Zack_M_Davis

• Zack_M_Davis 23 Sep 2025 18:55 UTC

  11 points

  6

  on: We Sup­port “If Any­one Builds It, Every­one Dies”

  I could be described as supporting the book in the sense that I preordered it, and I bought two extra copies to give as gifts. I’m planning to give one of them tomorrow to an LLM-obsessed mathematics professor at San Francisco State University. But the reason I’m giving him the book is because I want him to read it and think carefully about the arguments on the merits, because I think that mitigating the risk of extinction from AI should be a global priority. It’s about the issue, not supporting MIRI or any particular book.

• Zack_M_Davis 23 Sep 2025 2:18 UTC

  4 points

  1

  in reply to: Raemon’s comment on: This is a re­view of the reviews

  Thank you for clarifying.

  I think this was fairly obvious

  No, it was not obvious!

  You replied to a comment that said, verbatim, “what we should indeed sacrifice is our commitment to being anal-retentive about practices that we think associate with getting the precise truth, over and beyond saying true stuff and contradicting false stuff”, with, “This paragraph feels righter-to-me”.

  That response does prompt the reader to wonder whether you believe the quoted statement by Malcolm McLeod, which was a prominent thesis sentence of the comment that you were endorsing as feeling righter-to-you! I understand that “This feels righter-to-me” does not mean the same thing as “This is right.” That’s why I asked you to clarify!

  In your clarification, you have now disavowed the quoted statement with your own statement that “We absolutely should have more practices that drive at the precise truth than saying true stuff and contradicting false stuff.”

  I emphatically agree with your statement for the reasons I explained at length in such posts as “Firming Up Not-Lying Around Its Edge-Cases Is Less Broadly Useful Than One Might Initially Think” and “Heads I Win, Tails?—Never Heard of Her; Or, Selective Reporting and the Tragedy of the Green Rationalists”, but I don’t think the matter is “fairly obvious.” If it were, I wouldn’t have had to write thousands of words about it.

• Zack_M_Davis 23 Sep 2025 1:44 UTC

  −12 points

  −2

  in reply to: Raemon’s comment on: This is a re­view of the reviews

  This is important enough that you should clarify in your own words. Raymond Arnold, as a moderator of lesswrong.com, is it in fact your position that “what we should indeed sacrifice is our commitment to being anal-retentive about practices that we think associate with getting the precise truth, over and beyond saying true stuff and contradicting false stuff”?

• Zack_M_Davis 22 Sep 2025 5:17 UTC

  30 points

  12

  in reply to: MalcolmMcLeod’s comment on: This is a re­view of the reviews

  One is “We must never abandon this relentless commitment to precise truth. All we say, whether to each other or to the outside world, must be thoroughly vetted for its precise truthfulness.” To which my reply is: how’s that been working out for us so far?

  [...]

  We can win without sacrificing style and integrity.

  But you just did propose sacrificing our integrity: specifically, the integrity of our relentless commitment to precise truth. It was two paragraphs ago. The text is right there. We can see it. Do you expect us not to notice?

  To be clear, in this comment, I’m not even arguing that you’re wrong. Given the situation, maybe sacrificing the integrity of our relentless commitment to precise truth is exactly what’s needed!

  But you can’t seriously expect people not to notice, right? You are including the costs of people noticing as part of your consequentialist decision calculus, right?

• Zack_M_Davis 21 Sep 2025 22:45 UTC

  5 points

  0

  in reply to: Ishual’s comment on: Safety re­searchers should take a pub­lic stance

  How do you think norm enforcement works, other than by threatening people who don’t comply with the norm?

• Zack_M_Davis 21 Sep 2025 18:43 UTC

  10 points

  1

  on: And Yet, Defend your Thoughts from AI Writing

  Like feeling the rain on your skin, no one else can feel it for you.

  This is a deliberate reference to the lyrics of Natasha Bedingfield’s thematically-relevant song “Unwritten”, right? (Seems much more likely than coincidence or cryptomnesia.) I can empathize with it feeling too cute not to use, but it seems like a bad (self-undermining) choice in the context of an essay about the importance of struggling to find original words?

• Zack_M_Davis 21 Sep 2025 2:00 UTC

  3 points

  0

  in reply to: Vladimir_Nesov’s comment on: Com­ment on “Four Lay­ers of In­tel­lec­tual Con­ver­sa­tion”

  (Fixed; thanks for your patience.)

