Ben Pace

• Ben PaceMay 20, 2025, 6:39 AM

  4 points

  0

  in reply to: Veedrac’s comment on: leogao’s Shortform

  But not in full generality! This is a fine question to raise in this context, but in general the correct thing to do in basically all situations is to consider the object level, and then also let yourself notice if people are unusually insane around a subject, or insane for a particular reason. Sometimes that is the decisive factor, but for all questions, the best first pass is to think about how that part of the world works, rather than to think about the other monkeys who have talked about it in the past.

• Ben PaceMay 20, 2025, 6:38 AM

  2 points

  0

  in reply to: leogao’s comment on: leogao’s Shortform

  ou’re I am confused why you think my claims are only semi related. to me my claim is very straightforward, and the things i’m saying are straightforwardly converying a world model that seems to me to explain why i believe my claim. i’m trying to explain in good faith, not trying to say random things. i’m claiming a theory of how people parse information, to justify my opening statement,

  Thank you for all this. I still think your quick take is wrong on the matter of epistemology.

  I acknowledge that you make a fine point about persuasion, that someone who is primarily running the heuristic that “claims about the end of the world are probably crack-pots or scammers” will not be persuaded by someone arguing that actually 20:1 against and 20:1 in favor of a claim are equally extreme beliefs.

  A version of the quick take that I would’ve felt was just fine would read:

  Some people have basically only heard claims of human extinction coming from crackpots and scammers, and will not have thought much about the AI extinction idea on the object level. To them, this sort of argument I’ve discussed is unpersuasive at moving beyond the “is this a crackpot/​scam” part of the dialogue. In this quick take I’ll outline my model of how they’re thinking about it, and give recommendations for how you should argue instead.

  But your quick take doesn’t confine itself to discussing those people in those situations. It flatly says it’s true as a matter of epistemology that you should “use bigness of claim as a heuristic for how much evidence you need before you’re satisfied”, that you should “use reference classes that have consistently made good decisions irl” and that the crackpots/​scammers one is the correct one to use here otherwise you’ll risk “getting pwned ideologically”.

  These aren’t always the right heuristics (e.g. on this issue they are not for you and for me) and you shouldn’t say that they are just so that some people on Twitter will stop using rhetoric that isn’t working.

  I believe you’re trying to do your best to empathize with people who are unpersuaded by an unsuccessful rhetorical move, a move that people who believe your position are making in public discourse. That is commendable. I think you are attempting to cause other people who hold your position to stop using that rhetorical move, by telling them off for using it, but to acheive this aim you are repeatedly saying the people who do not hold your position are doing normatively correct epistemology, and you’re justifying it with Occam’s razor and reference class forecasting, and this is all wrong. In some situations for some people it is reasonable to primarily use theses heuristics, and in other situations for other people it is not. I’m not arguing that the people unpersuaded are being unreasonable, but (for example) your opening sentence makes fully-general statements about how to reason about this issue that I believe are false. Rule number of one of good discourse: don’t make false statements about epistemology in order to win an object level point.

  Yep, seems fine to drop this here; I make no bid of you to reply further.

Lighthaven Se­quences Read­ing Group #35 (Tues­day 5/​20)

• Ben PaceMay 19, 2025, 1:35 AM

  2 points

  0

  in reply to: leogao’s comment on: leogao’s Shortform

  You’re writing lots of things here but as far as I can tell you aren’t defending your opening statement, which I believe is mistaken.

  I claim it is a lot more reasonable to use the reference class of “people claiming the end of the world” than “more powerful intelligences emerging and competing with less intelligent beings” when thinking about AI x-risk. further, we should not try to convince people to adopt the latter reference class—this sets off alarm bells, and rightly so (as I will argue in short order) - but rather to bite the bullet, start from the former reference class, and provide arguments and evidence for why this case is different from all the other cases.

  Firstly, it’s just not more reasonable. When you ask yourself “Is a machine learning run going to lead to human extinction?” you should not first say “How trustworthy are people who have historically claimed the world is ending?”, you should of course primarily bring your attention to questions about what sorts of machine is being built, what sort of thinking capacities it has, what sorts of actions it can take in the world, what sorts of optimization it runs, how it would behave around humans if it were more powerful than them, and so on. We can go back to discussing epistemology 101 if need be (e.g. “Hug the Query!”).

