I’m less approving of the EA community now than before the FTX collapse
This is mostly a counterpoint to Will Aldred/Duncan Sabien’s post.
I’m not really an EA; haven’t taken the pledge, don’t work at an org, have been to no EA meetups in my life, (haven’t been a speaker at EA Global).
However, I have been close to the EA community for a while, agreeing with many of its goals and donating to many of its key causes.
Reading Duncan’s post made me want to defend the case for why I’m less approving of the EA community than before. I want to be a little specific about what I’m less approving of.
EA™ - CEA / EVF (I’d never heard it called EVF before FTX)
Will MacAskill personally
Donation practices during the “funding overhang” era
There are some people/orgs I am more approving of:
Peter Wildeford (and Rethink Priorities more generally)
Rob Wiblin
Dustin Moskovitz
I’m not going to say anything about Peter/Rob/Dustin in this post, although the amount I approve them more does not change the net effect, which is less approving of EA.
Lots of the reasons I am less approving of EA now than I was prior to FTX collapse are things I could have known. However I am aware of them because of the FTX collapse. Others might have already been aware of everything I’ll mention, in which case I would agree with Duncan—with few exceptions, the communities reaction to the FTX collapse has been very good and I largely approve.
So, what have I learned since the FTX collapse which makes me approve less of EA?
Will MacAskill:
Initial reactions thread I read this as both minimizing the event and distancing himself from SBF, which is not credible in light of:
Elon relationship—you don’t go and bat for someone with the richest guy in the world unless you are confident in who you’re batting for
Their long history of a close relationships (shared board memberships, his earlier mentorship etc)
Guzey’s review (this episode is why I’m posting as a throwaway)
Equivocal statements during the early crisis:
Holder Karnofsky (note the edit and Alexander Berger’s comments before blaming Holden)
There were more statements I was annoyed with at the time, but I can’t remember enough of the specifics to search them right now. If I remember them, I will add them here. (Part of my praise for Rob was his statement was in sharp contrast of what was coming out at the time)
Wytham Abbey—I don’t especially want to relitigate this but it makes me think less of EA—you can read many takes on EAF—probably the fairest place to start is the reasoning for why it was bought.
People who received funding in dubious ways and didn’t say anything (North Dimension donating on behalf of FTX Future fund)
EA (generally) treating FTX as something which “happened” to them, rather than being something they were intimately involved in. (Initial funding for Alameda coming from the EA community, …)
I suspect (but can’t prove) that the EA orgs are using the excuse of legal advice / legal threats to avoid saying things they don’t want to say for other reasons. (And my suspicion is that this is because they have done something which could be bad)
People in the community knew SBF’s image was manufactured to some degree but didn’t say anything (general take from forum—summarized in this New Yorker piece)
EA’s earlier relationship with a sketchy billionaire (and the degree to which this was covered up)
All of these things put together are enough for me to downgrade my opinion of EA. I’ve put this together from off the top of my head, so there are likely other things which have affected my view over the last month. To be clear—I am still very much in favor of EA principles, the EA community and will continue to donate to EA causes but will be a little more cautious around trusting the community’s view of people and especially cautious around organizations linked to CEA / Will MacAskill.
Edit 1 2022-12-18: Fixed typo in link
Thank you sharing this. As a distinct matter, the specific way FTX failed also makes me more concerned about the viability of a certain type of mindset that seems somewhat common and normalized amongst some in the EA community.
holding the belief that by being very very smart, you can work in areas you have minimal experience and know better than others
having (experienced) adults in the room/adhering to formal compliance norms is overrated
understating the risks posed by conflict of interest issues
accepting ends justify the means type reasoning
I believe Sam’s adherence to the above referenced beliefs played a critical role in FTX’s story. I don’t think that any one of these beliefs is inherently problematic, but I have adjusted downwards against those who hold all of them.
While I agree with the substance of this comment to a great extent, I want to note that EA also has a problem of being much more willing to tolerate abstract criticism than concrete criticism.
If I singled out a specific person in EA and accused them of significant conflicts of interest or of being too unqualified and inexperienced to work on whatever they are currently working on, the reaction in the forum would be much more negative than it was to this comment.
If you really believe the issues raised in the comment are important, take it seriously when people raise these concerns in concrete cases.
This comment exactly proves the point I’m talking about: it gets upvoted because it’s an abstract comment that doesn’t blame anyone in particular of anything. My comments that are critical of or adversarial towards specific people perform significantly worse, and plenty of my comments have fluctuated between having −10 karma to +10 or +20 karma.
I’m reasonably confident this is happening because people in EA hate conflict in general. I’m sorry to say that genuine criticism has to involve conflict of some kind. I understand people in EA are not used to this and prefer to shut their ears while accusing their counterparty of being “insufficienty truth-seeking” or “acting in bad faith” whenever they make use of adversarial discourse, and indeed this is exactly what happened to Guzey’s post from 2018.
I wish EA could move past these dysfunctional social norms, but I’m not getting my hopes up.
Relevant:
DMMF—I share your first concern, that many young EAs seem to have quite a bit of ageism, distrust of legacy systems, contempt for tradition, arrogance about being able to reinvent complex systems from first principles, and wariness of welcoming mid-career and senior experts into the EA community.
This youthful arrogance is partly merited. Traditional charity circa 2000 (before GiveWell) was an unaccountable, unempirical grift that was better suited to virtue signaling, status seeking, and wealthy in-group power games, than to actually helping people or animals. EA challenging this legacy system was entirely right, proper, and helpful. Consequentialist moral philosophy, sentientism (eg about animal welfare), randomized trials, cost-benefit reasoning, scope-sensitivity, etc were crucial and revolutionary developments at the foundation of EA.
However, applied to other domains, subcultures, professions, legacy systems, and cultural traditions, that youthful arrogance can be quite misguided. Especially in the domain of financial accounting, corporate management, and business ethics. SBF seemed to think that he could ignore all the accumulated wisdom of the past few centuries about how to run a company, and protect users and investors. The result was, from the perspective of experienced business managers, risk managers, and accountants, and highly predictable catastrophe (for anyone who actually knew the inner workings of Alameda/FTX).
So, one takeaway from this whole FTX debacle is that maybe EAs should be a little more selective about when to challenge tradition and try to reinvent things from first principles, versus when to respect tradition, expertise, and domain knowledge.
This is a tricky and delicate balancing act, but I think science provides an example of how to get that balance roughly right. Young scientists in grad school quickly learn that almost every new idea that they think is earth-shaking and ground-breaking was already thought about by the late 1800s, or 1950s, or whenever. With great scholarship comes great humility. Only after they’ve mastered a fair amount of the literature in some domain are they equipped to make even modest contributions at the cutting edge of thinking. And only in the rarest cases can a true genius overturn a major established paradigm.
The early EA geniuses managed to overturn the existing charity paradigm, and that was great. But with that initial success, we must be wary of developing a generalized hubris about overturning every tradition we see, and challenging every expert we meet.
This comment by Geoffrey Miller is one of the most insightful things I’ve read. I had a lot of these feelings, but lack the STEM background to say it as you have. EA did a great very specific thing, but then it went to their head and they saw themselves as able to do all things better. Hopefully this crisis will be the pivot where they see the difference and re-gather past collective wisdom where it’s still important.
And the fastest route to that is to bring in more seasoned veterans, just import the data straight in through their bodies in EA org’s.
Jeffrey—thanks for your kind comment! Appreciate it.
I agree and I am also concerned about this. I have witnessed this many times. I do think there are tremendous merits in vigorously thinking from first principal on some subject matters. But others such as risk management and regulation for example do require experise as we have now seen in the case of FTX.
