epistemology enthusiast
Zack_M_Davis
Karma: 20,329
in order to help people understand your perspective
I think this is inappropriately applying mistake theory to what is actually a conflict. If the problem was that people just didn’t understand the perspective expressed in one of Dai’s or Achmiz’s or my comments because it wasn’t expressed clearly enough, I think the natural solution would be to reply, “I don’t understand why you think this comment is relevant, can you spell it out for me?” I’m sure Wei would be happy to elaborate, as would I, as was Achmiz when he was still permitted to speak.
The point of the hobby-horse meme and the persecution of Achmiz is that we have things to say that you don’t want others to hear, because letting our perspectives be heard on the website we’ve been using for 17 years interferes with authors being able to control discussion of their own work. That’s a conflict. Maybe you think you can serve your side in the conflict by pretending that our side is making a mistake by not writing enough posts explaining our point of view? That trick worked to let you purge Achmiz without (yet) triggering a fatal loss-of-legitimacy for the whole site, but I don’t think it’s going to work against me (9x Curated, 4x Best of Less Wrong, 3x Less Online invited author) or Dai (7x Curated, 3x Best of Less Wrong, 3x Less Online invited author).
It’s likely that you think I’m mischaracterizing you, so I should probably make the standard disclaimer here that I’m speaking in functionalist terms, looking for the simplest explanations that I think predict people’s behavior, even when that contradicts their self-report. I’m not saying you self-identify as being in a conflict.
Rather, I’m looking at the comment you just wrote recommending that Dai write more posts explaining his perspective and making a judgment call that it’s sufficiently “unserious” in an important sense such that it makes sense to suppose (as a high-probability but by no means certain hypothesis) that you’re motivatedly refusing to see the conflict you’re engaged in (again in a functionalist sense).
Specifically, Dai already writes a lot of posts explaining his perspective. Telling him to do more of that is obviously not going to help. If a dishonest person who favored authors being able to control discussions was consciously aware of being in a conflict about that and wanted to strategically undermine Dai’s position, they might insincerely suggest that Dai expend even more labor writing the kinds of posts he already does in the hopes of deceiving third parties unfamiliar with the full record into thinking that that any censorship or delegitimization Dai faces is his own fault for not being clear enough.
A big part of the reason I think it’s important to be able to talk about functionally implied conflicts is that we don’t want to create incentives for self-deception. If it would be problematic for a dishonest person to insincerely suggest that Dai needs to do more of the thing he already does in order for his complaints to have standing, then it should also be problematic (maybe not as problematic, but problematic to some nonzero degree) if someone performs the exact same speech act without being consciously insincere.
I think that some people think that we shouldn’t talk about functionally implied conflicts on the grounds that the accusation is unfalsifiable: if I’m questioning your motives, how could you possibly defend yourself except with self-reports that I’ll also doubt? The reason I disagree with this is because accusations of hidden motives aren’t fabricated from nothing: they’re inferred from contradictions between unobservable stated motives and behavior—which, crucially, is observable. If you think I’m reading this whole situation wrong when I claim that there’s a functional conflict between Dai et al. and the mod team about whether authors should be able to control discussions, that’s something you can argue on the merits and potentially embarrass me in the eyes of third parties!
reduce my engagement with LW/rationalists, and spend more of my time elsewhere
Where do you have in mind? I’ve been committed to a “stay and fight” policy out of a perceived lack of alternatives. (I have my whitelist of personal friends who I trust to be saner than the “rationalist” center of memetic gravity, but there doesn’t seem to be an alternative scene with critical mass for the type of work we do.)
pretty defensive, and a mistake for that reason (as was Zack Davis’ original reply upthread).
What do you think would have been the superior play for me in that situation? When someone is deploying a “distasteful” epithet without specifics, I think asking for specifics (forcing them to take a stand on which side of the fact/value boundary the distaste is supposedly on) is the natural follow-up. (And then after he answered that fact-beliefs themselves aren’t culpable, there was nothing to do except thank him for answering the question.)
I think the idea is that we shouldn’t have to indiscriminately divest from the whole industry when we could “just” build fiction-writing AIs and not catastrophically risky AIs (as you suggested about Gwern’s company).
If one is an altruist (if you are purely self-interested there are different calculations of the risk), this seems to me to be fallacious and counter productive in the case of fossil fuels as well as AI.
Does your altruism take into account benefits from economic growth? I think the people whose behavior you’re puzzled by have an attitude more like “AI and fossil fuels are great except for the potential for catastrophic risk, which demands targeted mitigations” (as contrasted to “AI and fossil fuels are Bad in general; only use if truly unavoidable”).