  Secondly, insofar as someone believes you are a huckster or a crackpot, you should leave the conversation, communication here has broken down and you should look for other communication opportunities. However, insofar as someone is only evaluating this tentatively as one of many possible hypotheses about you then you should open yourself up to auditing /​ questioning by them about why you believe what you believe and your past history and your memetic influences. Being frank is the only way through this! But you shouldn’t say to them “Actually, I think you should treat me like a huckster/​scammer/​serf-of-a-corrupt-empire.” This feels analogous to a man on a date with a woman saying “Actually I think you should strongly privilege the hypothesis that I am willing to rape you, and now I’ll try to provide evidence for you that this is not true.” It would be genuinely a bad sign about a man that he thinks that about himself, and also he has moved the situation into a much more adversarial frame.

  I suspect you could write some more narrow quick-take such as “Here is some communication advice I find helpful when talking with friends and colleagues about how AI can lead to human extinction”, but in generalizing it all the way to making dictates about basic epistemology you are making basic mistakes and getting it wrong.

  Please either (1) defend and/​or clarify the original statement, or (2) concede that it was mistaken, rather than writing more semi-related paragraphs about memetic immune systems.

• Ben PaceMay 19, 2025, 1:09 AM

  2 points

  0

  in reply to: Veedrac’s comment on: leogao’s Shortform

  Thanks for the comment. (Upvoted.)

  a. I expect there is a slightly more complicated relationship between my value-function and the likely configuration states of the universe than literally zero-correlation, but most configuration states do not support life and we are all dead, so in one sense a claim that in the future something very big and bad will happen is far more likely on priors. One might counter that we live in a highly optimized society where things being functional and maintained is an equilibrium state and it’s unlikely for systems to get out of whack enough for bad things to happen. But taking this straightforwardly is extremely naive, tons of bad things happen all the time to people. I’m not sure whether to focus on ‘big’ or ‘bad’ but either way, the human sense of these is not what the physical universe is made out of or cares about, and so this looks like an unproductive heuristic to me.

  b. On the other hand, I suspect the bigger claims are more worth investing time to find out if they’re true! All of this seems too coarse-grained to produce a strong baseline belief about big claims or small claims.

  c. I don’t get this one. I’m pretty sure I said that if you believe that you’re in a highly adversarial epistemic environment, then you should become more distrusting of evidence about memetically fit claims.

  I don’t know what true points you think Leo is making about “the reference class”, nor which points you think I’m inaccurately pushing back on that are true about “the reference class” but not true of me. Going with the standard rationalist advice, I encourage everyone to taboo “reference class” and replace it with a specific heuristic. It seems to me that “reference class” is pretending that these groupings are more well-defined than they are.

• Ben PaceMay 18, 2025, 10:28 PM

  7 points

  3

  in reply to: leogao’s comment on: leogao’s Shortform

  Your points about Occam’s razor have got nothing to do with this subject^[1]. The heuristic “be more skeptical of claims that would have big implications if true” makes sense only when you suspect a claim may have been adversarially optimized for memetic fitness; it is not otherwise true that “a claim that something really bad is going to happen is fundamentally less likely to be true than other claims”.

  I’m having a little trouble connecting your various points back to your opening paragraph, which is the primary thing that I am trying to push back on.^[2]

  I claim it is a lot more reasonable to use the reference class of “people claiming the end of the world” than “more powerful intelligences emerging and competing with less intelligent beings” when thinking about AI x-risk. further, we should not try to convince people to adopt the latter reference class—this sets off alarm bells, and rightly so (as I will argue in short order) - but rather to bite the bullet, start from the former reference class, and provide arguments and evidence for why this case is different from all the other cases.

  To restate the message I’m reading here: “Give up on having a conversation where you evaluate the evidence alongside your interlocutors. Instead frame yourself as trying to convince them of something, and assume that they are correct to treat your communications as though you are adversarially optimizing for them believing whatever you want them to believe.” This assumption seems to give up a lot of my ability to communicate with people (almost ~all of it), and I refuse to simply do it because some amount of communication in the world is adversarially optimized, and I’m definitely not going to do it because of a spurious argument that Occam’s razor implies that “claims about things being really bad or claims that imply you need to take action are fundamentally less likely to be true”.

