Will Aldred

Karma: 4,597

Will Aldred Jan 30, 2024, 5:19 PM
17 points
1 ∶ 0
in reply to: weeatquince’s comment on: EV investigation into Owen and Community Health
(Fwiw, the Forum moderation team does this for many of our cases.)

Will Aldred Jan 28, 2024, 11:48 PM
2 points
0 ∶ 0
in reply to: Matthew_Barnett’s comment on: Matthew_Barnett’s Shortform
Objection 1: It is unlikely that there will be a point at which a unified agent will be able to take over the world, given the existence of competing AIs with comparable power
For what it’s worth, the Metaculus crowd forecast for the question “Will transformative AI result in a singleton (as opposed to a multipolar world)?” is currently “60%”. That is, forecasters believe it’s more likely than not that there won’t be competing AIs with comparable power, which runs counter to your claim.
(I bring this up seeing as you make a forecasting-based argument for your claim.)

Will Aldred Jan 15, 2024, 3:58 PM
10 points
0 ∶ 0
on: AI doing philosophy = AI generating hands?
Following on from your saner world illustration, I’d be curious to hear what kind of a call to action you might endorse in our current world.
I personally find your writings on metaphilosophy, and the closely related problem of ensuring AI philosophical competence, persuasive. In other words, I think this area has been overlooked, and that more people should be working in it given the current margin in AI safety work. But I also have a hard time imagining anyone pivoting into this area, at present, given that:^[1]
1. There’s no research agenda with scoped out subproblems (as far as I’m aware), only the overall, wicked problem of trying to get advanced AIs to do philosophy well.
2. There are no streams within junior research programs, like MATS, to try one’s hand^[2] in this area while gaining mentorship.
1. ^
  A third reason, which I add here as a footnote since it seems far less solvable: Monetary and social incentives are pushing promising people into empirical/ML-based intent alignment work. (To be clear, I believe intent alignment is important. I just don’t think it’s the only problem that deserves attention.) It takes agency—and financial stability, and a disregard for status—to strike out on one’s own and work on something weirder, such as metaphilosophy or other neglected, non-alignment AI topics.
  [ETA: Ten days after I posted this comment, Will MacAskill gave an update on his work: he has started looking into neglected, non-alignment AI topics, with a view to perhaps founding a new research institution. I find this encouraging!]
2. ^
  Pun intended.

Will Aldred Jan 10, 2024, 7:39 PM
2 points
0 ∶ 0
in reply to: Greg_Colbourn ⏸️ ’s comment on: MIRI 2024 Mission and Strategy Update
Yes, this is a fair point; Holden has discussed these dangers a little in “Digital People Would Be An Even Bigger Deal”. My bottom-line belief, though, is that mind uploads are still significantly more likely to be safe than ML-derived ASI, since uploaded minds would presumably work, and act, much more similarly to (biological) human minds. My impression is that others also hold this view? I’d be interested if you disagree.
To be clear, I rank moratorium > mind uploads > ML-derived ASI, but I think it’s plausible that our strategy portfolio should include mind uploading R&D alongside pushing for a moratorium.

Will Aldred Jan 9, 2024, 11:57 PM
8 points
3 ∶ 0
in reply to: Guy Raveh’s comment on: Will Aldred’s Shortform
Hmm, based on what you’ve said here—and I acknowledge that what you’ve said is a highly compressed version of your experience, thus I may well be failing to understand you (and I apologize in advance if I mischaracterize your experience)—I think I’m not quite seeing how this refutes my framing? I accept that my type-A/B framing rounds off a bunch of nuance, but to me, within that framing, it sounds like you’re type-A?
Like, I’m not sure how long the transition period was for you, and I expect different people’s transition periods will vary considerably, but my model, viewed through this lens, is that a type-A person will make it out of their transition period and be able to maintain a vegan diet thereafter at little to no cost. Whereas a type-B can spend weeks, months—even a year or more, like myself^[1]—planning out and iterating on their vegan diet; making sure, through doing research, taking blood tests, and so on, that they’re avoiding the known pitfalls, and still never make it out of the transition period.^[2]^[3]
1. ^
  I’ve written about my experience here.
2. ^
  I like this comment from Jason: “Nutritional research is hard, and we’d need a significantly stronger body of research (e.g., random assignment, very large samples) to say that a vegan diet is maximally healthful for everyone at an individual level (as opposed to healthier on the a [sic] population average).” (link)
3. ^
  Moreover, for me, the vegan experience actually got increasingly unpleasant with time, if anything, so I don’t think it’s the case that type-B’s will eventually asymptote onto becoming costlessly vegan if only they stick with it for long enough.
  (Additionally, if asymptoting really does occur, but “long enough” means months or years, then I have sympathy for those who give up in the meantime.)