But presumably you would care on the assumption that if Cowen had considered the question a good amount and concluded that what you are doing is super harmful, it would be because he came up with a more substantive objection than “it all feels a bit cultish”? You’ve already taken into account that some people think it all feels a bit cultish.
I’m saying that Shulman already took arguments against investing in AI capabilities into account before he decided to work for Situational Awareness. Yudkowsky’s disapproval would only matter to him insofar as it reflected objections that he hadn’t already “priced in.” Sorry if that wasn’t already clear due to my inexact analogy.
This is bad propaganda because it derives all of its persuasive force from the substitution of “demons” for AI, where people already know that “demons” are supposed to be bad; the reader is encouraged to sneer at the stupidity and wickedness of AI developers for working with the mythical incarnation of the concept of evil itself. But one of the main reasons the world is in so much danger from AI is precisely because the danger isn’t that obvious.
Imagine you’d never heard of “demons” before. If, as in this story, demons liked to “[k]ill, maim, and slaughter with squealing glee and ecstatic delight” and the degree of human control over demons was “50/50” “[o]n a good day”, then, indeed, those who summon demons would be bad people—and it also wouldn’t be very hard to get an international treaty banning demon-summoning. (No one likes killing, maiming and slaughtering with ecsatic delight. We banned CFCs for much less.)
If, as with real-world AI, millions of people had lots of everyday experience with demons routinely obeying human commands to answer questions, summarize documents, write code, and generate images, the case for banning demon-summoning altogether would be an extremely hard sell. Most of the people who understood some of the theoretical arguments for why demons could be dangerous would rather work on demon safety (and increase investments in demon safety and security efforts in response to the occasional rogue demon accident) than just give up on something so useful that had already become part of their lives. If people who thought demon safety research was doomed despaired of winning the technical argument on the merits with Society’s demon experts, they might switch messaging strategies to trying to demonize demons in the eyes of policymakers and the general public, but fighting the economic incentives would be an uphill battle.
I think readers intrigued by the premise of this story should save themselves the click and instead read “Love Stays Loved”, which explores the analogy between AI and the occult with serious literary merit rather than as cheap satire.
I hope Eliezer has let Shulman know that he’s lost a lot of respect in Eliezer’s eyes and done terrible things
You think Carl doesn’t already know that?
I’m not sure you’re modeling how uncompelling moral disapproval is to people who have their own considered views on a topic (as Carl does on the machine intelligence transition and how to intervene on it) rather than deferring to the same authorities that you do. Imagine if Tyler Cowen told you that by choosing to work at Lightcone Infrastructure, you’ve lost a lot of respect in his eyes and done terrible things. Not convincing, right?
If others are interested, I would happily share my takeaways at greater length
Please do! (I find your students’ perspective alien and would be interested to read more.)
I’ve been following the NorCal case closely. (There’s a separate case in the D.C. Circuit under a different statute.)
My impression as a non-expert is that the impact of adding a single high-profile name to an amicus curiae brief is negligible. The amicus brief in question was organized by a group called Protect Democracy. One can only assume that if Turner hadn’t asked Dean, the same document would be prepared but without Dean’s name on it. Judge Lin’s preliminary injunction order in the NorCal case collectively acknowledges the various pro-Anthropic amici a few times (“Several amicus briefs detail the chilling effect”, “Several amicus briefs support this conclusion”), but there’s no indication that she thought anything like, “Well, if Jeff Dean thinks the government is wrong …” (as of course she shouldn’t!).
I find some of Richard’s attitudes very distasteful
Examples? What attitudes, specifically? Can you link to something Ngo wrote that expresses the attitude you find distasteful? (Same question to any of the agree-voters.)
I feel like this should be easy. (I think that when I find someone’s attitudes distasteful, I usually don’t have a problem retrieving a link; I need the links to write a convincing critical blog post explaining why the views are bad, as I so often do.)
Taking a wild guess (I don’t know if this is what you had in mind), if the distasteful attitude is that he thinks it’s plausible that there are ancestry-cluster (“racial”) differences in socially-relevant traits like cognitive ability (mentioned in footnote 2 of the OP), I think that’s more of a “hypothesis” than an “attitude”. (It’s a claim about the world that could be true or false, not a claim about how we should behave or feel.)
If the distasteful attitude is that he thinks scholars should be permitted to reason in public about decision-relevant consequences of that hypothesis while also getting paid to do decision theory research, I can see reasons why someone would find that attitude distasteful (see my discussion of “the Schelling point for preventing group conflicts” in a 2020 book review), but to the extent that those reasons aren’t about the truth or falsity of the hypothesis, I think it’s important to notice the rationality implications of silencing discussion of the implications of a hypothesis for reasons other than its truth.