  You are often in an environment where people are trying to use language to describe reality, and in that situation the primary thing to evaluate is not the “bigness” of a claim, but the evidence for and against it. I recommend instead to act in such a way as to increase the size and occurrence of that environment more-so than “act as though it’s correct to expect maximum adversarial optimization in communications”.

  (Meta: The only literal quotes of Leo’s in this comment are the big one in the quote block, my use of “” is to hold a sentence as object, they are not things Leo wrote.)

  1. ^

     I agree that the more strongly a claim implies that you should take action, then the more you should consider that it is being optimized adversarially for you to take action. For what it’s worth, I think that heuristic applies more so to claims that you should personally take action. Most people have little action to directly prevent the end of the world from AI; this is a heuristic more naturally applied to claims that you need to pay fines (which are often scams/​spam). But mostly, when people give me claims that imply action, they are honestly meant claims and I do the action. This is the vast majority of my experience.

  2. ^

     Aside to Leo: Rather than reply point-by-point to the each of the paragraphs in the second comment, I will try restating and responding to the core message I got in the opening paragraph of the first comment. I’m doing this because the paragraphs in the second-comment seemed somewhat distantly related /​ I couldn’t tell whether the points were actually cruxy. They were responding to many different things, and I hope restating the core thing will better respond to your core point. However I don’t mean to avoid key arguments, if you think I have done so feel free to tell me one or two paragraphs you would especially like me to engage with and I will do so in any future reply.

• Ben PaceMay 18, 2025, 1:47 AM

  37 points

  1

  in reply to: leogao’s comment on: leogao’s Shortform

  This all seems wrongheaded to me.

  I endeavor to look at how things work and describe them accurately. Similarly to how I try to describe how a piece of code works, or how to to build a shed, I will try to accurately describe the consequences of large machine learning runs, which can include human extinction.

  I personally think AGI will probably kill everyone. but this is a big claim and should be treated as such.

  This isn’t how I think about things. Reality is what exists, and if a claim accurately describes reality, then I should not want to hold it to higher standards than claims that do not describe reality. I don’t think it’s a good epistemology to rank claims by “bigness” and then say that the big ones are less likely and need more evidence. On the contrary, I think it’s worth investing more in finding out if they’re right, and generally worth bringing them up to consideration with less evidence than for “small” claims.

  on the other hand, everyone has personally experienced a dozen different doomsday predictions. whether that’s your local church or faraway cult warning about Armageddon, or Y2K, or global financial collapse in 2008, or the maximally alarmist climate people, or nuclear winter, or peak oil. for basically all of them, the right action empirically in retrospect was to not think too much about it.

  I don’t have the experiences you’re describing. I don’t go to churches, I don’t visit cults, I was 3yrs old in the year 2000, I was 11 for the ’08 financial crash and having read about it as an adult I don’t recall extinction being a topic of discussion, I think I have heard of climate people saying that via alarmist news headlines but I have not had anyone personally try to convince me of this or even say that they believe it. I have heard it discussed for nuclear winter, yes, and I think nukes are quite scary and it was reasonable to consider, I did not dismiss it out of hand and wouldn’t use that heuristic. I don’t know what the oil thing is.

  In other words, I don’t recall anyone seriously trying to convince me that the world was ending except in cases where they had good reason to believe it. In my life, when people try to warn me about big things, especially if they’ve given it serious thought, usually I’ve found it’s been worthwhile for me to consider it. (I like to think I am good at steering clear of scammers and cranks, so that I can trust the people in my life when they tell me things.)

  The sense I get from this post is that, in it, you’re assuming everyone else in the world is constantly being assaulted with claims meant to scare and control them rather than people attempting to describe the world accurately. I agree there are forces doing that, but I think this post gives up all too quickly on there being other forces in the world that aren’t doing that that people can recognize and trust.

• Ben PaceMay 15, 2025, 5:10 PM

  2 points

  0

  on: From Com­ments on Ac­countabil­ity Sinks

  Oops, I didn’t send my reply comment. I’ve just posted it, yes, that information did change my mind about this case.

• Ben PaceMay 15, 2025, 5:09 PM

  3 points

  0

  in reply to: Martin Sustrik’s comment on: Ac­countabil­ity Sinks

  Thank you for the details! I change my mind about the locus of responsibility, and don’t think Wascher seems as directly culpable as before. I don’t update my heuristic, I still think there should be legal consequences for decisions that cause human deaths,

  My new guess is that something more like “the airport” should be held accountable and fined some substantial amount of money for the deaths, to go to the victim’s families.
  
