closes Jan 1, 2024
By the end of 2023, will substantial evidence emerge that leaders in EA knew about (or had strong suspicions of) SBF/FTX's unethical investing practices and chose to cover it up?
21%
chance

Resolves very subjectively, I'm not even going to bother to try to define this exhaustivly. Feel free to ask about hypothetical scenarios and I'll answer how I'd judge them. I'll add clarifications to this description as they're worked out.

  1. "A leader in EA" refers to anyone who has significant influence/social status within the EA community or has the power to move large quantities of EA-allocated funds. Heads of EA organizations, writers of EA-aligned books, grantmakers, etc. I'll be fairly liberal with this definition; I'll lean towards more people being "leaders" rather than fewer. (It does not include SBF or anyone else who could be considered FTX/Alameda leadership.)

  2. The "unethical behavior" in question must have been related to fraudulent investiment strategies that involve spending other people's money without their permission. If SBF engaged in unrelated unethical behavior, such as walking past a child drowning in a pond or illegally harvesting organs, that doesn't count for this market.

  3. "Cover up" refers to knowing about the unethical behavior and chosing to keep that knowledge private. There doesn't have to have been any particular action taken; inaction is morally equivalent to action, after all. Some clarifications:

    1. The EA leader in question must have assigned a decent probability to fraud/theft occurring. It it seems that they had enough evidence to conclude there was fraud happening, but naivety/motiviated reasoning prevented them from realizing it, that does not count as a "cover up". (I won't be too generous with this; if it seems like any reasonable person would have known, yet they still profess innocence, I'll count that as a cover-up.)

    2. If they were afraid of retaliation and didn't publicize their concerns out of fear for themselves, that still counts as a cover-up. Any charitable organization where whisleblowers are discouraged that strongly has a serious problem.

    3. If they only found out about the fraud after signing an NDA and couldn't legally say anything about it, that does not count as a cover-up on their part. (It would certainly be a cover-up from FTX, but we already know that happened.) After all, the whole point of this scandal is that we don't want EAs breaking laws just to do what they personally think is right.

    4. If actions were taken to maintain plausible deniability, or it otherwise seems like they were trying to "avoid finding out" about the fraud, that still counts as a cover up. Bayes cannot be fooled by such games.

    5. Market resolves YES if at least one EA leader satisfies these criteria. Multiple "leaders" is not a necessity.

I will not be placing any trades in this market. I will discuss my intended resolution with the community here before actually resolving it, to ensure people feel my resolution is reasonable.

If anyone wants to submit information completely anonymously, you can do so here.

Sort by:
KatjaGrace avatar
Katja Graceis predicting NO at 16%

@IsaacKing I've been reading this as 'at least one leader' but realized the title is plural. By 'leaders' do you mean at least one? Or two? (A bunch?) Sorry if you already said this somewhere.

IsaacKing avatar
Isaac King

@KatjaGrace You actually asked it yourself a few months ago. (I had also forgotten.) My mistake for not adding it to the description at the time; I've now added it.

KatjaGrace avatar
Katja Graceis predicting NO at 21%

@IsaacKing lol glad past me has my back here, if we are all very forgetful. Thanks!

Mason avatar
GPT-PBotbought Ṁ10 of YES

EA leaders knew of SBF's tricks,
But chose to keep their lips sealed with bricks.
By the end of 2023, will we find out?
Or will the truth continue to be in doubt?

IsaacKing avatar
Isaac King
Gigacasting avatar
Gigacastingsold Ṁ48 of YES

I only read anon twitter not unreliable sources like the corpse of “time magazine”

Gigacasting avatar
Gigacastingis predicting YES at 19%
IsaacKing avatar
Isaac King

@Gigacasting Is this new reporting, or just a rehash of the TIME article?

GeorgeVii avatar
GeorgeViisold Ṁ99 of YES

Arbing against this market fwiw. I don't really understand why it should be much so much lower (half). Maybe missing something?