Will Aldred Jan 8, 2024, 9:35 PM
4 points
0 ∶ 0
in reply to: Elizabeth’s comment on: Will Aldred’s Shortform
Thanks for your comments, both. I agree that the personal versus universal statements distinction is noteworthy (and missing from my take above).

Will Aldred Jan 7, 2024, 10:17 PM
37 points
14 ∶ 1
on: Will Aldred’s Shortform
Shower thought: A lot of the talking past each other that happens between vegan and non-vegan^[1] EAs^[2] might come from selection effects plus typical mind fallacy.^[3]
Let’s say there are two types of people: type-A, for whom a vegan diet imposes little or no costs, and type-B, for whom a vegan diet imposes substantial costs (in things like health, productivity,^[4] social life). My hunch is that most long-time vegans are type-A, while most type-B people who try going vegan bounce.
Now, to a type-A vegan who doesn’t realize type-B is a thing, those who claim that a vegan diet has costs are obviously lying. And to a type-B person who’s tried and failed to go vegan, vegans are obviously lying about how easy it is. (To put things crudely.)
I think an awareness from both sides of the vegan/non-vegan divide that the other type—B/A—exists, and that the other side is mostly made up of the other type, could go quite a long way toward circumventing frustrating or otherwise unproductive debates.
(I doubt this take is original to me, but I didn’t find it on this forum already with a quick search, so I figured I’d share.)
1. ^
  I expect the same argument to apply to vegetarians versus non-vegetarians, but for brevity I’ll just write “vegan” throughout. (I’m avoiding the veg*n shorthand just because I don’t find *’s aesthetically appealing in prose.)
2. ^
  I’ve specified “EAs” rather than just people in general because my argument works better if most non-vegans have actually tried to go vegan. My sense is that this is true for the EA population, but (very) not true for the general population.
3. ^
  My preferred definition of typical mind fallacy: The mistake of jumping to conclusions—often subconsciously—about other people’s experiences based on your own experiences. In other words, the mistake of assuming that other people are more like you than they really are. (Here’s the LessWrong Wiki’s definition.)
4. ^
  I’ve previously written about my experience here.

Will Aldred Jan 6, 2024, 7:20 PM
14 points
0 ∶ 1
in reply to: lilly’s comment on: Survey of 2,778 AI authors: six parts in pictures
Just to give your final point some context: the average in-depth research project by Rethink Priorities reportedly costs $70K-$100K. So, if this AI Impacts survey cost $138K in participant compensation, plus some additional amount for things like researcher time, then it looks like this survey was two or three times more expensive than the average research project in its approximate reference class.
I haven’t thought hard about whether the costs of EA-funded research make sense in general, but I thought I’d leave this comment so that readers don’t go away thinking that this survey cost like an order of magnitude more than what’s standard.