Maybe others were following developments more closely than me / knew more stuff in advance than I did, but from my perspective Will’s reaction came out pretty early, at least slightly before I had seen enough discussion to feel confident that there had been serious wrongdoing at FTX. I remain unclear about various important details, and about how to divvy up fault between various key players at FTX and Alameda.
If as of 3:55pm PT on Nov 11 you were already extremely confident on these points, then I can better understand your view. But speaking as someone who needed to hear more discussion first—and as someone who thinks of himself as having the opposite views from Will MacAskill on almost every EA debate! -- Will gained some points in my book for being willing to say “I am not certain that this is what happened” and “If new information vindicates FTX, I will change my view and offer an apology” at the start of an obvious “find and burn the witches” social-media wave. Even when a witch-hunt-ish social process has in fact found a witch, it’s still important at the process level that people feel safe expressing their actual uncertainty, saying things that go against the social gradient, etc., or you end up with echo chambers that socially filter out all counterevidence and calls for moderation.
FWIW, I agree with Duncan’s post, and Wytham Abbey is a positive update for me about Owen/EAV/OP’s judgment. I share some of your other worries (did Will really not know more? why didn’t EA share more info about the early Alameda conflicts? are the legal risks really a good reason not to say more now?), but I don’t have private information, so I mostly still feel in-the-dark about those topics.
I can confirm that both me and the author of the post (who are separate people) were extremely confident that there had been serious wrongdoing at FTX as of 3:55pm PT on Nov 11.
I think EA’s failure to be extremely confident of this given the strength of the available evidence was extremely problematic, and I’ve seen EAs expressing skepticism that Sam Bankman-Fried was involved in any wrongdoing (other than just being incompetent) way after Nov 11.
I can tell you that I don’t feel safe expressing my views about how broken EA norms are in public under my own name, and I find MacAskill gaining positive points in your book to be absurd given MacAskill’s personal involvement in suppressing inconvenient information about the activities of FTX and Alameda executives in the Bahamas.
That’s good to know! I’m not shocked if some people were rationally confident of this at that time, since other people may have known more than me.
How long after / can you give an example of what you have in mind? At a glance, it sounds to me like you’re interacting with a different subset of EA than I’ve seen.
I said that he gained positive points from me for the tweet throwaway791 linked to, not that he’s net positive in my books for the entire FTX saga. It’s pretty important that those two questions not be conflated, given how halo/horns bias (and social pressures that reinforce the bias at the group level) works, and how soldier mindset works in general.
You think Alice is bad, so you punish anyone who says “I agree with Alice about X” (even if you too think Alice is right about X) or “I think it was good that Alice did Y” (even though you agree), because you want to drive down Alice’s reputation as far as you possibly can, and you’re worried that any positive thing said about Alice could be used strategically to protect Alice from proper punishment.
This is a really bad system, one that almost always leads to wild misappraisals of exactly how good or bad something or someone is, and that unnecessarily frequently leads to villains being idolized and innocents being vilified, because the group doesn’t trust itself to do a fair and accurate accounting of each local incident, and people hold back their impressions and their data lest they be punished for giving aid and comfort to the Wrong Side.
Indeed, I’d claim that this is one of the key errors of process that protected SBF before things went to shit. When SBF is good, sharing any contrary take or data point is punished; and when he’s bad, it’s likewise punished, just with a different assumption about which take is wrong.
I don’t see any value in starting down that road; we can just evaluate evidence on its own terms, without tilting the scales in a confirmation-bias-y way.
Separately: What is “MacAskill’s personal involvement in suppressing inconvenient information” alluding to?
Not only was I not talking about whatever you’re talking about—I don’t even know what you have in mind.
Sadly I can’t give any explicit examples without giving people material information about my identity that I would not be comfortable revealing. I agree that it’s possible I’ve been interacting with a different subset of EA. I also can’t share the exact date but you can assume it was some weeks after Nov 11.
I assume you know that it’s possible for E[X] < E[X | A] and E[X | B] > E[X | A, B] to hold simultaneously. I agree that to a totally uninformed observer MacAskill’s tweet thread could be slightly positive evidence, but if you put his tweets within the broader context of his activities connected to Sam Bankman-Fried in the past decade or so, they give cause for an update in the opposite direction.
MacAskill has previously been dishonest by pushing the misleading image that Bankman-Fried had a frugal lifestyle despite knowing better, and he has a further track record of instrumental dishonesty e.g. when promoting What We Owe the Future to the public. In general, EA often covers up facts about the philosophy or community that would not be palatable to a general audience, and MacAskill has engaged in such behavior far too frequently for me to enumerate all examples of it here.
For him to say that honesty is very important is laughable and is only further evidence of his dishonesty. If he wants to provide Bayesian evidence that he’s actually honest, he should try being honest instead of linking passages from his old books that extol honesty as a virtue. Talk is cheap, action is not.
How is it punished to share anything positive about SBF? I could post on this forum under my real name that SBF had a positive impact in some areas despite him running the equivalent of a Ponzi scheme and I wouldn’t be concerned at all. Your impression of which claims are punished and which claims are not punished is wildly miscalibrated.
I agree that the error you’re talking about is one of the key errors that protected SBF, but I don’t agree that it’s now operating in the opposite direction with anything like the same intensity.
I know for a fact that MacAskill knew SBF’s “frugal lifestyle” image was fabricated and he fully assisted him in perpetuating this myth, effectively acting as a PR agent for him. I also suspect MacAskill knew a substantial amount of information about events that happened in early Alameda and had good cause to suspect SBF’s character. He also knew that FTX operated in a shady way in which they would deliberately conceal relevant facts about their business from the venture capitalists who had invested in FTX. He chose to reveal none of this.
I don’t think he knew anything about the fraud, but he knew that attaching EA’s reputation to a figure like SBF was a reckless move with serious potential downsides even if SBF had ultimately run a legitimate business and donated significant resources to EA causes. He didn’t disclose these details in public because of his willingness to be instrumentally dishonest whenever he thinks dishonesty furthers the Cause.
OK! I’m sad about that, and very much want to make EA a space where people can rationally spill their guts more and not fear big social costs.
Hmm. In the quoted article he says:
“Heavily considering what you show as well as what you do, especially if you’re in a position of high visibility. ‘Signalling’ is often very important! For example, the funding situation means I now take my personal giving more seriously, not less. I think the fact that Sam Bankman-Fried is a vegan and drives a Corolla is awesome, and totally the right call.”
This could mean ‘I think it’s great that Sam is so frugal overall (with the Corolla being just one example), since that sends a great signal’, but a more literal reading is just ‘I think it’s great that Sam was frugal in one particular way (the Corolla), because this sends a great signal’. Like, I come away with
twothree questions:When Will wrote this (in May 2022? sometime earlier?), did he know how unrepresentative the Corolla was in a ton of ways?
(
I assume so, butPlausibly so, but it would be nice to confirm?)If Will knew the Corolla was unrepresentative, did he intend to deceive the reader into thinking the Corolla was representative?
(This isn’t obvious to me at all. If Will knew SBF was unfrugal and had praised SBF in the article for his frugality, then I’d go “yeah, trying to trick the reader”. But the thing Will actually praised was SBF’s signaling, not his frugality; this is not the phrasing I’d have chosen if I were trying to trick readers into thinking SBF is amazingly frugal.)
Did Will think it mattered how representative the Corolla is, for purposes of signaling to the larger public?
Like, a cynical take on the Corolla is “SBF is trying to trick people by being frugal in one or two minor ways and then plastering the Internet with those cherry-picked examples, while being totally unfrugal otherwise”.