  Having looked into it a little more I see they were sued substantially for these, so it sounds like that broadly happened.

Lighthaven Se­quences Read­ing Group #34 (Tues­day 5/​13)

• Ben PaceMay 5, 2025, 7:26 AM

  6 points

  0

  on: Ac­countabil­ity Sinks

  I liked reading these examples; I wanted to say, it initially seemed to me a mistake not to punish Wascher, whose mistake led to the death of 35 people.

  I have a weak heuristic that, when you want enforce rules, costs and benefits aren’t fungible. You do want to reward Wascher’s honesty, but I still think that if you accidentally cause 35 people to die this is evidence that you are bad at your job, and separately it is very important to disincentivize that behavior for others who might be more likely to make that mistake recklessly. There must be a reliable punishment for that kind of terrible mistake.

  So you must fire her and bar her from this profession, or fine her half a year’s wages, or something. If you also wish to help her, you should invest in supporting her get into a new line of work with which she can support her family, or something. You can even make her net better off for having helped uncover a critical mistake and saving future lives. But people should know that there was a cost and there will be if they do so in future.

  Or at least this is what my weak heuristic says.

  What links here?

  • From Com­ments on Ac­countabil­ity Sinks by Martin Sustrik (May 15, 2025, 10:20 AM; 15 points)

• Ben PaceMay 1, 2025, 9:42 PM

  6 points

  9

  in reply to: Casey_’s comment on: “The Ur­gency of In­ter­pretabil­ity” (Dario Amodei)

  I don’t think that propaganda must necessarily involve lying. By “propaganda,” I mean aggressively spreading information or communication because it is politically convenient /​ useful for you, regardless of its truth (though propaganda is sometimes untrue, of course).

  When a government puts up posters saying “Your country needs YOU” this is intended to evoke a sense of duty and a sense of glory to be had; sometimes this sense of duty is appropriate, but sometimes your country wants you to participate in terrible wars for bad reasons. The government is saying it loudly because for them it’s convenient for you to think that way, and that’s not particularly correlated with the war being righteous or with the people who decided to make such posters even having thought much about that question. They’re saying it to win a war, not to inform their populace, and that’s why it’s propaganda.

  Returning to the Amodei blogpost: I’ll happily concede that you don’t always need to give reasons for your beliefs when expressing them—context matters. But in every context—tweets, podcasts, ads, or official blogposts—there’s a difference between sharing something to inform and sharing it to push a party line.

  I claim that many people have asked why Anthropic believes it’s ethical for them to speed up AI progress (by contributing to the competitive race), and Anthropic have rarely-if-ever given a justification of it. Senior staff keep indicating that not building AGI is not on the table, yet they rarely-if-ever show up to engage with criticism or to give justifications for this in public discourse. This is a key reason why it reads to me as propaganda, because it’s an incredibly convenient belief for them and they state it as though any other position is untenable, without argument and without acknowledging or engaging with the position that it is ethically wrong to speed up the development of a technology they believe has a 10-20% chance of causing human extinction (or a similarly bad outcome).

  I wish that they would just come out, lay out the considerations for and against building a frontier lab that is competing to reach the finish line first, acknowledge other perspectives and counterarguments, and explain why they made the decision they have made. This would do wonders for the ability to trust them.

  (Relatedly, I don’t believe the Machines of Loving Grace essay is defending the position that speeding up AI is good; the piece in fact explicitly says it will not assess the risks of AI. Here are my comments at the time on that essay also being propaganda.)

• Ben PaceMay 1, 2025, 6:58 PM

  4 points

  2

  in reply to: Casey_’s comment on: “The Ur­gency of In­ter­pretabil­ity” (Dario Amodei)

  I’m saying that he is presenting it as something he believes from his place of expertise and private knowledge without argument, because it is something that is exceedingly morally and financially beneficial to him (he gets to make massive money and not be a moral monster), rather than because he has any evidence, and he stated it without evidence.

  It is a similar sentence to if a President of a country who just initiated war said “If there’s one thing I’ve learned in my life it’s that war is inevitable, and there’s just a question of who wins and how to make sure it’s over quickly”, in a way that means they should be absolved of responsibility for initiating war.