But it also seems like rationalists and EAs do better at this than almost any similarly sized subculture?
What is the relevance of this? It is written of the eighth virtue of humility that “it is useless to be superior [...] There is no guarantee that adequacy is possible given your hardest effort; therefore spare no thought for whether others are doing worse.”
For the fraction of our conversation that was frustrating due to my grammatical error, and unclear communication in that comment, as well as further frustration caused by me not noticing that what I wrote was unclear earlier in this conversation, I apologize.
Thanks, and I appreciate it, but unfortunately, this does not fully resolve my concerns about the degree to which you are trustworthy. Let me explain—and thanks for your patience. (I thought the present comment would benefit from some focused, cool-headed care and attention, which it took me some days to get around to.)
How I’m Thinking About Discouragement Claims
I agree that in as much you interpreted my statement as “these authors have all stopped posting on the site because of you”, then of course my statement is obviously blatantly wrong. But I don’t understand how that hypothesis would even be available
Indeed, I didn’t interpret it that way; I agree that it’s possible to be discouraged but still use the site.
“Being discouraged” is a much weaker proposition than “has mentioned being annoyed by specifically that user to the head-admin”
Is it? I think “discouraged from using the site in some non-trivial way” and “annoyed by” are different things, because a lot of annoyances are trivial. The question was expressing skepticism (which I share) that finding a commenter annoying would be a legitimately non-trivial barrier to participating; it’s normal and expected that not everyone on a public forum is going to be your best friend. [1]
I’m worried about a potential motte-and-bailey pattern where the reality in the motte is “Habryka has a vague memory that someone either made or agreed to some sort of negative-valence statement about Achmiz, at some time, in some context” and what gets reported in the bailey—without any details of what the person allegedly said—is that that person was discouraged from using the site by Achmiz specifically, such that they get cited as an example of “authors who find [a commenter’s] very presence in a discussion so ‘unpleasant’ that … it’s enough to discourage them from posting on LW altogether”.
To be clear, I think we do have better examples of “authors who find [Achmiz’s] very presence in a discussion so ‘unpleasant’ that … it’s enough to discourage them from posting on LW altogether”, because we have direct statements from them in their own words (which I of course quoted in §III.2 of my post): for instance, Duncan Sabien (“It’s not on LessWrong because of you, specifically”), DirectedEvolution (“one of three people who are readily top of mind at having a net negative impact on my LW experience”), and Lucas-Gloor-as-of-2023 [2] (“Said’s way of asking questions, and the uncharitable assumptions he sometimes makes, is one of the most off-putting things I associate with LW”).
(All three of those users have posted comments within the last 70 days, so you can see I’m not using the “stopped posting on the site” criterion.)
I think it really matters that in these cases, we have the specific, first-party statements about discouragement and not just vague hearsay (that you remember the user making some sort of complaint, without telling us what the complaint was). If it’s not already obvious why I’m so distrustful of hearsay, see the next section.
How My Distrust Is Paying Rent in Anticipated Experiences
I continue to strongly disagree that in the world where I did make an unambiguous statement of general inference (that these authors were discouraged from posting on LW) that you finding the kind of comment that Jacob made, or the kind of statement that Scott made, that this would be any substantial evidence against my integrity
The issue is that my distrust of you keeps paying rent in anticipated experiences: I made better predictions because I distrusted you. (I’m writing about “trustworthiness” rather than “integrity” here, but I doubt that’s a crux. If necessary, see the section below on what I mean by “trustworthiness” in this context.)
Consider how your sequence of statements about Alexander would look to observers who don’t have access to your private memories and who believe that memories about what other people said can be unreliable. You were asked about “authors who find [a commenter’s] very presence in a discussion so ‘unpleasant’ that … it’s enough to discourage them from posting on LW altogether”, and responded with a list including Scott Alexander.
Even if there had been no grammatical ambiguity (and the comment had read “have complained along these lines” per your revision in the parent), I would have still been motivated to contact Alexander. I think that someone who trusted you would have predicted that I was wasting my time, that Alexander would corroborate your story by reporting that he had a negative firsthand impression of Achmiz. In fact, I wasn’t wasting my time: Alexander didn’t remember complaining and said he had “no direct opinion” on Achmiz. (He did say that he “deferred” to you about “what [you two] might have talked about in 2019″, but that’s not a corroboration.)