A less cynical take is “SBF is willing to spend huge amounts when he thinks it’s useful for EA, but a nicer car wouldn’t actually help even a tiny bit with any EA thing, so he picked the Corolla just on the merits, not trying to paint a coherent picture or push a narrative”. Like, it’s not as though it would have been more virtuous for SBF to buy a fancier car.
If Will had the cynical model of SBF’s motives, then I definitely object to him singling this out as a good example of EA virtue signaling; good EA virtue signaling should be honest, not dishonest and cherry-picked!!
But if Will had the less-cynical model, then I don’t actually object to him mentioning it to other EAs, as a good thing for us to emulate. I do think that letting the Corolla example catch on and go viral, without regularly chiming in with examples of SBF being less frugal (given that it’s in fact unrepresentative!), would be very bad. But I don’t actually know the details here, and if he just mentioned it in an 80K page the one time and wasn’t running around bragging about it to everyone, and didn’t have the cynical model, then I’m not bothered. Examples of marginal virtue (“Bob did a good thing that one time!”) are fine, as long as they don’t metastasize into false “that good thing is a major defining feature of Bob!” mythologies.
I’d be interested to hear some concrete examples? I expect Will to spin and cherry-pick some things in ways that I wouldn’t, but you might have something more serious in mind (or less serious, if your bar for instrumental dishonesty is low enough).
This argument seems structurally valid to me. Like, if you think that a friend of yours is a drug addict, then hearing them claim to have hardline anti-drug views out of the blue could be a negative update (“oh jeez, they’re lying now too”), where hearing them come clean and talk about their past mistakes would update you to thinking they may be on the path to changing their behavior.
To an innocent who has no idea Carol is an addict, the drug confession would be a shocking negative update; to the person who knows Carol better, it’s an important positive update.
I’d indeed expect this to cost you more than if you just wrote an article criticizing SBF. Not sure how to operationalize this into a bet.
Interesting!
To be clear, I’m not making a claim about the exact intensity levels. I just think that there was a large pressure to only say positive things before, and that there’s a large pressure to only say negative things now. I don’t feel confident about which of those pressures was larger.
I’d love to see a longer write-up somewhere that provides more evidence and details about any or all of these claims—I don’t know who you are and don’t currently know how much weight to put on them. I could see them turning out to be totally true, or totally false, depending on how honest and rigorous you (and/or your sources, if this isn’t direct info) are.
There are plenty of views held by EA leadership that would be bad if they were to be brought up on, say, Trevor Noah’s show. Do you really want explicit examples here? I’m sure you know exactly what I’m talking about.
I don’t think this by itself is that problematic because it’s a common mode of dishonesty, and if it were the only problem I would think MacAskill is just engaging in marketing, where deception can be a prerequisite for success. However, for me this is only one piece of the puzzle.
The reasons I’m suspicious of MacAskill in particular are:
His willingness to deceive both the public and ordinary EAs on many occasions: about the views of EA leadership and his own views in particular, about the fact that SBF’s “frugal lifestyle” image was fabricated, potentially about events that took place at early Alameda, etc.
His close relationship with SBF: the fact that he was the one to offer SBF an EA job after SBF left Jane Street, the support given by EA organizations to Alameda Research early on, including both substantial financial resources and human capital; along with e.g. his intermediation between SBF and Elon Musk
The shady nature of Future Fund grantmaking channels when he would have been in a position to know who received funding from where
The Guzey confidential draft leak episode
Vague “bad rumors” about him that have been circulating recently—I can’t elaborate on this further because I also don’t know the origin of these rumors, but you might hear such things if you go to real-life EA meetups and it still seems like relevant Bayesian evidence
Conditional on the above, the fact that his initial reaction to the FTX scandal was both delayed and ambiguous makes him look even worse
None of this is conclusive evidence, but together they paint a picture of the utilitarian who is convinced he is doing good and who has no problems employing deception and lying as part of his toolkit.
I have some more evidence that I can’t share here due to confidentiality concerns, but none of it is more decisive than the public evidence I’m able to share above. I think the evidence is clear that MacAskill is instrumentally dishonest in many important situations, and there’s unclear but suggestive evidence that he was dishonest about SBF’s character or information he may have had about the leadership of FTX and Alameda Research.
If you have a highly favorable view of MacAskill, the above is probably insufficient to update you to an extent that you would agree with me about his character. In this case I would implore people to have more reasonable priors and to hold EA leadership to higher standards.
I concede that this speaks for “he isn’t always careful with what he conveys.” However, for the reasons Rob Bensinger explained in one of his comments, I think this could be more innocent. And there’s another likely explanation of this that is also pretty innocent. It’s a cool story with the cheap car, it’s an image that “sticks.” Maybe he just didn’t notice confusion, didn’t say “oops” when he saw things that were in tension with this mental image of Sam. I think this happens pretty commonly. (E.g., let’s say you see your parents about twice per year and they remember from when you used to live with them that you’re vegan. Then, they see you eat animal products a couple of times when you visit. The parents may not update strongly enough on the obvious evidence that “vegan” is no longer accurate, so they may tell other people that their son is vegan by accident because that’s the mental image they’ve “saved you under.” That doesn’t mean the parents are deceivers covering up your relapse! They simply aren’t paying close attention and are slow to update on the image that’s still salient in their mind.)
I think this is referring to things like allegations that Sam tricked co-founders when he registered entities in his own name rather than theirs, that he seemed reckless with risks, and that he was bending numbers with accounting when this was favorable for him/for the company? I have no idea if Will had information about this. It seems possible to me, but I don’t know! If he knew about it, then it really makes you wonder why he vouched for Sam. In my view, the vouching was already the worst thing I’m aware of because if you vouch for someone, you’re saying that you know them well enough that you would know if something was amiss (so ignorance of some bad past behavior is not even a good excuse). However, I think you’re being uncharitable when you imply that knowing about the early Alameda stuff (hypothetically!) would mean that Will was deceiving people about Sam’s character. It’s about twenty times more likely, IMO, that rather than thinking Sam has bad character, he simply didn’t update enough on the evidence (if he at all knew). This is also a common pattern: people observe red flags but the flags go against the mental image they have formed of the other person as someone decent, so they allow the bad actor to make excuses and explain away any incongruencies. For instance, I’m sure Sam had his own story about what happened. Let’s say Will heard rumors and asked Sam about it. Sam would have said something that sounded reassuring. He probably would have said stuff like “Well, one of the other co-founders was a bit of narcissist and we generally didn’t get along. And they seemed unusually risk-averse and not made for the crypto environment.” I’m just making stuff up here, but you can see how it’s possible that there were two sides of the story. I’m not saying it would have been smart to dismiss any red flags, but that’s often how things go. Many people are not cynical enough (and sometimes bad actors are really good at spinning up a counter-narrative).
Nothing about this seems in any way bad to me. It’s not always easy to tell how someone’s going to behave. Also, maybe Sam got “corrupted” over time and some tendencies became worse as he got more successful? At best, I count this as evidence of “not great people judgment,” but sometimes you get unlucky. Not every time something goes wrong it means that you made a horrible mistake ex ante.
He may not spend much attention on following or understanding logistics like that. Even if he saw that the bank account wasn’t labelled “FTX future fund,” he may have thought there was some random logistical reason behind it. This really seems like the sort of thing that only looks suspicious with hindsight. Or, okay, maybe it should look suspicious to someone who has accounting/legal knowledge (it wouldn’t have been salient to me that “money comes from a different entity” is a potential red flag, but maybe this would be salient to someone who does accounting or law). For what it’s worth, I’m definitely sympathetic to the claim that someone at the FTX future fund should have been on top of the situation and should have done better due diligence. However, my guess would be that Will’s personal role there wouldn’t have made him responsible for this.