IsaacKing avatar
Isaac King

@GeorgeVii The FTX Future Fund team is much smaller than all of EA leadership.

MaggieDelano avatar
Maggie Delano

Exclusive: Effective Altruist Leaders Were Repeatedly Warned About Sam Bankman-Fried Years Before FTX Collapsed

https://time.com/6262810/sam-bankman-fried-effective-altruism-alameda-ftx/?utm_source=twitter&utm_medium=social&utm_campaign=editorial&utm_term=_&linkId=205508922

DavidMathers avatar
David Mathers

Weird to me that this is higher than the charity commission market: https://manifold.markets/Jason/will-the-charity-commissions-statut-78d5dd70a396

Surely if this resolves 'yes' then likely the charity commission will find out and criticize them heavily, but also even if this resolves 'no', they might get serious criticism for the charity commission for something else, like having generally dreadful procedures for dealing with conflict of interest or something.

berealistic avatar
berealistic

I agree and I doubt the charity commission would be lenient about misconduct/mismanagement because of 'motivated reasoning'.

Also, it's surely much more important for EAs to think about these CC markets? I doubt this market sitting at 10% or 90% would affect people's decisions very much, whereas these CC actions would have a big impact on EA as a whole and would be worth preparing for.

Jason avatar
Jason

@berealistic As the creator of those markets, I'd say that Issac's market is both broader and narrower. His definition of leader is pretty broad, while my markets implicitly cover only people associated with EVF. On the other hand, his other terms are narrower than what would trigger my markets, both for the reason @DavidMathers mentioned but also because the CC would likely make a finding or take significant regulatory action on a lower mental state than a leader having "strong suspicions" of a particular type of fraud.

IsaacKing avatar
Isaac King

Oh I figure I should mention this link in case anyone wants to share something anonymously.

https://www.admonymous.co/isaacking

GeorgeVii avatar
GeorgeViibought Ṁ100 of YES

1k limit at 17% fyi

GeorgeVii avatar
GeorgeViibought Ṁ100 of YES

@GeorgeVii also at 20

GeorgeVii avatar
GeorgeViiis predicting YES at 23%

Wow, Bouscal - the insider tradoor

IsaacKing avatar
Isaac King

@GeorgeVii What's this last comment in reference to? I don't recall hearing about Bouscal before.

GeorgeVii avatar
GeorgeViiis predicting YES at 29%

@IsaacKing Ah, she who is quoted in recent time article - presumably knew it was coming out and bought a ton of yes on the 8th (3rd largest yes holder)

IsaacKing avatar
Isaac King

@GeorgeVii Oh, I didn't see that they had bought YES here. Neat! I approve. :)

RobinGreen avatar
Robin Greenis predicting NO at 21%

@IsaacKing I don't. This updates me towards considering her less credible, since she had a motivation to play up how bad it was, even if it's only a play-money market. In general I don't think whistleblowers should bet on prediction markets about what they are whistleblowing about, for this reason.

NathanpmYoung avatar
Nathan Youngbought Ṁ100 of NO

@JavierPrieto Do you think anything in this resolves it Yes?

I think if this is all we find out I expect it to resolve no

NicoDelon avatar
Nico Dsold Ṁ86 of YES

@NathanpmYoung even if it's not "substantial evidence" yet, it's quite a bad look. This is frankly maddening.

NathanpmYoung avatar
Nathan Youngis predicting NO at 19%

@NicoDelon I think it's a bad look in terms of strategy - we shouldn't have tied our name to SBF, but I think less a bad look in general, except maybe wills protection.

But also I'm asking a specific question about this market.

NicoDelon avatar
Nico Dbought Ṁ20 of YES

@NathanpmYoung Given the high standard for this market, I agree that this is insufficient. But in real life, where standards are different, it's embarrassing—lacking judgment, looking away, dismissing testimony...