If there had been no grammatical error, I don’t think I would have used the phrase “false claims” (because it would have been clearer that you were only reporting what you remember being told, which isn’t directly contradicted by Alexander not remembering telling you that, because he could have forgotten), but I don’t think that would have changed the subsequent sequence of events very much; I still would have posted a comment in an accusatory register (whereas I wouldn’t have if Alexander had told me that he was discouraged by Achmiz).
In our actual timeline, you then clarified that “It wasn’t an incredibly intense mention” but that you “were talking about what makes LW comment sections good or bad, and [Achmiz] was a commenter we discussed in that conversation in 2019 or so.”
I think that someone who trusted you would infer that you meant that Alexander had mentioned Achmiz’s name. Because I didn’t trust you, I noted in §III.2 of my post that “it’s conspicuous that Habryka’s elaboration does not claim that Alexander volunteered Achmiz’s name, only that the ‘name came up’ and that Achmiz ‘was a commenter we discussed.’”
As it happened, my suspicion that Alexander did not name Achmiz at all seems to have been well-founded: in your 9 July comment above, you write that to the best of your knowledge, “Scott brought up Said, not by name but indirectly (referring to a specific commenter whose exact name he didn’t remember)” and only then did you “provide[ ] the name, which seemed to match.” (Seemed to match how? What did Alexander actually say, such that you could confidently infer who he meant?)
I notice a pattern in these events: twice I doubted the interpretation of your words that I think would come most naturally to someone who trusted you, and both times my distrust was vindicated (first by Alexander’s “no direct opinion”, then by your admission that you were the one who supplied Achmiz’s name, not Alexander): every time I’ve poked at the claim, it’s become less compelling as a justification for the ban. [3]
And after all that, I still don’t know what you’re alleging that Alexander said (which he has no memory of), only that you recall that there was some sort of “complaint”—and you’re still telling me (with respect to the claim about Alexander) [4] that you “would continue to write that same statement today.”
I hope you can see why this sort of behavior (where details that weaken the claim have to be extracted under adversarial questioning rather than volunteered up front) should (quantitatively) decrease people’s trust in you (such that they assign less credence when you make non-auditable claims on the basis of private evidence), even if every sentence you said permitted a true interpretation. (Because in the absence of someone distrustful and motivated enough to do the audit, the details would never surface, and people would have less accurate beliefs.) I think that a trustworthy speaker in this situation would be telling me that they wished the original 15 June 2025 comment had mentioned the “wasn’t an incredibly intense mention” part and that they were the one who supplied Achmiz’s name (such that Alexander turning out not to remember it would be less surprising to readers who trusted the original comment) and what they remembered Alexander having said more specifically than “complaints along these lines”. [5]
Appendix: What I Mean By “Trustworthiness” in This Context
If a speaker is trustworthy, then the beliefs I adopt when I interpret their words as being motivated by an intent to inform me should be approximately the same as the beliefs I would adopt after sending an auditor to check up on their claims. If someone is trustworthy, the auditor won’t predictably turn up counterevidence that lowers my credence in the beliefs the speaker is trying to get me to adopt (because a trustworthy speaker would have told me the counterevidence up front, without me needing to hire an auditor to get it).
Notably, this sense of “trustworthiness” is stronger than merely trusting the speaker to not tell explicit lies, because it’s possible for a speaker to induce false beliefs in a trusting audience while only using sentences that permit a true interpretation (as I wrote about in the Curated post “Firming Up Not-Lying Around Its Edge-Cases Is Less Broadly Useful Than One Might Initially Think”): for example, by selectively presenting evidence that supports their preferred conclusion (as I wrote about in the Best of Less Wrong 2019 post “Heads I Win, Tails?—Never Heard of Her; Or, Selective Reporting and the Tragedy of the Green Rationalists”). Furthermore, this sense of trustworthiness isn’t about the speaker’s subjective conscious intent, because I think behavior can be functionally optimized to mislead in the absence of such subjective conscious intent (as I wrote about in “Algorithmic Intent: A Hansonian Generalized Anti-Zombie Principle”).
- ↩︎
I would even argue that not finding anyone annoying would be a red flag that the forum is suffering from the kind of failure mode I discuss in “Hazards of Selection Effects on Approved Information”. As an umeshism: if you don’t have any annoying critics, then you don’t have enough critics. For example, I’ve found some of gjm’s criticisms of my posts to be subjectively annoying, but for that very reason, I think it’s important that I’ve often replied to them (although not always due to time constraints). If I didn’t have strong replies to my most annoying critics (or worse, didn’t have any annoying critics at all), then I couldn’t be so confident that I’m right.