Overall, I agree that Will looks bad in many ways, so please don’t take my comment as saying that there’s “nothing to see here.” Still, the thing I mainly want to push back on is the idea that Will personally thought Sam was probably committing fraud or at least that he was otherwise a horrible person and he (Will) was okay with that. I could be wrong, but I really doubt this. I think all of Will’s behavior so far (that I’m aware of) is easily explained by “he had bad people judgment + maybe had a tendency to believe things that are convenient for him.” It’s possible that you know things I don’t know, but I can’t take your word for it with regard to the things I don’t know, given that we seem to disagree about how to interpret the evidence that’s already been openly discussed.
All you’re saying, as far as I can see, is that the available evidence is consistent with MacAskill simply having made mistakes, having bad judgment, not making deductions that he reasonably should have made, et cetera. I agree with that, which is why I noted in my original comment that
I just don’t understand why people have the prior that obviously MacAskill should be a nice person who always acts in good faith. The circumstantial evidence I’ve outlined is insufficient to overcome such a strong prior, and you can always explain it away by attributing any number of honest mistakes and oversights to MacAskill.
In addition, I don’t think the way you’re casting the issue is appropriate. I think MacAskill didn’t spend his days thinking about how SBF was a horrible person who EA nevertheless needed because of the resources at his disposal, and I don’t believe he knew about the fraud. I think only some of his deception is conscious and deliberate. I suspect his failures in many cases looked more like “choosing not to know” or “choosing not to focus on” certain things, but I also don’t think these choices were entirely innocuous as you’re making them sound. Often when you choose not to know something it’s because of self-serving or otherwise immoral reasons.
I don’t believe for a second that he pushed SBF’s frugal image because he wasn’t careful with the information he conveys. I think this was a fully deliberate decision on his part and you attempting to make it look otherwise requires stories which are, on their face, much less plausible than that.
As for other issues I raised, I believe that some of it was not deliberate on his part. For instance, I don’t think MacAskill concocted a conscious scheme to try to kill Guzey’s post. He simply felt personally attacked by the hostility in the review and those emotions got him to conceal the fact that Guzey’s private draft had been leaked against what should have been his better judgment. Still, what people do in such situations is often informative about how they generally act when under pressure, and MacAskill seems to have fared quite badly in this particular test.
There were likely also plenty of things about FTX and SBF that he “chose not to know”. These would be odd observations that he might have made that he never followed up on, for instance. He may or may not have contemplated the possibility that FTX was engaged in some serious wrongdoing; I’m honestly quite uncertain about this. Either way, I believe he didn’t have any proof and all the evidence he had access to would have been circumstantial.
I agree with almost everything you say here (still have my doubts about the frugality thing).
I didn’t mean to imply that all these things were innocuous. “Choosing not to look carefully” isn’t innocuous – but most of Germany did it in WW2. I think there’s an important distinction between “dark agency” (what Sam had) vs. a more passive “lack of high integrity,” which is how I’d describe the patterns you’re describing in your last comment here. I feel like “lack of high integrity” is a good phrasing. You may say it’s a euphemism, but I think it’s important to highlight that “high integrity” is difficult to cultivate and therefore rare. Also, I think people often develop high integrity after they fuck something up and are given a second chance. (I think this becomes less likely if they make mistakes later in life – but you never know and maybe some people can turn their mental habits at age 30+!)
In any case, I agree that both things (“dark agency” and “lack of high integrity”) are bad. I’d even claim that people tend to overrate the importance of whether bad things are done with fully-conscious bad intent/bad faith vs. (partly) unconscious hidden motives and self-deception. Still, I feel like it’s useful to be precise in what we’re accusing someone of, because the stuff that isn’t fully conscious seems quite common and isn’t always “irredeemable.” I would strongly vote against people who don’t have high integrity to have leadership positions, but I don’t think it warrants much more drastic steps like “I wouldn’t work with a person who is like that.” In fact, I could very much see myself from age 21-25 making mistakes very similar to Will’s if I had been in his situation, and I certainly don’t think my past self was a terrible person. (And I’m kicking myself for not having become more worried about FTX and about the implications of a potential collapse of FTX for EA after I pointed out to others how it seems very suspicious that Sam was defending the cryptocurrency tether. It’s long been a pet peeve of mine that tether seems like a potentially big fraud and people in crypto are totally insane to tolerate it!) I don’t know what your exact view is on these distinctions. I could image that you’re mostly upset about this particular issue because of the important position of responsibility that Will is in. And that makes perfect sense to me – we should hold leaders to especially high standards! So, maybe we don’t disagree too much. I think some of your phrasings in earlier comments, or maybe just the strongly condemnatory attitude, conveyed to me that you might be accusing will of “dark agency” and I thought that was unwarranted.
I’m pretty sure there are multiple sources confirming that Will knew about early Alameda events. One is that Semafor reported that the ex-Alameda employees reported allegations about SBF to CEA trustees in 2018. Will was a CEA trustee in 2018 according to Wayback Machine.
Notice that, as usual, these allegations are considered privately and nobody from CEA responds to requests for comments, even after FTX goes bankrupt.
People should be demanding from all EA organizations to make their internal deliberations on this subject public. What allegations did they consider in 2018? Why did they decide that they were ultimately not serious enough to warrant action?
The tolerance this community has shown for the silence of EA organizations in light of what has happened is astonishing to me. I think it’s partly that they simply don’t know the extent of the connection between early Alameda and central EA infrastructure, but that can’t be the whole story. There’s some slim chance that some of this will come out in court if the discovery process ever extends that far, but it seems much less likely at this point than I would like.
At least MacAskill and Beckstead need to resign from the EVF board, take a leave of absence, publicly recuse from responding to the FTX situation, or submit a detailed explanation of the facts and circumstances that render none of these actions appropriate. They are just too intertwined in the events that happened to be able to manage the conflict of interest and risks to impartiality (or at least the appearance of the same).
That’s not me saying that I think they committed misconduct, but I think the circumstances would easily “cause a reasonable person with knowledge of the relevant facts to question [their] impartiality in the matter,” and that’s enough. Cf. 5 CFR § 2635.502(a) (Standards of Conduct for Employees of the Executive Branch [of the U.S. Government]). Although I’m not going to submit that EA officials should always follow government ethics rules, this is a really clear-cut case.
Hopefully they have recused and this just isn’t being stated due to PR concerns (because some people might misinterpret recusal as an admission of wrongdoing rather than as respect for a basic principle of good governance).
Yes — since the first week of the crisis, Nick and Will have been recused from the relevant discussions / decisions on the boards of both EV entities to avoid any potential conflict of interest. Staff in both EV entities were informed about that decision in mid-November.
Thanks, Howie. That is reassuring.
My suspicion is that FTX fallout may be the defining issue of EVF’s corporate governance for the next few years, so having nearly half the board recused ain’t ideal but is certain better than non-recusal.
Why don’t you think the EA-Alameda connection will come out in court proceedings? I thought it was pretty likely.
I, at least, would be happy for those of examples.
What do you mean by this?
North Dimension is an example of what I mean. In general, the way FTX & Alameda Research structured their business affairs was extremely messy, and we’ve found out more about this since the Chapter 11 filing by FTX and affiliated entities. I think MacAskill was in a position to know about some of this and chose not to say anything. At the very least he would have known which entities are used to disburse funding to grantees, and he may also have known more about the informal way in which FTX conducted business dealings in general.