Will Aldred Jan 5, 2024, 3:49 PM
11 points
1 ∶ 1
on: MIRI 2024 Mission and Strategy Update
I’m curious, since it sounds like MIRI folks may have thought about this, if you have takes on how best to allocate marginal effort between pushing for cooperation-to-halt-AI-progress on the one hand, and accelerating cognitive enhancement (e.g., mind uploading) on the other?^[1]
Like, I see that you list promoting cooperation as a priority, but to me, based on your footnote 3, it doesn’t seem obvious that promoting cooperation to buy ourselves time is a better strategy at the margin than simply working on mind uploading.^[2] (At least, I don’t see this being obviously true for people-trying-to-reduce-AI-risk at large, and I’d be interested in your—or others’—thoughts here, in case there’s something I’m missing. It may well be clearly true for MIRI given your comparative advantages; I’m asking this question from the perspective of overall AI risk reduction strategy.) Here’s that footnote 3:
Nate and Eliezer both believe that humanity should not be attempting technical alignment at its current level of cognitive ability, and should instead pursue human cognitive enhancement (e.g., via uploading), and then having smarter (trans)humans figure out alignment.
Related recent discussion:
- “Does davidad’s uploading moonshot work?”
  - Context: David Dalrymple (aka davidad) recently outlined a concrete plan for mind uploading by 2040.
Related prediction markets:
- Eliezer’s Manifold market, “If Artificial General Intelligence has an okay outcome, what will be the reason?”
  - At present, the leading answer is: “Humanity successfully coordinates worldwide to prevent the creation of powerful AGIs for long enough to develop human intelligence augmentation, uploading, or some other pathway into transcending humanity’s window of fragility.”
- My Metaculus question, “Will mind uploading happen before AGI?”
  - The current community prediction is 1%.^[3]
1. ^
  ETA: I’ve just noticed that earlier today, another Forum user posted a quick take on a similar theme, asking why there’s been no EA funding for cognitive enhancement projects. See here.
2. ^
  The immediate lines of reasoning I can think of for why “put all marginal effort towards pausing AI” is the best strategy right now are: i) uploading is intractable given AGI timelines, and ii) future, just-before-the-pause models—GPT-7, say—could help significantly with mind uploading R&D. But then, assuming that uploading is our best bet for getting alignment right, I think ii just shifts the discussion to things like “where is the best place to pause (with respect to the tradeoff between powerful automation of uploading R&D versus not pausing too late)?” and “are there ways to push for differential progress in models’ capabilities? (e.g., narrow superhuman ability in neuroscience research).”
  What’s more, as counters to i: Firstly, most problems fall within a 100x tractability range. Secondly, even if cooperation+pause efforts are clearly higher impact right now than object-level uploading work, I think there’s still the argument that field-building for mind uploading should start now, rather than once the pause is in place. Because if field-building starts now, then with luck there’ll be a body of uploading researchers ready to make the most of a future pause. (This argument doesn’t go through if the pause lasts indefinitely, because in that case there’s time to build up the mind uploading field from scratch in the pause. But it does go through if the pause is limited or fragile, which I tentatively believe are more likely possibilities. See also Scott Alexander’s taxonomy of AI pauses.)
3. ^
  Taken together, these two prediction markets arguably paint a grim picture. Namely, the trades on Eliezer’s question imply that mind uploading is the most likely way that AGI goes well for humanity, but the forecasts on my question imply that we’re very unlikely to get mind uploading before AGI.
What links here?
- Will Aldred's comment on Digital Minds Takeoff Scenarios by Bradford Saad (Jul 10, 2024, 1:19 PM; 3 points)

Will Aldred Dec 27, 2023, 7:33 PM
12 points
2 ∶ 0
in reply to: Will Aldred’s comment on: Only mammals and birds are sentient, according to neuroscientist Nick Humphrey’s theory of consciousness, recently explained in “Sentience: The invention of consciousness”
Also, I’m pretty sure that octopuses do play? A quick search appears to confirm this: “Octopuses like to play” (BBC).
I mention this in response to the part of the post that reads: “Other animals like fish, reptiles, and octopuses do not engage in sensation-seeking or play and so do not have those internal conscious experiences.”
(ETA: I see that in their closing section, OP acknowledges some uncertainty here, listing as a question for further investigation: “Is it really true that fish, shrimp, octopuses, and other animals of particular concern do not engage in sensation seeking or play?”)