• Zack_M_Davis 12 Sep 2025 7:02 UTC

  2 points

  0

  in reply to: mesaoptimizer’s comment on: AllAmer­i­canBreak­fast’s Shortform

  Followup question: you thought criticism was useful in April 2023. What changed your mind?

• Zack_M_Davis 11 Sep 2025 1:30 UTC

  6 points

  1

  in reply to: mesaoptimizer’s comment on: AllAmer­i­canBreak­fast’s Shortform

  your attempts at posting good-faith critiques in the comments of most LW posts are costlier to [...] the community you care about, than they are beneficial.

  Why? What are the costs to the community?

• Zack_M_Davis 7 Sep 2025 17:08 UTC

  2 points

  0

  in reply to: philh’s comment on: Critic Con­tri­bu­tions Are Log­i­cally Irrelevant

  Thanks for commenting!

  Describing them as posts versus comments probably isn’t ideal, but I think it’s mostly okay.

  Yes, in retrospect, I wish I had done a better job of flagging the metonymy. I’m glad the idea got through despite that.

  I claim that yes, these two different types of writing are significantly different activities.

  Different in what respect? When I write a critical post (arguing that author $X$ is wrong about $Y_1$ because $Z_1$), it feels like relevantly the same activity as when I write a “non-critical” post (just arguing that $Y_2$ because $Z_2$ without reference to any reputedly mistaken prior work) in terms of what cognitive skills I’m using: the substance is about working out how $Z_i$ implies $Y_i$. That’s the aspect relevant to the playing/​coaching metaphor. Whether there happens to be an $X$ in the picture doesn’t seem to change the essential character of the work. (Right? Does your subjective assessment differ?)

  The effect of rendering these bytes as text preceded by my username does not need to be the same as the effect of rendering these bytes as text preceded by another username!

  It doesn’t need to, but should it? The section titled “However, Critic Contributions Can Inform Uncertain Estimates of Comment Value” describes one reason why it should. My bold philosophical claim is that that’s the only reason. (I’m counting gjm’s comment about known expertise as relevantly “the same reason.”)

  Alternatively, for the purpose of the argument in that section, we can instead imagine that we’re talking about a blog where the commenting form has a blank “Author name” field, rather than a site with passworded accounts: the name could be forged just as easily as the comment content, and the “comment” is the (author-name, content) pair. That would restore the screening-off property.

• Zack_M_Davis 3 Sep 2025 7:19 UTC

  5 points

  1

  in reply to: Said Achmiz’s comment on: Ban­ning Said Ach­miz (and broader thoughts on mod­er­a­tion)

  When discussing rationality, I typically use the word normative to refer to what idealized Bayesian reasoners would do, often in contrast to what humans do.

  (Example usage, bolding added: “Normatively, theories are preferred to the quantitative extent that they are simple and predict the observed data [...] For contingent evolutionary-psychological reasons, humans are innately biased to prefer ‘their own’ ideas, and in that context, a ‘principle of charity’ can be useful as a corrective heuristic—but the corrective heuristic only works by colliding the non-normative bias with a fairness instinct [...]”)

  As Schopenhauer observes, the entire concept of adversarial debate is non-normative!

  “[N]ot demand[ing] [...] that a compelling argument be immediately accepted” is normatively correct insofar as even pretty idealized Bayesian reasoners would face computational constraints, but a “stubborn defense of one’s starting position—combined with a willingness [...] to change one’s mind later” isn’t normatively correct, because the stubbornness part comes from humans’ innate vanity rather than serving any functional purpose. You could just say, “Let me think about that and get back to you later.”

• Zack_M_Davis 3 Sep 2025 5:24 UTC

  8 points

  0

  in reply to: Ben Pace’s comment on: Ban­ning Said Ach­miz (and broader thoughts on mod­er­a­tion)

  And yet here you demand I immediately change my mind in response to reason and evidence.

  I think this is an improperly narrow interpretation of the word now in the grandparent’s “I’ll take that retraction and apology now.” A retraction and apology in a few days after you’ve taken some time to cool down and reflect would be entirely in line with Schopenhauer’s advice. I await the possibility with cautious optimism.

  Zack Davis describes that position as “laughable, obviously wrong, and deeply corrosive”

  I mean, I do think that (recall that I actually did the experiment with an LLM to demonstrate), but do you understand the rhetorical device I was invoking by using those exact words in the comment in question?