I also believe people over at CEA may have known this, because to my knowledge the Future Fund was partnering with CEA to process grants at some point. Integrating their infrastructure would have required CEA to have some exposure to exactly how FTX handles payments. The fact that nobody has thought it important to clarify this situation following the revelations that e.g.
FTX employees asked for their salary through Slack messages
at some point FTX customers had to wire money to Alameda’s bank accounts to deposit money at FTX
is quite strange to me.
By itself this is not a red flag; I think someone elsewhere in the comments called it a “yellow flag”. It has to be seen in the context of the other observations I mentioned to be worthwhile.
Lots of strong evidence-free claims in this comment.
Which particular claims do you think are “evidence-free”?
I hope you understand that some of what I know is from private conversations and I can’t disclose all of the details, not just because I want to remain private but I also don’t want to betray the confidence of those who trusted me with concrete information. I understand that makes my claims less reliable to outsiders who have no ability to verify that parts of what I’m saying are correct, but I view this as a desirable trade-off to make given the circumstances.
If you want further details, you should make EA as a community more receptive to proper criticism and stop punishing people who are critical. Perhaps then people will start to be more forthcoming with verifiable details.
Virtually all of it. You have one link in the entire long comment.
Same.
I don’t fully understand this. I think it’s okay not to obsess over on-going events and try to figure out what’s going on as quickly as possible. For people in a position to make public statements, you could argue that they should inform themselves carefully before making the statement. But appearances can be deceiving, so it’s though to form confident opinions even if you do a few hours of research. To become confident, you arguably had to make “What happened at FTX?” your main focus of the first couple of days after it happened (or maybe already have a lot of crypto-specific knowledge as your background and to inform your priors, which many EAs didn’t have).
I think the strongest case that something was badly off about some EAs’ early reactions is that some EAs probably/allegedly knew about some red flags at early Alameda, and depending on how serious and credible you think these reports are, it’s arguably unreasonable to say things like “there’s a good chance it’s all innocent” once it all came crashing down.
The people I’m talking about have been following the FTX saga from day one, so lack of familiarity with the specific events is not an excuse for them. However, I have noticed a pattern that the people who thought “maybe SBF was just incompetent” were always people with no domain expertise in finance. This is an insufficient excuse for their lack of sound judgment, however, because there were domain experts all over Twitter whose expertise was easily verifiable for outsiders and who were being transparent about their thoughts on this subject.
Among other issues I’ve discussed in the comments so far, I also view this as an instance of the general problem that EA doesn’t give sufficient weight to domain expertise.
I strongly agree with this comment because I personally created an account to comment on several posts that seemed wildly ignorant about the entire situation. It was incredible to me that the reaction was just “lets see how this plays out, they are going through a tough time” versus immediately reevaluating all of their priors about FTX and Sam. It was very clear that something was incredibly wrong and likely fraudulent once he tried to sell to Binance.
Why was trying to sell to Binance a sign of fraud?
It was a sign that FTX was in deep trouble, but at this time it was not yet clear whether there had actually been fraud or not.
That the problem was a solvency problem and not a liquidity problem was revealed, I think, by the following sources of evidence (among others):
Binance refusing to go through with the deal after only cursory DD
On-chain data showing only on the order of $1B of assets owned by FTX against estimated liabilities on the order of $10B
The leaked FTX balance sheet, which is what started the whole mess
The on-chain estimates turned out to be more or less accurate: FTX also held some illiquid venture investments but these were not sufficient to change the order of magnitude of the assets at their disposal.
Probably until around Nov 9 it remained plausible that FTX was in trouble as a result of some kind of asset mismatch between the two sides of its balance sheet, which would be more of a liquidity problem and something that a Binance bailout could have solved. However, around Nov 9 the evidence was just overwhelming that around $10B of customer assets were missing, and such a gap was impossible to explain by anything other than FTX having used customer deposits for their own purposes.
> The leaked FTX balance sheet, which is what started the whole mess
I think you mean the leaked Alameda balance sheet, on November 2nd?
There was pretty much no distinction between FTX and Alameda’s assets and liabilities by this point. It doesn’t matter what you want to call it, but it’s true Coindesk initially reported this document as “Alameda’s balance sheet”.
I find it very strange that you thought this insignificant detail is worth commenting on and ignored everything else in the comment.
That, and there was later a separate set of news articles around “FTX’s balance sheet”, on November 12th through 15th.
I don’t think it’s insignificant: the main interpretation I saw of the November 2nd news was that this was bad news about Alameda, not FTX. Which is pretty reasonable: at the time the true extent of their interconnection was not public.
If you’re reading through the forum and see a point where you have something to add, I think it’s generally good to respond even if there are other points where you don’t. It’s also not a strong signal that the addressed point is the most important one.
I very strongly agree with this.
That balance sheet (if it can be called that) is also not FTX’s balance sheet by your standards; it’s a balance sheet of all FTX-related entities (including Alameda).
This is true but I don’t see how it matters in the context we’re talking about here.
I agree. Binance walking out of the deal after cursory DD also gave plenty of information about how bad the situation at FTX was at the time.
Is there more context on the Guzey/Wise/Macaskill breach of confidentiality thing that’s worth being aware of? (Lots of the relevant posts/commentary have been deleted). Because on its face it seems bad, and I’m a bit surprised I hadn’t come across it.
I hadn’t realized the original apologies from Will and me are no longer visible since Alexey took down the post. Here’s what Will and I wrote in November 2018:
(Alexey and others then replied with more thoughts, but I haven’t sought permission to repost those comments.)
Since the initial posts are deleted, could you clarify whether these apologies came before or after MacAskill publicly posted a rebuttal he wrote based on the confidential draft?
My understanding is that this is after he posted the rebuttal and Alexey called him out on it, if I remember the situation correctly.
I was surprised by this since a decent chunk of the relevant discussion happened quite recently (maybe a year ago). I checked back and it seems like Guzey has deleted both the post in question and most of his relevant comments, both here and on LessWrong.
I’m sad about this, since it makes it quite hard to follow the dispute unless you were there at the time (e.g. I had some back-and-forth with Guzey at the time of the Decade Review, that I found quite informative, that is now unreadable). That said, the archived link includes the original post, the original discussion, and Guzey’s depiction of the confidentiality screwup, which is a decent chunk of the relevant context; the main thing that is missing is (the more detailed version of) CEA’s side of the story.
(FWIW, I think it’s reasonable to update negatively about the relevant parts of CEA to some degree as a result of the confidentiality screwup, but IMO the subsequent reactions to and discussion of the event reflected more badly on Guzey than on CEA.)
I don’t understand why you think subsequent reactions and discussion reflected worse on Guzey than on CEA. I think the incident seems pretty clear-cut.
I had the opportunity to discuss this incident with someone who knew Julia Wise personally and they also said that they don’t know what to make of it as they didn’t think Wise would do what Guzey was accusing her of. However, even if Julia Wise just made a honest mistake, MacAskill appears to have tried to conceal the fact that he had access to the draft Guzey sent to Wise in confidence.
For me this incident is telling because it shows MacAskill is willing to be dishonest even when the issue at stake is entirely trivial. The response to this blaming Guzey instead of MacAskill and Wise was very disturbing to me and I think it’s a good example of EA not taking criticism that bites seriously. Even if Guzey’s post was of low quality, and I think in some parts it indeed was, the events surrounding the publication of the post ought to dwarf any substantive concern you might have with Guzey’s criticism of Doing Good Better.
(Note that I’m not the author of the post, but someone else posting under a throwaway account.)
Obviously, I disagree with this. I think Guzey pulled together some fairly flimsy fragments of evidence of wrongdoing, and spun them into a story of deliberate, malicious conspiracy wildly out of proportion with the strength of the available evidence. His accusations here were entirely lacking in the kind of epistemic and rhetorical care I expect when making allegations like this, and his persistent refusal to moderate his claims or adduce additional evidence in response to pushback was a substantial negative update for me.