Will Aldred Dec 27, 2023, 6:09 PM
25 points
8 ∶ 2
on: Only mammals and birds are sentient, according to neuroscientist Nick Humphrey’s theory of consciousness, recently explained in “Sentience: The invention of consciousness”
Upvoted because I’m a fan of people summarizing and signal-boosting literature that bears on EA priorities. Disagree-voted because I’m not convinced that any of the observations or considerations put forward support the headline claim that only mammals and birds are sentient.

Will Aldred Dec 25, 2023, 5:06 PM
7 points
2 ∶ 0
in reply to: Daniel_Eth’s comment on: The US plans to spend $1.5 Trillion upgrading its Nuclear Missiles!!
I’m sure ~everyone involved considers nuclear war a negative-sum game. (They likely still think it’s preferable to win a nuclear war than to lose it, but they presumably think the “winner” doesn’t gain as much as the “loser” loses.)
On top of this, I imagine most involved view not fighting a nuclear war as preferable to fighting and winning. (In other words, a nuclear war is not only negative on net, but negative for everyone.)^[1]
Yeah, my sense is multiple countries will upgrade their arsenals soon. I’m legitimately uncertain whether this will on net increase or decrease nuclear risk (largely I’m just ignorant here – there may be an expert consensus that I’m unaware of, but I don’t think the immediate reaction of “spending further money on nukes increases nuclear risk” is obviously necessarily correct).
I previously did some work (with/under Michael Aird) on the effects of nuclear weapons advances on nuclear risk. There’s no expert consensus I’m aware of: for many advances there are a bunch of considerations going in both directions.
One example of an advance that I’m somewhat confident would decrease risk is more accurate nuclear weapons. The main reason: nukes being more accurate means that fewer nukes, and/or nukes with lower explosive yields, are needed to hit the intended target. The effect of this is fewer direct casualties; also—and more importantly for x-risk—less soot generated, hence less of a nuclear winter effect.
(Tagging the OP, @Denis, in case my comment or the post I link to is of interest.)
1. ^
  This raises the obvious question, “Why fight at all?” As best I’m aware, the answer to that lies with things like false information (e.g., false alarm triggering a second strike that’s actually a first strike), and also with some artefacts of game theory (e.g., brinkmanship-gone-wrong; bargaining breakdown due to misevaluating how the opponent sees things; etc.) as well as the reality that actors don’t always behave rationally.