  Edit: Just as Casey B was writing his reply below, I edited an example out of Mark Zuckerberg saying something like “If there’s one thing I’ve learned in my career, it’s that social media is good, and the only choice is which sort of good social media we have”. Leaving this note so that ppl aren’t confused by his reply.

• Ben PaceMay 1, 2025, 1:17 AM

  4 points

  0

  in reply to: Wei Dai’s comment on: Wei Dai’s Shortform

  Can you expand on this? How can you tell the difference, and does it make much of a difference in the end (e.g., if most people get corrupted by power regardless of initial intentions)?

  But I don’t believe most people get corrupted by power regardless of initial intentions? I don’t think Francis Bacon was corrupted by power, I don’t think James Watt was corrupted by power, I don’t think Stanislav Petrov was corrupted by power, and all of these people had far greater influence over the world than most people who are “corrupted by power”.

  I’m hearing you’d be interested in me saying more words about the difference in what it looks like to be motivated by responsibility versus power-seeking. I’ll say some words, can see if they help.

  • I think someone motivated by responsibility often will end up looking more aligned with their earlier self over time even as they grow and change, will often not accept opportunities for a lot of power/​prestige/​money because they’re uninteresting to them, will often make sacrifices of power/​prestige for ethical reasons, will pursue a problem they care about long after most would give up or think it likely to be solved.

  • I think someone primarily seeking power will be much more willing to do things that pollute the commons or break credit-allocation mechanisms to get credit, and generally game a lot of systems that other people are earnestly rising through. They will more readily pivot on what issue they say they care about or are working on because they’re not attached to the problem, but to the reward for solving the problem, and many rewards can be gotten from lots of different problems. They’ll be more guided by what’s fashionable right now, and more attuned to it. They’ll maneuver themselves in order to be able to politically work with whoever has power that they want, regardless of the ethics/​competence/​corruption of those people.

  > As a background model, I think if someone wants to take responsibility for some part of the world going well, by-default this does not look like “situating themselves in the center of legible power”.

  And yet, Eliezer, the writer of “heroic responsibility” is also the original proponent of “build a Friendly AI to take over the world and make it safe”.

  Building an AGI doesn’t seem to me like a very legible mechanism of power, or at least it didn’t in the era Eliezer pursued it (where it wasn’t also credibly “a path to making billions of dollars and getting incredible prestige”). The word ‘legible’ was doing a lot of work in the sentence I wrote.

  Another framing I sometimes look through (H/​T Habryka) is constrained vs unconstrained power. Having a billion dollars is unconstrained power, because you can use it to do a lot of different things – buy loads of different companies or resources. Being an engineer overlooking missile-defense systems in the USSR is very constrained, you have an extremely well-specified set of things you can control. This changes the adversarial forces on you, because in the former case a lot of people stand to gain a lot of different possible things they want if they can get leverage over you, and they have to be concerned about a lot of different ways you could be playing them. So the pressures for insanity are higher. Paths that give you the ability to influence very specific things that route through very constrained powers are less insanity-inducing, I think, and I think most routes that look like “build a novel invention in a way that isn’t getting you lots of money/​status along the way” are less insanity-inducing, and I rarely find the person to have become as insane as some of the tech-company CEOs have. I also think people motivated by taking responsibility for fixing a particular problem in the world are more likely to take constrained power, because… they aren’t particularly motivated by all the other power they might be able to get.

  I don’t suspect I addressed your cruxes here so far about whether this idea of heroic responsibility is/​isn’t predictably misused. I’m willing to try again if you wish, or if you can try pointing again to what you’d guess I’m missing.

Lighthaven Se­quences Read­ing Group #33 (Tues­day 5/​6)

• Ben PaceApr 29, 2025, 5:33 AM

  10 points

  1

  in reply to: Knight Lee’s comment on: “The Ur­gency of In­ter­pretabil­ity” (Dario Amodei)

  Not sure I get your overall position. But I don’t believe all humans are delusional about the most important questions in their lives. See here for an analysis of pressures on people that can cause them to be insane on a topic. I think you can create inverse pressures in yourself, and you can also have no pressures and simply use curiosity and truth-seeking heuristics. It’s not magic to not be delusional. It just requires doing the same sorts of cognition you use to fix a kitchen sink.