It’s probably easier at this point if I just quote one of my comments to Guzey from last year (now sadly hard to find due to him deleting his posts & comments):
I continue to basically believe this. I think the lack of a credible story by which Guzey was or could credibly have been harmed by CEA’s actions here is a big red flag, and the stories of harm that Guzey has provided have been very weak.
I also have a further claim, which is that what Guzey asked Wise to do in the first place is pretty strange and unreasonable. That doesn’t make it right to break his confidentiality—I think the correct response would have been to say “no, sorry, it doesn’t make sense for me to do this, I’d need to share it with Will”—but it makes the downstream mistakes much more understandable to me.
I really wish you’d chosen a more distinct username. I think not doing so is a disservice to the original post author.
I don’t think the evidence for wrongdoing is “flimsy”. It only looks that way to you because you’re starting with a very high prior confidence in Wise and MacAskill.
I think the likely story of wrongdoing looks something like the following: Julia Wise sends the confidential draft to MacAskill as the result of an oversight. She realizes afterwards that she wasn’t actually supposed to do this, and tries to cover up the fact that she did this. MacAskill feels personally attacked due to the hostile tone of Guzey’s confidential draft (which he has admitted to in public) and decides to exploit this opportunity to draft a preemptive response to kill Guzey’s post as quickly as possible. In the process, he overlooks that he is responding to something that Guzey has cut from the draft prior to publication, and therefore the entire incident is exposed.
The initial oversight of Wise is not the problem here. The problem is that neither she nor MacAskill felt it necessary to inform Guzey that his confidential draft had been leaked to exactly the person that he didn’t want to have access to the draft. Guzey has a habit of writing private drafts in a highly adversarial tone and toning this down after feedback from reviewers, so he would understandably be reluctant to share these private drafts with the targets that they are critical of. I don’t understand why you think this is “strange and unreasonable”.
It’s true that the issue at stake there was of trivial importance, but that makes MacAskill’s bizarre refusal to be transparent about reading a confidential draft even more disturbing. What would he do if he actually had a strong motive to deceive others?
As a final remark, I’ll tell you that there are “bad rumors” about MacAskill circulating in EA circles right now. People are reluctant to express exactly what they are concerned about and I can only guess at this as I have no inside information here, but I’ll leave you to draw your own conclusions about what that could mean.
I think it’s strange to ask for feedback on a post about a book from someone uninvolved in publishing the book, and unreasonable to ask them to give that feedback without consulting anyone who might actually be able to provide an informed response.
The most natural thing for Wise to do in this situation was to share the draft with MacAskill (or MacAskill’s research team for the book, if he had one). With that option excluded, sharing the draft with Wise was pretty pointless. I think it’s totally reasonable for Wise to have been confused here. (ETA: But it was still wrong to share it given Guzey’s request for confidentiality, as I’ve said several times.)
This is the part that seems extremely weak to me. I see zero evidence of anyone trying to “kill Guzey’s post”, and I think preparing a pre-emptive response is both totally reasonable and actively good for the discourse.
By “cover up” here, do you mean “not mention”? Because “cover up” to me strongly implies active destruction of evidence, silencing of witnesses, &/or other nefarious activity, and I haven’t seen evidence of this.
I hope that I never have to share a confidential draft with you if those are your standards of proper conduct when someone does so.
In that case, MacAskill would have had no reason to conceal the fact that he had read Guzey’s draft. He even placed a line in his response claiming that “Guzey’s post keeps changing” to hedge against the possibility that he might appear to be responding to something that’s not actually in the post.
I can’t understand how you can defend this blatantly dishonest behavior. It’s possible Julia Wise acted in good faith the entire time, but reasonably making this claim about MacAskill seems impossible to me.
I don’t understand what “evidence” you think they could have attempted to destroy. They actively tried to hide the fact from Guzey by not mentioning it and not admitting it when they were confronted on the subject by Guzey after he caught MacAskill responding to something that was cut from the post prior to publication.
What else would you be looking for to classify what happened as a “cover up”? I don’t understand.
It’s pretty difficult for me to respond to this kind of personal rudeness politely. I don’t think this is the kind of discourse that will make our community better.
I already said several times that Wise shouldn’t have shared the draft with MacAskill, that it was a serious error, that the breach of confidentiality is per se bad, et cetera et cetera. This is not in dispute, even by CEA.
I don’t care if you’re personally rude to me or not. I’m just a throwaway account and I assure you that my feelings won’t be hurt by you failing to be polite on an online forum. If my comment comes across as rude, keep in mind that I don’t write anything I don’t mean to say—I legitimately would not share a confidential draft with you given the views you’ve expressed in this thread so far, because I consider you to be insufficiently trustworthy.
Setting that aside, what I want to understand is how you think Guzey somehow comes across worse in this incident than MacAskill and Wise. I think it’s very hard to tell a story of this incident in which MacAskill was not intentionally deceptive on an issue with only trivial stakes. Julia Wise was at best incompetent and at worst complicit ex post after having leaked the draft by accident, as she didn’t reveal that the draft had been leaked to Guzey.
She looks even worse because this incident wasn’t added to the page of CEA’s mistakes until after Guzey discussed the incident in public 3 years after it took place. I can imagine Wise acting in good faith throughout this episode, but I can’t imagine the same for MacAskill, and it’s MacAskill that I really care about.
I already stated, IMO fairly clearly, why I think Guzey comes across poorly in this incident. I think there’s insufficient evidence of malice here, that your & Guzey’s stories of what a conspiracy to harm Guzey aimed to actually achieve are way too thin, and that the strength of Guzey’s reaction & his accusations are seriously out of proportion to both the strength of his evidence and the actual harm that occurred (in a way that I think is consistent with the many other online dust-ups that Guzey regularly gets into).
I could start going through your comments and responding line by line, but given how low my credence is that you’re actually truth-seeking here, it doesn’t seem like a good use of my time.
This is a typical way in which people in EA shut down discussions of subjects they don’t like, especially when they involve members of the outgroup. You simply accuse the person disagreeing with you of being insufficiently truth-seeking (whatever that means).
I’ve told you what I think most likely happened, and you’ve offered no counterargument other than “I see no evidence of anyone trying to kill Guzey’s post”, even though what happened is obviously evidence (in the Bayesian sense) for MacAskill attempting to do exactly that. It’s not like MacAskill could have prevented the publication of the post on the forum. What he had the power to do was influence the reception of the post, and I believe his response did influence the reception in a negative direction.
Guzey’s post indeed “died” and some of his substantive criticisms of the book were not addressed at all. I think most of this was due to natural dynamics on the forum that were beyond MacAskill’s control, but he certainly tried to accelerate this process and (I believe) succeeded in doing so. Whether that means he tried to kill the post or not is up to your judgment.
https://forum.effectivealtruism.org/posts/hYh6jKBsKXH8mWwtc/a-contact-person-for-the-ea-community?commentId=h6ra3qdAGr5W2zcwE, linked from https://www.centreforeffectivealtruism.org/our-mistakes
Thanks for sharing this! Sorry if I’m being dense or unnecessarily persnickety, but was Will aware or unaware the draft was intended to be confidential when he read it? I’m realizing that when I first read the post, I thought the answer must be “aware” (because there was a blinking “keep confidential” at the top of the email), but now see that which email was forwarded is unclear, so maybe my initial understanding was inaccurate.