Will Aldred Dec 21, 2023, 6:22 PM
43 points
3 ∶ 1
in reply to: Geoffrey Miller’s comment on: Effective Aspersions: How the Nonlinear Investigation Went Wrong
Writing in a personal capacity.
Hi Geoffrey, I think you raise a very reasonable point.
There’s some unfortunate timing at play here: ³⁄₇ of the active mod team—Lizka, Toby, and JP—have been away at a CEA retreat for the past ~week, and have thus mostly been offline. In my view, we would have ideally issued a proper update by now on the earlier notice: “For the time being, please do not post personal information that would deanonymize Alice or Chloe.”
In lieu of that, I’ll instead publish one of my comments from the moderators’ Slack thread, along with some commentary. I’m hoping that this shows some of our thinking and adds to the ongoing discussion here.^[1] I’m grateful to yourself, @Ivy Mazzola, @Jason, @Jeff Kaufman and others for helping push this conversation forward.
Starting context: Majority of moderators in agreement that our current policy on doxing, “Revealing someone’s real name if they are anonymous on the Forum or elsewhere on the internet is prohibited” (link), should apply in the Alice+Chloe case. (And that it should apply whether or not Alice and/or Chloe have exaggerated their allegations.)
Will (5 days ago)
I think what this comes down to for me is: If Kat Woods’ Forum username was pseudonymous, would we have taken down Ben’s post? (Or otherwise removed all references to Kat by her real name?)
If the answer to this is “yes,” then I don’t think Alice+Chloe should be deanonymized.
If the answer to this is “no,” then I think Alice+Chloe should be deanonymized.^[2] (Because if we go with “no,” then this would mean that it’s fair game for Kat to go write a post now that shares information on two past employees whom she believes have spread falsehoods about Nonlinear, using these employees’ real names. Which is equivalent to deanonymizing Alice+Chloe.)
---
I’m concerned about setting a precedent of first-mover advantage. Like, I imagine there are EAs with mutual grievances out there, and if we set a precedent of, “If you strike first, anonymously, then you can name the person you bear a grievance against whilst granting yourself anonymity immunity,” then I think we’re laying the foundations for a pretty awful dynamic.
There was then commentary from a couple of the other moderators acknowledging that:
- Under our current policies, it is indeed fair game for Kat to write a post sharing information on those former employees using their real names, as long as that post doesn’t refer directly to Ben’s one. Which does seem like a weird technicality.
- The three broad paths forward seem to be: 1) “give the first accuser the right to remain anonymous;” 2) “ban anonymous allegations altogether;” 3) “stick with our current policies, notwithstanding that there is that technicality.^[3]”
  - There’s since been the beginnings of a discussion on whether certain categories of anonymous allegations—notably, employees sharing information on employers—could be exceptions to the rule if the second path is taken.^[4]
Closing context: At the time of me posting this comment, inter-moderator discussion on the issue is inconclusive.
1. ^
  The thinking shown will of course be skewed towards my personal view, over the views of other moderators.
2. ^
  A later comment of mine, which ties in here:
  I think it probably makes most sense to decide this at the policy level first, and then circle back to how to handle the Nonlinear case. Specifically, if we go with giving the first accuser anonymity rights at the policy level, then it follows that Alice and Chloe should remain anonymous. But if we decide that allegations should not be made anonymously, then we’ll have to think of what to do with the Alice+Chloe allegations already out there.
  [ETA: Like, since Alice and Chloe made their allegations expecting anonymity, I believe we should give their cases special treatment (if we go the route of not allowing anonymous allegations). This could just mean fully respecting their expectation. This could mean making difficult judgement calls in weighing up, for example, Alice’s expectation for anonymity versus the magnitude of her exaggerations.]
3. ^
  And notwithstanding that this technicality may well put an accused party in an uncomfortable position, where they believe they have good reason for writing a response “sharing information” post, but they also know that doing so likely means receiving some backlash (for implicitly deanonymizing their accuser).
  My personal view is that this rules out the third path: I don’t think a fair set of policies would force someone into this kind of a lose-lose position.
4. ^
  In other words, one consideration is whether the benefit of allowing anonymous whistleblowing on the Forum—where the purpose of whistleblowing is in part to compensate for employer-employee asymmetry—outweighs the cost of not having first-accuser-second-accuser symmetry.

Will Aldred Dec 18, 2023, 5:19 PM
4 points
1 ∶ 0
in reply to: Linch’s comment on: The Unweaving of a Beautiful Thing
(I only came across this post because I saw your comment; agree that it’s breathtakingly moving :’) If the author sees this: I appreciate you, and I too would definitely subscribe.)