• Ben PaceApr 29, 2025, 12:59 AM

  11 points

  5

  in reply to: Knight Lee’s comment on: “The Ur­gency of In­ter­pretabil­ity” (Dario Amodei)

  Not only would most people be hopelessly lost on these questions (“Should I give up millions-of-dollars-and-personal-glory and then still probably die just because it is morally right to do so?”), they have also picked up something that they cannot put down. These companies have 1,000s of people making millions of dollars, and they will reform in another shape if the current structure is broken apart. If we want to put down what has been picked up more stably, we must use other forces that do not wholly arise from within the companies.

• Ben PaceApr 28, 2025, 11:37 PM

  51 points

  20

  in reply to: Wei Dai’s comment on: Wei Dai’s Shortform

  My sense is that most of the people with lots of power are not taking heroic responsibility for the world. I think that Amodei and Altman intend to achieve global power and influence but this is not the same as taking global responsibility. I think, especially for Altman, the desire for power comes first relative to responsibility. My (weak) impression is that Hassabis has less will-to-power than the others, and that Musk has historically been much closer to having responsibility be primary.

  I don’t really understand this post as doing something other than asking “on the margin are we happy or sad about present large-scale action” and then saying that the background culture should correspondingly praise or punish large-scale action. Which is maybe reasonable, but alternatively too high level of a gloss. As per the usual idea of rationality, I think whether you are capable of taking large-scale action in a healthy way is true in some worlds and not in others, and you should try to figure out which world you’re in.

  The financial incentives around AI development are blatantly insanity-inducing on the topic and anyone should’ve been able to guess that going in, I don’t think this was a difficult question. Though I guess someone already exceedingly wealthy (i.e. already having $1B or $10B) could have unusually strong reason to not be concerned about that particular incentive (and I think it is the case Musk has seemed differently insane than the others taking action in this area, and lacking in some of the insanities).

  However I think most moves around wielding this level of industry should be construed as building an egregore more powerful than you. The founders/​CEOs of the AI big-tech companies are not able to simple turn their companies off, nor their industry. If they grow to believe their companies are bad for the world, either they’ll need to spend many years dismantling /​ redirecting them, or else they’ll simply quit/​move on and some other person will take their place. So it’s still default-irresponsible even if you believe you can maintain personal sanity.

  Overall I think taking responsibility for things is awesome and I wish people were doing more of it and trying harder. And I wish people took ultimate responsibility for as big of a thing they can muster. This is not the same as “trying to pull the biggest lever you can” or “reaching for power on a global level”, those are quite different heuristics; grabbing power can obviously just cost you sanity, and often those pulling the biggest lever they can are doing so foolishly.

  As a background model, I think if someone wants to take responsibility for some part of the world going well, by-default this does not look like “situating themselves in the center of legible power”. Lonely scientist/​inventor James Watt spent his early years fighting poverty before successfully inventing better steam engines, and had far more influence by helping cause the industrial revolution than most anyone in government did during his era. I think confusing “moving toward legible power” for “having influence over the world” is one of the easiest kinds of insanity.

• Ben PaceApr 28, 2025, 9:05 PM

  13 points

  10

  in reply to: Knight Lee’s comment on: “The Ur­gency of In­ter­pretabil­ity” (Dario Amodei)

  I think that Anthropic is doing some neat alignment and control work, but it is also the company most effectively incentivizing people who care about existential risk to sell out, to endorse propaganda, silence themselves, and get on board with the financial incentives of massive monetization and capabilities progress. In this way I see it as doing more damage than OpenAI (though OpenAI used to have this mantle pre-Anthropic, while the Amodei siblings were there and with Christiano as researcher and Karnofsky on the board).

  I don’t really know the relative numbers, in my mind the uncertainty I have spans orders of magnitude. The numbers are all negative.

• Ben PaceApr 27, 2025, 4:53 AM

  15 points

  23

  on: “The Ur­gency of In­ter­pretabil­ity” (Dario Amodei)

  I couldn’t get two sentences in without hitting propaganda, so I set it aside. But I’m sure it’s of great political relevance.

  What links here?

  • AI #115: The Evil Ap­pli­ca­tions Division by Zvi (May 8, 2025, 1:40 PM; 31 points)