Apologies if I’m just being paranoid in light of recent events (and I hate to reopen/relitigate an unpleasant past incident, recognize that no one’s perfect and mistakes happen, etc.), but I find this story (or at least my original understanding of the story) unsettling and the context seems important.
FWIW: I’ve known Julia for nearly 10 years. She is scrupulously honest and careful. I would bet
all the money I have (literally not figuratively)that she just made an honest mistake. I also don’t understand how this is relevant to the current complaint (FTX). This happened years ago and she’s already publically commented on it and apologized.Edited to add: I am seeing a pattern where commentaries lately pull out individual mistakes from any time frame and of any type to support a broader image of EA as committing malfeasance. I think this is epistemically shaky and beneath the standard of kind and collaborative criticism, I broadly hope for from members of this community.
Second edit: I’ve added a strikethrough after my husband pointed out some language in our Terms of Service.
Josh was asking about Will, not Julia.
While I respect the inside view information you would have about Julia Wise’s personality given your position, you should keep in mind that what you’re saying is exactly what William MacAskill would have told us about Sam Bankman-Fried two months ago.
I found her apology unpersuasive and the events surrounding the publication of Guzey’s post, including MacAskill’s behavior, were quite strange. I wouldn’t say I’m confident of anything but I wouldn’t be surprised if leaking Guzey’s private draft was planned on their part.
That she is honest is evidence that it was an honest mistake. That she is careful is evidence that it was not.
Against how much?
Edited to add: PM’d the person to offer to figure out a bet.
Yea—My guess is that KMF didn’t mean that part fully literally. Clarifying is good though.
Ooh, touche… Like 5k to AMF that she didn’t intentionally forward it knowing it was confidential and instead made an innocent mistake? Any more than that and I’m sleeping in the shed… ;)
Apparently my wife has been making huge bets on the internet with money from a hidden ‘fiat@’ account.
When you forget bae also reads the EA Forum...
I would personally strongly prefer a bet where the money ends up either in my pocket or in a charity of my choice.
Initially I was asking against how much you would bet al your money, i.e., against $1, $10, $50, etc, i.e., how much I would have to bet on my side. But I think you answered the other side, i.e., how much you would bet on your side?
So I agree that in particular seems very likely (though I’d still bet, e.g. $1 against your $5k). I’m more interested in betting that at some point between the initial mistake and the post being published she had some “deceitful intent” (chosen so as to mean the opposite of an “honest mistake”).
To resolve this, we could either ask Julia, or we could ask three persons we both trust about which one is more likely.
Going to sleep now, will follow up tomorrow
Being willing to bet the money you have is much less strong evidence than the odds you would bet at. I too would be willing to bet very large amounts of money … if offered sufficiently high odds. The key issue is what odds you would make the bet at. Are you 95% confident? 99% confident? 99.99% confident?
(This is of course rather moot if you are not actually willing to bet.)
edited to add: note that the Kelly betting formula does not apply in a simple way here because for most young people ‘all the money they have’ is a small fraction of the net present value of their future income.
So it’s relevant to me because I only put two-and-two together about this after the FTX implosion. As I said in my initial post:
My point isn’t that this is related to FTX. My point is prior to the collapse I didn’t know some things about the EA community. Now I know them and I think less of the community. My general point is it seems absolutely nuts to me that Duncan can have an unchanged opinion of the community unless he already knew a bunch of things I was unaware of.
Following up on my bet offer below, seems good to either figure out the odds, or to edit the “would bet all the money I have (literally not figuratively)” claim. No rush on my end, though, I’ll follow up in a few days.
Can you say more about “[p]eople who received funding in dubious ways and didn’t say anything (North Dimension donating on behalf of FTX Future fund)”?
“FTX Future Fund” wasn’t a legal entity (although there was “FTX Philantrophy” a/k/a “FTX Foundation”). I think it was disclosed on the “Fund’s” website that that the monies might actually come from various entities and people. I didn’t save a copy of the website before it went down, though. So, it’s reasonable that a recipient would think any potentially dubious information had been publicly disclosed already.
Second, it’s not immediately obvious to me why the exact funding source is “dubious.” At least for a natural person, there are potentially solid (tax) reasons certain persons might choose to make some donations out of the entity that actually realized the associated income rather than out of a private foundation. Without the benefit of hindsight, I don’t think I (who am a lawyer) would have caught this as a flag that needed to be discussed with the community—and I wouldn’t expect the grantees (who are generally not) to have flagged it either.
So I wouldn’t lower my opinion of grantees, or EA in general, based on the existence of people who did not mention their exact source of funding.
As a comparator, over the years I’m pretty sure I’ve received funding “from/through OpenPhil” that actually legally came from a few different entities.
Did any of that funding come from defunct phone stores, which is what North Dimension legally was?
Big corporations have lots of subsidiaries, and I don’t think it’s super-rare to acquire what amounts to a dummy/shell corporation for non-fraudulent reasons. I’d call it a yellow flag if the grantee knew North Dimension had been “defunct phone stores.” Except for perhaps recipients of the largest grants, I don’t think that flag is bad enough that the grantee couldn’t have reasonably relied on others’ due diligence—both in the investing world and in EA.
But more to the point: the only thing the grantee knew was that the money came from an entity with the generic name of North Dimension. I don’t think that information alone put them on notice that they needed to dig or discuss with others.
To be clear, it seems that far too little due diligence was done as a whole, but I don’t think the original poster should think less of most individual grantees for that.
Do you have a link for that claim?
https://twitter.com/bax1337/status/1590916763549503488
Sure, I don’t know anything about North Dimension and am open to the claim that it in particular was dubious.
I don’t think I’d have spotted this which in hindsight seems like a dumb error. I guess there are a range of red flags like this?
I don’t think you would have been obliged to spot it, but someone should have. Obviously the people who were in a prime position to spot this were the people over at the Future Fund, and if I recall correctly Nick Beckstead posted a comment on the forum saying that Future Fund grants would be disbursed through a variety of legal entities.
I haven’t seen confirmation of this, but I would guess they knew the list of entities they used most frequently for this purpose, and North Dimension seems to have been one of them. Explaining why a defunct phone store was being used to disburse payments to grantees seems like it should have been a reasonably high priority.
I appreciate the conciseness of this post. It’s an under-appreciated skill on the EA Forum.
The reorganization was announced on 2022-09-13 (post) so it predates the FTX collapse by about two months. But it wasn’t very well publicized and some public information was slow to be updated.
Previously there was:
CEA (the umbrella org)
CEA (the community building project)
80k
GWWC
etc
And as of September it’s now:
EVF (the umbrella org)
CEA (the community building project)
80k
GWWC
etc
Did anyone else find the title of that post (“CEA Ops is now EV Ops”) a bit confusing?
When I first saw it, I thought, “oh, it seems like a team/sub-org within the CEA changed its name. Cool” and didn’t think much else about it. It didn’t strike me as important enough for me to click the title and read the post.
It was only months later, by reading an EA Forum comment in another post, that I noticed that CEA UK (the umbrella org) had changed its name to Effective Ventures. When I read the comment, I spent some time trying to find an official announcement, and it took me a while to realize that it was that post.
I’m confused about why that title was chosen over titles that made it clear that it was the CEA umbrella org that changed its name, rather than only a specific team within it.
Wrote up a disambiguation post
This is even worse than I thought, this is right in the window of “rats fleeing the boat”
Sam Trabucco left Alameda on 24aug (https://twitter.com/AlamedaTrabucco/status/1562519114979356673)
Bret left FTX.us on 27sep (https://twitter.com/BrettHarrison88/status/1574782901748727809)
If I were a cynic, I’d have thought they were trying to demonstrate they were separable entities for legal reasons when they knew the risk was hitting the fan.