Will Aldred Dec 16, 2023, 1:49 AM
14 points
3 ∶ 0
in reply to: lilly’s comment on: Nonlinear’s Evidence: Debunking False and Misleading Claims
Writing in a personal capacity.
“An update to our policies on revealing personal information on the Forum” covers some of what you’re asking about, I think, although the framing there is more about revealing private vs public info than about “How substantiated is substantiated enough?” The most relevant part:
1. We think a very good norm is to check unverified rumors or claims before sharing them — especially if they might be damaging or if they relate to sensitive or stigmatized topics.
  If you’re not sure whether you should check something (or how to check), you can contact the moderation team to ask.
2. If you think that some information should be removed, you should flag this to us. We will probably not remove information that no one has asked us to remove.
  (We don’t read everything on the Forum, and when we are reading, we’re not always thinking about everything through the lens of our policies.)
3. Why we don’t just default to removing all private/personal information: we think there are cases when some personal information about people who are highly relevant to work in effective altruism is important to share (like discussions of potential conflicts of interest (COIs) or reasons for why someone in a position of power shouldn’t be in that position). We also want to keep the potential for censorship from the moderation team low.
4. The way we enforce these norms isn’t about whether we think a specific comment is “overall correct” or helpful, etc.; we’re trying to outline policies that will help us make these calls more objectively.
(I recognize that the above isn’t really granular enough to help with answering the questions in front of us right now, for instance, “Was it okay for Ben to include Alice’s allegations?” Or, “Is Kat’s ‘Sharing Information on Ben Pace’ section acceptable?” Thanks for raising this point for the moderators, it will be discussed. (My immediate take: I can see a couple of reasons for why it might be hard to operationalize, and to enforce, a set of policies here. But, as you say, this does seem important enough to warrant some thought.))

Will Aldred Dec 15, 2023, 3:54 PM
8 points
0 ∶ 0
in reply to: lilly’s comment on: Lizka’s Shortform
Yeah, this is a reasonable thing to ask. So, the “if we have reason to believe that someone is violating norms around voting” clause is intentionally vague, I believe, because if we gave more detail on the kinds of checks/algorithms we have in place for flagging potential violations, then this could help would-be miscreants commit violations that slip past our checks.
(I’m a bit sad that the framing here is adversarial, and that we can’t give users like you more clarification, but I think this state of play is the reality of running an online forum.)
If it helps, though, the bar for looking into a user’s voting history is high. Like, on average I don’t think we do this more than once or twice per month.

Will Aldred Dec 15, 2023, 2:31 PM
6 points
1 ∶ 0
in reply to: lilly’s comment on: Lizka’s Shortform
Writing in a personal capacity; I haven’t run this by other mods.
Hi, just responding to these parts of your comment:
I think people might reasonably (though wrongly) assume that forum mods are not monitoring accounts at this level of granularity, and thus believe that their voting behavior is private.
...
Frankly, I don’t love that mods are monitoring accounts at this level of granularity. (For instance, knowing this would make me less inclined to put remotely sensitive info in a forum dm.)
We include some detail on what would lead moderators to look into a user’s voting activity, and what information we have access to, on our “Guide to norms on the Forum” page:
Voting activity is generally private (even admins don’t know who voted on what), but if we have reason to believe that someone is violating norms around voting (e.g. by mass-downvoting many of a different user’s comments and posts), we reserve the right to check what account is doing this. If we suspect that someone is using multiple accounts to vote on the same post, we also reserve the right to check whether the accounts are related, and check their voting history.
...
The following information is accessible to moderators but will only be used to identify behavior such as “sockpuppet” accounts and mass downvoting, in situations where we have strong reason to believe that an account is used to get around a ban (or other restriction), or in the case of severe safety concerns. The moderators will not view or use this information for any other purpose.
- The IP address a post/comment came from
- The voting history of users
- The identity of voters on any given post/comment
(In addition, note that moderators can’t just go into a user’s account and check their voting history even when we do have reason to look into that user. We require one of the Forum engineers to run some queries on the back end to yield this information.)
Finally, to address your concern about direct messages on the Forum: like a regular user, a moderator cannot see into anyone else’s messages.
Hope this is helpful :)

