself that may end up degrading coordination capacity
- 'honor culture' was historical default-> reputation isn't purely personal, it is carried 'by association' to kin+community.
- classic bayesian issue - the prob of someone with a true 'bad brain' manifesting unusual eating much higher than a normie + risk avoidance -> signal gets over-interpreted -> externalities via. reputation
This is not to say tradition cultures are *optimal* in these regards (see e.g. https://www.thepsmiths.com/p/review-sick-societies-by-robert-b)...but these were the ancestral env + resulting brain circuits for most humans.
I mostly agree with you, but as an intellectual exercise I can imagine a few possibilities depending on which ethical framework you use to analyze:
Utilitarian: Shared eating situations (dinner parties, going to restaurants together, banquets and other catered social occasions, etc) require some degree of coordination of food preferences, and sharply differing preferences lead to lower total utility than if you have the same preferences as everyone else in the group. Skipping occasions you would otherwise attend deprives you of the utility of attending and other attendees of the utility of your presence. Eating stuff you dislike has obvious disutility to you, as does skipping the meal (or the course) while still attending. Asking others to accommodate your preferences by tailoring the menu to you or providing multiple options imposes costs of them (either providing sub-optimal choices to others or requiring more planning, effort, and expense to provide parallel options). And self-segregating based on food preferences leads to group formation that is sub-optimal in other dimensions. If you could reasonably self-modify to not dislike the food that those around you like, that is arguably the ethical thing to do from a utilitarian perspective.
Dentological: Not sure. I'll leave this for others.
Aristotelian Virtue Ethics: Disliking healthy foods leads you towards Intemperance.
Objectivism: It is in the nature of humans to be omnivores. To dislike meat is to deny your nature.
Divine Command: Jesus told us to eat bread and drink wine in remembrance of him. If you don't like bread and wine, you don't like Jesus.
Bretty good, all are on point except divine command (think you're reaching there). Deontology seems tricky, but then, I feel you could make the virtue ethics case apply to it too, could be said you have an obligation to take care of yourself and eat healthy food (not sure why virtue ethics isn't a subset of deontology).
I never could quite wrap my mind around the internal logic of Deontology, even in areas where I think I have a pretty good handle on its prescriptions. Sometimes it feels like an extreme interpretation of Rule Utilitarianism and other times it feels like Virtue Ethics with extra steps.
Yes, I'm reaching on divine command. I was looking for religious dietary rules (ideally rooted in scripture) where a certain food was obligatory, but everything I could think of off the top of my head beside the Lord's Supper was a food being forbidden rather than mandatory.
The other candidates I came up with were the Catholic practice of eating fish on Fridays and the Jewish practice of eating matzo during Passover, but I think both of those are substitutions for prohibited food and eating the substitute is a cultural practice rather than a religious mandate. Catholicism forbids eating meat on Fridays during Lent, and before 1966 the rule was all Fridays year-round which is still observed by many individuals and communities. "Meat" is defined as coming from land animals, so seafood is traditionally substituted. I think the Jewish rule forbids eating risen bread during Passover, so matzo or another flatbread is traditionally substituted.
I'd point out that according to the latest, Catholicism suggests some sort of sacrifice on Fridays to promote virtue through the practice of sacrifice. Meat was given as a suggestion but it's been taken as an order for a long time. (How this could be mistaken for so many years is a bit beyond me, but that's what I'm hearing.)
Interesting. My knowledge on the subject comes mostly from my father, who was raised Catholic but has been non-practicing since before I was born. Supplemented a bit by stuff by a number of random bits that I recognized as part of the pattern. When I worked for Microsoft a decade or so ago, the cafeterias had a lot of seafood options on Fridays for the benefit of Catholics who kept the older practices. I've also encountered odd historical tidbits like the chronicles attributing the cause of death for King Henry I of England to "a surfeit of lampreys", with the backstory being that lampreys have an unusually beefy flavor profile for fish and thus were a popular luxury food for rich people to eat on Fridays, but were thought to be unhealthy due to something to do with the Four Humours theory of medicine.
None of these things, you may notice, inform me well of modern theology on the subject. So thank you for passing that along.
Fwiw, Exodus 12 commands they eat unleavened bread (as well as forbidding anything with yeast). That should count.
Yes, it does seem to do so, and it looks like there are a number of Jewish religious websites that talk about eating Matzo being theologically mandatory at certain points during Passover, although opinion seems to vary a bit as to what those points are.
Your comment is top-level, but you probably meant to ask Brad.
Indeed, thanks