(I don’t believe this, but I’m 20% this is true)
They weren’t and aren’t separate legal entities. The change clarified the one actual org’s internal structure.
Thanks for posting this. I think it’s useful for calibration and I appreciated the measured tone & takeaways. I don’t agree with everything you included in your list of negative updates (and have gotten into some conflict in the comments here about a couple of them), but your overall update...
...seems pretty reasonable to me.
throwaway790 what’s your reason for approving of Rob Wiblin’s statement about FTX, but not Will Macaskill’s, or Holden Karnovsky’s? I read all their statements as ‘we strongly condemn this’ in summary.
So this is more from memory at the time although I do have messages I was sending to other EAs so I know I felt like this at the time. For example, my first reaction to Rob’s statement for example was to forward it to Nathan Young saying “That’s what Holden should have said”
My sense of Holden and Will’s statements were they were very equivocal about whether or not anything bad had happened. “If something bad has happened then we condemn it, we must reserve judgement” etc. eg Holden’s statement:
Holden worked in finance before going into direct work. There is no way I can believe that at the time he didn’t know better than this. (And I think he did know better and toned down his language because Alexander suggested he should).
Strong disagree.
I would have qualified my statements even if I was 99% confident that SBF had committed outright fraud. I would expect more and more information to come to light which could nuance my views (if not outright change them) which would lead me to want to change my public position.
Moreover, I felt Will’s statement was quite strong in outright denouncing SBF, and had this qualifier as an afterthought; not in the foreground.
I didn’t take issue with Holden’s post. For me the subtext was “I’m going to be giving cold takes, because another loud voice isn’t needed right now” at a time when Twitter, EA Forum and media overall was alight and not with the most sensible takes. I respect that as his choice, and I don’t feel he minimised what was going on. I also think it was sensible for him to focus energies on ‘here’s what this means for OpenPhil, and for the EA funding ecosystem, which is what most of you care about’.
Personally, I felt most aligned with Rob’s take, and I probably would have posted something similar. To note, he also included a qualifier: “If media reports of what happened are at all accurate, what at least two people high up at FTX and Alameda have done here is inexcusable (e.g. wsj.com/articles/ftx-t…).”
I just don’t understand how you can say that. Almost every tweet had the qualification in the foreground. Certaintly it wasn’t an afterthought. His offer to apologize was in the 4th tweet, before he even condemns Sam. (Edit: link to thread)
I read “more likely than not” to be ~60% confidence, which doesn’t show a very high level of understanding of the situation (which is not very acceptable for someone who should have been monitoring this situation very closely). It also colors the rest of what I read as someone who doesn’t think that there is much certainty to be had here.
Emphasis mine, but I’m afraid we’re either talking about different statements, or read these tweets very differently.
ok, thanks for the explanation. I’m also working from my memory of that time. And my impression at the time was that things were moving very quickly—FTX had gone in a few days from one of the biggest crypto trading firms, to insolvent—and I was very unclear what was happening, and most of the information seemed to be coming from twitter. So I was more approving, not less, when Will and Holden used qualifiers like “I’m not 100% certain what has happened”. But it’s possible in hindsight I should expect them to have had better access to information than me.
Yeah, I can understand how some people might have been uncertain about what happened, but given what was public at the time, anyone following closely would not have been as uncertain as you at that time.
This is the first time I’m hearing about this. Am I right in understanding that EA has got involved with not one, but two crypto billionaires who are on the wrong side of the law?
Wikipedia link in the original quote is broken btw.
IIRC compared to Muskovitz and SBF, Delo was/is a fairly small-scale billionaire, but yes. He was a big shot for ~2 EAGs around 2019, & took the Giving Pledge at a very young age. There was a lot of excitement about him because he was the first ~major funder to pledge to make major donations entirely to longtermist causes.
In 2020 he and his co-founders were indicted for failing to establish adequate anti-money-laundering procedures on their online platform, and he dropped off the map.
My cached memory of this incident is that Delo’s wrongdoing was far more minor/procedural/unimpressive than what SBF is accused of, and more an example of “regulators & crypto people wrangling over how crypto should be regulated” than clearly immoral behaviour. But other conversations on the Forum (which I can’t find right now, but which involved people who seemed much more knowledgeable than me) have claimed that he was also accused of significantly more shady stuff, so IDK.
(ETA: Whether or not the preceding paragraph is true, it is nevertheless the case that “EA has got involved with not one, but two crypto billionaires who are on the wrong side of the law”.)
Thanks for the explanation. According to that wiki article, Ben Delo was charged with violating the Bank Secrecy Act on 1 October 2020.
I think EA leadership should have learned from this and been much more cautious about the financial, legal and reputational risks of having large amounts of their funding provided by crypto billionaires.
I’m now much less sympathetic to the view that “The FTX collapse and it’s effect on EA was a freak unforeseeable event”, and more sympathetic to the view that “EA has a major blind spot in this area, and some aspects of EA culture or ideology may encourage or enable the kind of fraud that apparently happened at FTX.”
Yeah, but I feel like the situation with Ben Delo wasn’t perceived as a highest-category catastrophe at the time because Delo never became a huge public figure within EA and wasn’t particularly involved with any big project. (He worked with some EA funders briefly and it probably didn’t reflect well on them, but they seemed like a smaller and newer org at the time and it was overall an extremely different situation than what happened with FTX and the Future fund.) This could explain why people didn’t draw a big enough update.
Also, while I think “accepting money from someone and it later turns out it was based on fraud” is bad, I view it as the type of bad that I don’t think one has to spend sleepless nights over, at least not under all circumstances. The circumstances where it’s most unambiguously bad are when a donor is using their donations to build up a positive public image and they become more active and powerful within the EA ecosystem. In those instances, it seems absolutely crucial to make sure they’re not fraudsters.
However, if a donor is keeping a low profile and the amount of money doesn’t make you too dependent on them, then IMO there’s a strong argument to be made for “minimal vetting is perfectly okay.” By “minimal vetting,” I mean something like you check if there’s an ongoing criminal investigation into the person in question and you comply with any laws regarding money laundering and so on. And if there’s no criminal investigation and if the donation is compliant with money laundering laws, then that’s it and you take the money without worrying about it too much. It’s not your job to figure out if someone’s a financial criminal – there are government agencies whose job that is. If it later turns out that there was fraud, you stop taking the money! (And maybe have to give some of it back for moral and legal reasons – so prepare for the possibility in your planning.) In practice, I think there’s a problem where, by taking the money, you automatically “endorse” the donor in some sense, and this could be used by them to white-wash their reputation. So maybe you could write a disclaimer that describes the type of vetting that your org does for big donors, so it’s apparent to outsiders that accepting a donation doesn’t necessarily mean “we vouch for this donor being an A+ high-integrity person.” But that should be pretty self-explanatory, since it would be an unreasonable standard to vet all donors in that way. I guess it’s a special situation with crypto given how crazily high the fraud rates are. There are many honest people in crypto, but the risk is maybe elevated enough to warrant the awkwardness of a “how we vet crypto donors” disclaimer.
I’m aware that some people think there’s a moral obligation to go beyond the process I described above, but IMO that’s a legitimate thing to disagree over and we also have to consider the costs of turning down money or of asking big donors to go through time-intensive auditing for extra vetting. It’s easy to virtue-signal that we should never do anything that could turn out to look bad, but that’s not the right stance to take either given that the world is full of suffering and inaction is often costly, too.
Thanks—I will fix the link.
This is kind of my point about why I think less of EA (despite still being a big fan) now than I did before FTX collapse. It’s the same with the Guzey thing I mentioned earlier