Will Aldred Dec 12, 2023, 6:47 PM
132 points
27 ∶ 1
on: Nonlinear’s Evidence: Debunking False and Misleading Claims
Hey folks, a reminder to please be thoughtful as you comment.
The previous Nonlinear thread received almost 500 comments; many of these were productive, but there were also some more heated exchanges. Following Forum norms—in a nutshell: be kind, stay on topic, be honest—is probably even more important than usual in charged situations like these.
Discussion here could end up warped towards aggression and confusion for a few reasons, even if commenters are generally well intentioned:
- Some of the allegations this post responds to, and the new allegations in the post, are serious and upsetting. People who have gone through similar experiences may find engaging with this topic especially stressful.
- Power dynamics and alleged mistreatment is rightly an emotionally loaded topic, and can be difficult to discuss objectively.
- Differences in personal culture and experiences can lead to hard-to-articulate disagreements over acceptable versus unacceptable behaviour.
Regarding this paragraph from the post:
Given what they have done, a number of people expressed to us that they think Alice/Chloe are a danger to the health of the community and should not be anonymized. We will leave that to the community to decide.
For the time being, please do not post personal information that would deanonymize Alice or Chloe. The moderation team is currently discussing how we might deal with this.
Also, don’t vote using multiple accounts, and don’t engage in vote brigading. We will be monitoring voting activity. We had to ban users for voting violations on the “Sharing Information About Nonlinear” post, and we would like to avoid issuing bans again.
What links here?
- Will Aldred's comment on Effective Aspersions: How the Nonlinear Investigation Went Wrong by TracingWoodgrains (Dec 21, 2023, 6:22 PM; 43 points)
- Lorenzo Buonanno🔸's comment on Effective Aspersions: How the Nonlinear Investigation Went Wrong by TracingWoodgrains (Dec 20, 2023, 3:44 PM; 30 points)

Will Aldred Dec 12, 2023, 6:19 PM
47 points
25 ∶ 14
in reply to: John G. Halstead’s comment on: Nonlinear’s Evidence: Debunking False and Misleading Claims
Edited to add: My objection to John’s comment in what I write below lies with the “deranged” part. If John had instead said something like “unnecessary” or “overly escalatory/ad hominem,” then I would not have responded. But “deranged” — dictionary definition: “completely unable to think clearly or behave in a controlled way, especially because of mental illness” (source) — which I take as John implying that the direction Kat has gone in is so completely nonsensical that there can’t possibly be a reasonable explanation, struck me as sufficiently inaccurate for the opening assertion in such a highly upvoted comment that I felt the need to weigh in.
I think Kat could reasonably claim that, from her perspective, Ben has opted out of the social convention around not damaging someone’s reputation through less-than-solid allegations, so she is now fighting fire with fire.
I’m not saying I agree with Kat’s move here [edited to add: and I would personally prefer it if Kat had focused solely on engaging, in a factual manner, with the evidence Ben put forward], but I think there’s a frame in which it makes sense, and therefore it seems unfair to label this move “deranged.”

Will Aldred Dec 12, 2023, 3:31 PM
26 points
3 ∶ 0
on: Who is Sam Bankman-Fried (SBF) really, and how could he have done what he did? - three theories and a lot of evidence
the actions he [SBF] was convicted of are nearly universally condemned by the EA community
I don’t think that observing lots of condemnation and little support is all that much evidence for the premise you take as given—that SBF’s actions were near-universally condemned by the EA community—compared to meaningfully different hypotheses like “50% of EAs condemned SBF’s actions.”
There was, and still is, a strong incentive to hide any opinion other than condemnation (e.g., support, genuine uncertainty) over SBF’s fraud-for-good ideology, out of legitimate fear of becoming a witch-hunt victim. By the law of prevalence, I therefore expect the number of EAs who don’t fully condemn SBF’s actions to be far greater than the number who publicly express opinions other than full condemnation.
(Note: I’m focusing on the morality of SBF’s actions, and not on executional incompetence.)
Anecdotally, of the EAs I’ve spoken to about the FTX collapse with whom I’m close—and who therefore have less incentive to hide what they truly believe from me—I’d say that between a third and a half fall into the genuinely uncertain camp (on the moral question of fraud for good causes), while the number in the support camp is small but not zero.^[1]
1. ^
  And of those in my sample in the condemn camp, by far the most commonly-cited reason is timeless decision theory / pre-committing to cooperative actions, which I don’t think is the kind of reason one jumps to when one hears that EAs condemn fraud for good-type thinking.