- ↩︎
Gloor updated his comment in May 2024 to say that in the subsequent year and change, he “liked a couple of comments by Said and I don’t remember any particular ones that I thought exhibited the above pattern.”
- ↩︎
If Alexander felt as Sabien did, that would be a big deal in the minds of a lot of stakeholders; if, as is actually the case, Alexander doesn’t remember the complaint he allegedly made about someone whose name he couldn’t remember, that’s barely anything.
- ↩︎
I appreciate the statement that you would link Falkovich’s 2018 comment. I do consider the “some countervailing evidence” characterization a significant understatement (because the comment in Falkovich’s own words evidentially outweighs mere hearsay and is quite explicit in his support for Achmiz “bring[ing] a new flavor to the community” despite the previous impression of “disagreeableness [...] colored negatively in [his] mind”), but this is already a very long comment deep into a very long discussion and it’s not worth spending any more wordcount on this.
- ↩︎
I furthermore maintain that a trustworthy speaker wouldn’t suppress visibility of audits of the evidence for their claims on the grounds that it would incentivize them to spend time replying, “which [they] do not want to do”, as you did when you delisted Sting’s question post in contradiction of your claim in the OP that users can “make a post about how you disagree with some decision we made”, as I discuss in §VI.4.
- ↩︎
Some people posted about how to do cheaper better faster vision and it ended up in fable.
Some people posted about how to do cheaper better faster speculative decoding and it ended up in grok 4.5.
Details? Evidence? This is important and interesting if true, but if you just post the assertion without any substantiation, that’s not very helpful.
The thing you link to that I said was that “these people complained about Said”.
That’s not my understanding of what you were being asked. You were asked a question about “authors who find [a commenter]‘s very presence in a discussion so ‘unpleasant’ that … it’s enough to discourage them from posting on LW altogether”, and you answered with a list of names.
On my understanding of the meaning of that question, if someone is correctly named as an example of such an author, and I go ask them, “Did that commenter’s presence discourage you from posting on Less Wrong?”, I anticipate the experience that they’ll say “Yes.” (Here I’m making an assumption, which you seem to disagree with, that the author would remember having been discouraged and who discouraged them.)
I furthermore do not anticipate the experience of finding a comment by a correctly named example telling the commenter in question that their opinion of them has “flipped entirely to become positive” and encouraging them to “Do your own thing.” Even though the author is reporting that they used to have a negative opinion of the commenter at an earlier point in time (before it “flipped entirely to become positive”), that does not make it correct to say that their opinion was so negative that it was enough to discourage them from posting on the website. (I think it would be really surprising if such an extreme negative opinion could be so easily “flipped entirely to become positive.”)
It is perhaps a crux that I’m interpreting “find[ing] [a commenter]‘s very presence in a discussion so ‘unpleasant’ that … it’s enough to discourage them from posting on LW altogether” as a much stronger claim than “complain[ing] about [a commenter]”, such that correct answers to questions about the latter would be very often incorrect answers to questions about the former.
The reason I think those things are very different is because they’re very different in my own case, and I imagined that other people would be similar. I have some negative opinions (or one could as well use the word “complaints”) about lots of users of this website, but there’s no one I find so unpleasant that it’s enough to discourage me from posting on the website altogether. For example, I’ve complained about, say, Eliezer Yudkowsky (often, actually). If someone said, “Davis complained about Yudkowsky,” that would be a true claim. If someone said, “Davis finds Yudkowsky’s presence so unpleasant that it discourages him from using the website,” that would be a false claim. The claim would still be false even if someone erroneously thought the former implies the latter (and therefore wasn’t lying when they said it). I think it’s epistemically sloppy to collapse those two things (and that epistemically sloppy people are less trustworthy), even if epistemic sloppiness isn’t lying.
Regarding the assumption that an author would remember having been discouraged and who discouraged them, part of the reason that that seems like a reasonable assumption to me is that the central and uncontroversial case of an author finding Achmiz so unpleasant that it discourages them from using the website is Duncan Sabien, who is on the record saying as much. Sabien definitely remembers Achmiz, and has a direct negative opinion of him! I think it would be weird to put someone who says they have “no direct opinion” on Achmiz on the same list of discouraged authors as Sabien.
Which is just not the same as “readers would likely walk away with a wrong belief from Habryka’s comment which is wrong”.
I see. In retrospect, I wish I had gone with “misleading claims” rather than “false claims.” (Or maybe better, asked a clarifying question, as Richard Ngo suggests.) When a typical reader of my words walks away with a wrong belief, then the thing I said was “misleading” (independently of my conscious intent), even if it might not have been unambiguously “false” (because there exists a construal of my words that would make them true: for example, because my understanding of the question I was being asked differed from how typical readers interpreted the question).