  You had just disparagingly characterized Achmiz as “describing [interlocutors’] positions as laughable, obviously wrong, deeply corrosive, etc”. I was deliberately “biting the bullet” by choosing to express my literal disagreement with your hyperbolic insult using those same words verbatim, in order to stick up for the right to express disagreement using strong language when appropriate.

  Just checking that you “got the joke.”

  “normatively correct”. You guys

  Please note that I had put a Disagree react on the phrase “normatively correct” on the comment in question. (The react was subsequently upvoted by Drake Morrison and Habryka.)

  My actual position is subtler: I think Schopenhauer is correct to point out that it’s possible to concede an argument too early and that good outcomes often result from being obstinate in the heat of an argument and then reflecting at leisure later, but I think describing the obstinacy behavior as “normatively correct” is taking it way too far; that’s not what the word normative means.

• Zack_M_Davis 2 Sep 2025 16:53 UTC

  2 points

  0

  in reply to: habryka’s comment on: Which top au­thors did Said Ach­miz drive away?

  Thank you for answering my question.

• Zack_M_Davis 2 Sep 2025 4:58 UTC

  10 points

  0

  in reply to: habryka’s comment on: Which top au­thors did Said Ach­miz drive away?

  Allowing lots of top-level posts

  As it happens, I was planning (in due time) to write my own top-level reaction post to your post of 22 August. I had assumed this would be allowed, as I have written well-received top-level reaction posts to other Less Wrong posts many times before: for example, “Relevance Norms” (which you evidently found valuable enough to cite in your post of 22 August) or “Firming Up Not-Lying Around Its Edge-Cases Is Less Broadly Useful Than One Might Initially Think” (which was Curated).

  Will I be permitted to post?

  will inevitably then cause me to have to spend another 100+ hours on this

  I don’t think “have to” is warranted. You don’t have to reply if you don’t want to. But other people have a legitimate interest in publicly discussing your public statements among themselves, independently of whether you think it’s worth your time to reply.

• Zack_M_Davis 2 Sep 2025 3:11 UTC

  6 points

  3

  in reply to: habryka’s comment on: Which top au­thors did Said Ach­miz drive away?

  Sting, the author of the post, thought a top-level question post was the right choice. If you think the algorithm should deprioritize showing other people Sting’s post because you think it’s bad, you have a strength-10 strong downvote. Why isn’t that enough? On any other topic, if someone makes a post about something other people have also made posts about, you don’t demote later posts to comments. Why is this topic different?

• Zack_M_Davis 18 Aug 2025 1:50 UTC

  9 points

  5

  in reply to: AnthonyC’s comment on: My In­ter­view With Cade Metz on His Re­port­ing About Lighthaven

  What makes you describe this as a “typical reporter perspective”? One would expect that people who write for a living are sensitive to the effects of word choices (such that, if they’re nudging readers one way or the other, it’s probably on purpose rather than on accident).

• Zack_M_Davis 18 Aug 2025 1:49 UTC

  34 points

  17

  in reply to: orthonormal’s comment on: My In­ter­view With Cade Metz on His Re­port­ing About Lighthaven

  the only important thing is ensuring that these weirdos don’t get status

  Seems too self-centered to be the real explanation. (Most of the time, people who do things that hurt you aren’t doing it because they hate you; it’s because you’re in the way.)

  As a technology reporter whose job is to cover what rich and powerful people in Silicon Valley are up to, the fact that companies your readers have heard of (DeepMind and OpenAI and Anthropic) are causally downstream of this internet ideology that no one has heard of, is itself an interesting story that the public deserves to hear about.

  It is a legitimate and interesting story that the public deserves to hear about! The problem, from our perspective, is that he doesn’t accept that the object level is a relevant part of the story. He’s correct to notice the asymmetry in vibes between people at MATS trying to save the world and people at Meta trying to make money as being “a key part of the debate” as far as the psychology of the participants goes—and by writing a story about that observation, he’s done his job as a technology reporter. Simulacrum 1 isn’t in scope.

• Zack_M_Davis 17 Aug 2025 2:31 UTC

  20 points

  1

  on: My In­ter­view With Cade Metz on His Re­port­ing About Lighthaven

  (The original recording is available, but there’s a lot of background noise.)

My In­ter­view With Cade Metz on His Re­port­ing About Lighthaven

In­ter­view with Kel­sey Piper on Self-Cen­sor­ship and the Vibe Shift