I regret my word choice—by which I mean: I think that this experience will make me more likely to think carefully about whether I should say “misleading” rather than “false” in analogous future situations.
Importantly, both “misleading claims” and “false claims” need not entail lying. If I tell people “Munich is in Russia”, then I made a false claim, because actually, Munich is in Germany. It doesn’t matter whether I thought I was telling the truth (for example, because I misremembered something I read). Claims about “false claims” are about the claim, not the speaker’s private intent. People who see me making that mistake should regard me as less trustworthy about geography.
Indeed, the whole conversation is centrally about the degree to which I am trustworthy.
Yes, this whole conversation is centrally about the degree to which you are trustworthy. However—
whether I lied [...] the discussion of whether I am lying [...] evaluate whether I am lying to people
I think there are ways to be (somewhat, quantitatively, in a topic-dependent way) untrustworthy without consciously lying, but simply by being biased: for example, by overestimating the degree to which other people share your dislike of Achmiz and interpreting ambiguous statements from them in the light of that prior without being clear in your reports to others about the interpretive lens that you’re adding. (I’ve been writing about this kind of phenomenon for years.)
I think I’ve been very careful to not sloppily misuse the l-word. That’s why I made sure to explain above (and in my post) that “the doubt is agnostic as to the reason for the false reports” because “[w]hat matters is the likelihood ratio”. If it helps, think about a machine learning classifer rather than a human: if a classifer assigns positive labels to data points that are confirmed to be negative, that does make the classifier less trustworthy.
I don’t think I’m holding you to standards that I wouldn’t hold myself. If someone asked me, “Who dislikes Alice?” and I replied, “Bob dislikes Alice,” on the basis of my fuzzy memories of a conversation that I had with Bob seven years ago (in which Bob said something negative about someone whose name he didn’t know, and I supplied Alice’s name, which seemed to match the description Bob gave) and then Dave came to me and said, “You are making false claims; I talked to Bob, and he said he has no direct opinion of Alice”, I think I would be embarrassed! In addition to giving my side of the story about why I said what I did (about what I remembered about that conversation with Bob seven years ago), I think I would apologize for having replied to the literal question “Who dislikes Alice?” with the literal answer “Bob dislikes Alice” when Bob isn’t corroborating that. I think if I stood by my original answer and insisted I had done nothing wrong, people would be right to (quantitatively) distrust me more because of that!
If you find it impossible to think that I had a conversation with Falkovich in which he complained to me about Said, given that comment
No, that’s not what I’m saying. Let me try again.
Your 15 June 2025 comment replied to a question about “authors who find [someone]‘s very presence in a discussion so ‘unpleasant’ that … it’s enough to discourage them from posting on LW altogether” with a list that included Falkovich’s name.
I think that readers who read that comment were likely to walk away with the belief that, as of June 2025, Falkovich found Achmiz’s very presence in a discussion so unpleasant that it was enough to discourage Falkovich from posting on Less Wrong altogether. I’m saying that that belief is contradicted by Falkovich’s October 2018 comment. I agree that this is compatible with you having had an earlier conversation with Falkovich in which he made some sort of complaint about Achmiz. If you think the wording in my post is unclear, I’m happy to consider suggested edits.
(This is a rhetorical question, I am not actually interested in engaging in a longer conversation here)
Sure. I’m not asking you to engage in a longer conversation. I’m correcting your public characterization of my position.
To clarify for anyone reading—
Also, what is going on with you taking two statements by authors I talked to, neither of which directly contradicted how I summarized them, with the summary “was not true when I actually checked” and “having made false claims about people”.
The claim I was checking was the list of allegedly discouraged authors in your 15 June 2025 comment:
My guess is something like more than half of the authors to this site who have posted more than 10 posts that you commented on, about you, in particular. Eliezer, Scott Alexander, Jacob Falkovich, Elizabeth Van Nostrand, me, dozens of others.
I’m saying that I don’t think the additional context in your 12 July 2025 comment rescues the discouragement claims in the 15 June comment.
Jacob literally started his sentence with “his negative affect towards Achmiz”
Specifically, the sentence in question was (bold in original) “But now that you’ve stated that you’re disagreeable on purpose, the negative effect flipped entirely to become positive.”
In §III.2, I discuss in more detail why I don’t think your interpretation in your 12 July comment rescues the claim in your 15 June comment:
In response to being presented with Falkovich’s comment contradicting his claim about Falkovich’s opinion, Habryka replied:
I think you can clearly see how the Jacob Falkovich one is complicated. He basically says “I used to be frustrated by you, but this thing made that a lot better”. I don’t remember the exact time I talked to Jacob about it, but it had come up sometime some context where we discussed LW comment sections. It’s plausible to me it was before he made this comment, though it would be a bit surprising to me, since that’s pretty early into LW’s history.
In fact, I do not see how the Jacob Falkovich one is “clearly” complicated. Indeed, I dispute Habryka’s characterization of what Falkovich “basically says”: it is tendentious to paraphrase “negative [...] flipped entirely to become positive” as “made that a lot better.” The latter is compatible with Habryka’s original claim that Falkovich is discouraged from using the website by Achmiz, but the former directly contradicts it: if something that’s bothering you is “made [...] a lot better”, the implication is that it’s still bothering you a little bit (although not as much as before); if your attitude towards something “become[s] positive”, that implies that it’s not bothering you.
I gave you more context on the Scott conversation which Scott didn’t dispute
I also discuss this in more detail in §III.2:
Notably, [Alexander’s reported lack of memory] is not a corroboration of Habryka’s original claim that Achmiz “in particular” discouraged Alexander from using Less Wrong: if Achmiz’s comments were so noxious as to drive Alexander off the website, one would have expected Alexander to have at least some memory of it. Indeed, it’s conspicuous that Habryka’s elaboration does not claim that Alexander volunteered Achmiz’s name, only that the “name came up” and that Achmiz “was a commenter we discussed.”
Thanks for commenting!
because I’m worried that you believing in Speech to this extent will end up being an unproductive crux for us
I don’t think so: while I concede that my deciding to talk to Metz and my opposition to the Achmiz ban are “correlated” in some sense, I fully expect that many people who would disapprove of the former would agree with my case on the latter. “Whether to talk to a biased journalist” and “whether to ban a Less Wrong user” are actually just pretty different situations, even if they both involve speech. Indeed, Achmiz himself does not share my obsession with transparency maximalism, and has a much dimmer view than me on the merits of talking to journalists, whom he has collectively described as “the scum of the earth”. [1]
It’s something more like “Habryka has a job, and he seems to be doing his job in a reasonable way in this case”.
Great, happy to start there. I think that one particularly legible requirement for doing the job of moderator in a reasonable way—not necessarily the most important requirement, but one that’s easy to check whether it’s being fulfilled—is, “Don’t make false claims to justify moderation decisions, and if you do accidentally make false claims, you should apologize and credibly express intent to not mislead people in that way again.”
As I document in §III.2 of my post, Habryka claimed in June 2025 that Scott Alexander and Jacob Falkovich, among “dozens of others”, were discouraged from using the website by Achmiz’s comments.
However, when I checked with Alexander, he testified that he had “no direct opinion” on Achmiz. I also found an October 2018 comment from Falkovich in which Falkovich wrote that his previous negative affect towards Achmiz had “flipped entirely to become positive” and urged Achmiz to “Do your own thing, and own it.”
When I presented Habryka with Alexander and Falkovich’s statements, rather than apologizing for having made false claims about other people’s stances on Achmiz, Habryka said of Alexander, “My guess is he doesn’t remember. It wasn’t an incredibly intense mention” (apparently putting Habryka’s own word against Alexander’s on the question of Alexander’s opinions about Achmiz?) and claimed that Falkovich’s statement was “complicated.”
Habryka continues to stand by a claim stated in the ban announcement post that “many top authors cit[e] [Achmiz] as a top reason for why they do not want to post on the site, or comment here”, while noting that most complaints about other users are private. The problem here is that when someone claims that some things are examples of a phenomenon, and the things turn out not to be examples when checked, that casts doubt on claimed examples that we can’t check: if the thing Habryka claimed about Alexander and Falkovich was not true when I actually checked, then what should we believe about the “dozens of others” or “many top authors” who weren’t named? As I explain in footnote 17, the doubt is agnostic as to the reason for the false reports. (What matters is the likelihood ratio
To be clear, this is not a particularly important point in itself. (The stated basis for the ban is not “Scott Alexander said so.”) The reason I’m re-explaining it in this comment (separately from the longer discussion in §III.2) is because it’s particularly legible: you don’t have to evaluate a complicated argument to check the linked statements and see for yourself that Habryka’s claims about authors’ opinions about Achmiz were contradicted by those authors’ own statements. My hope is that this small token of evidence injects enough of a shadow of a doubt into your prior belief that Habryka is doing his job in a reasonable way in this case, that you might find it worth your time to reconsider that belief after reading §II, §III.1, and §IV, which I think are important (but less trivial to evaluate).
If you find any of it persuasive (or unpersuasive), I think it would be in the public interest for you to say so in public.
- ↩︎
Specifically, Achmiz wrote to me in a January 2024 email (quoted with permission):
That’s journalist thinking—the idea that as long as you didn’t say the magic words “this is off the record”, you can publish anything anyone says to you in any context. And we all understand quite well by now that this (among other things) is what makes journalists, as a class, the scum of the earth. (“Never talk to a journalist”, one often hears—but why not? Because when you talk to a journalist—unlike when talking to normal people who understand the ideas of private communication and discretion and basic decency—you might well find your words broadcast publicly.)
- ↩︎
Right.
I don’t think that’s the axis. Obviously, if no one had any positive vision, then there would be nothing to critique. People who care about intellectual progress should definitely be trying to have positive visions.
It’s just that people who want their positive visions to be correct should also care about fixing flaws as they’re pointed out.
For example, I’m currently working on several interrelated posts in which I articulate a positive vision about signaling theory. (It’s about the conditions under which “honest” agents can successfully distinguish themselves from “dishonest” agents, but the details aren’t important.)
I think my positive vision is basically on the right track, but that doesn’t mean I’ve got it exactly right. I ran the drafts by my friend Said Achmiz via email, and he had a lot of detailed criticisms, some of which I disagree with, but some of which I recognized as serious. It spurred me to do a lot of laborious rewriting which I’m still not done with.
I think Achmiz’s criticism was really helpful for refining my positive vision. I don’t think he was being nice to me in his comments because I’m his friend. I think he was saying the same kind of things to me that he says to everyone. [1] If some people are not only personally uninterested in reading this kind of detailed, incisive engagement with their work, but don’t even want it to be published for others to read, I think the most plausible reason for that is that they don’t care whether their positive vision is correct and don’t want anyone else to find out, either.
Sorry, I know that seems like a really mean or “uncharitable” thing to say. But, well, I’m not sure what alternative theory of people’s behavior that you (Richard) think I should be embracing instead. (More below on my understanding of others’ theories.) Just, where would someone even get the idea that “critiquing flaws” and “having a positive vision” are somehow in a zero-sum competition with each other on opposite ends of an axis, such that people who are in favor of critiquing flaws are therefore less in favor of positive visions?
As far as I can tell, it seems to have something to do with a psychological hypothesis about morale. When Elizabeth von Nostrand writes about “a general trade off between authors’ experience and improving correctness” or Raymond Arnold writes that “it’d basically be the wrong call for any platform to ignore that” “content creation is harder than critique”, the idea seems to be that if people’s attempts to formulate positive visions get critiqued too vigorously, they’ll get discouraged and give up.
The implied psychological model of Less Wrong authors reminds me of my attempts to teach chess to my five-year-old niece this week. It’s not just that I let her capture material [2] in order to keep her engaged with the endeavor of learning how to make legal moves. It’s that the idea of losing material was so aversive that she didn’t seem to be able to process my attempted instruction of the form, “Okay, I’m not playing seriously; I’m going to let you capture my pony in a moment; but if I were playing seriously, I would recapture your rook here.”
Basically, I think Less Wrong authors are capable of having more emotional maturity than a five-year-old: we don’t need to falsely “let them win” in order to induce them to participate at all. If you think I’m being unrealistically optimistic about that, then I feel like I’m not the one who doesn’t believe in intellectual progress.
But you can’t seriously have expected me to disagree that intellectual progress is in fact a valuable thing which we can aim for.
Uh, correct me if I’m misreading this, but it seems like you’re trying to deescalate the conflict by proposing mutual concessions: you’ve accepted a belief that favors my side, so I should reciprocate and accept a belief of your choice.
That’s not how it works. Treating epistemics as social exchange—trying to see things the other guy’s way on his request, in exchange for him trying to see it your way on your request—doesn’t create true maps; it creates false maps representing a compromise between the parties’ preferred lies. I put a lot of effort into explaining why I don’t think Habryka’s observed behavior is well-described as trying to protect intellectual progress. If you think I’m wrong about that, you should be able to explain why I’m wrong. It’s not a trade!
Selected excerpts: “So why would this argument work? Why should it work? It’s deceptive, isn’t it? And transparently so!”, “I am just not convinced by almost anything else you write in these two posts”, “Until you figure out just what is going on there, I don’t think that anything you write on this topic will manage to form any kind of coherent or sensible whole”, “I still think that you’re misinterpreting/misconstruing rather than miscommunicating”.
It seems imprecise to say “let her win”, since I despaired of trying to explain the concept of checkmate.