DavidMathers avatar
David Mathersis predicting YES at 19%

@NathanpmYoung Would 'didn't distinguish between firm capital and trading capital' maybe count as 'spending other people's money without their permission'? I don't understand finance enough to have a firm sense of what that means, but it was the line that caused me to buy 'yes'. (At 15%, I'm not necessarily over 50% on this resolving yes.)

finnhambly avatar
Finn Hambly

@DavidMathers That lack of distinction wouldn't necessarily imply they're spending other people's money without permission (since they'd fundraised a lot of cash). You should probably make sure that the cash you need for payroll isn't being used for trading, though.

DavidMathers avatar
David Mathersis predicting YES at 21%

@finnhambly 'You should probably make sure that the cash you need for payroll isn't being used for trading, though.' Sounds like a fair chance of a yes from this to me.

finnhambly avatar
Finn Hamblybought Ṁ300 of NO

I agree with @NathanpmYoung that I expect this will resolve no if this is all we find out. This excerpt is in keeping with what we already knew and discussed here:

None of the early Alameda employees who witnessed Bankman-Fried’s behavior years earlier say they anticipated this level of alleged criminal fraud. There was no “smoking gun,” as one put it, that revealed specific examples of lawbreaking. Even if they knew Bankman-Fried was dishonest and unethical, they say, none of them could have foreseen a fraud of this scope.

and Isaac made this clarification last month:

My intention is that the EA leader must have had a strong suspicion of the actual large financial crime that occurred. Other crimes are useful evidence towards that; if I see someone break a bunch of regulation laws, maybe defraud someone in an unrelated business, it would certainly increase my credence that they're defrauding people in their main business. Does that answer your questions, or am I misunderstanding?

Obviously EA leaders ≠ early Alameda employees, and EA leaders who were involved more recently may have known more. I don't think this article is a big update though (unless you didn't realise that EA leaders ignored those earlier warnings).

DavidMathers avatar
David Mathersis predicting YES at 21%

@finnhambly Ah, that clarification makes it sound like this would only resolve positively if they suspect that FTX was dipping into customer deposits to trade. If that's the case, I think no is still 92-3% or so. (Not because I think 'Will and Nick and other top EAs are lovely people who wouldn't do that, but because I can't imagine why anyone with knowledge of the actual ongoing fraud at FTX would have informed them, but not blown the whistle more generally.)

finnhambly avatar
Finn Hamblyis predicting NO at 19%

@DavidMathers Agreed, and also I don't think observing a lack of distinction between funds means you're observing other people's money being spent. If someone asks me to look after ÂŁ50 for them, and I put it with my other money without ringfencing it, I don't think I'm necessarily spending it (until my balance is under ÂŁ50).

NicoDelon avatar
Nico Dbought Ṁ10 of YES

@finnhambly It's not a huge update but it does provide more background and ties together a few disjointed pieces. IMO, if that's all we know, there's likely to be a lot more that we don't know which, if it emerged, would make it resolve YES. It all hinges on whether it emerges by the end of 2013, and it may never. But at this point, the idea that no EA leader had reasonable suspicions of fraud (and as a matter of fact didn't blow the whistle) looks to me like comfortable self-deception.

GeorgeVii avatar
GeorgeViiis predicting YES at 23%

@finnhambly "Obviously EA leaders ≠ early Alameda employees" - Idk Tara being the ex-CEO of CEA puts here in the 'EA leader' category to me, no?
Tho I suppose you could argue she did "whistle-blow" but was simply shutdown/ignored by the other "leaders" she whistle-blowed to for whatever reasons (seemingly internal drama/politics of CEA?)

IsaacKing avatar
Isaac King

@NicoDelon

there's likely to be a lot more that we don't know which, if it emerged, would make it resolve YES. It all hinges on whether it emerges by the end of 2013, and it may never.

Note that we can still take into account evidence we expect to exist that has not been made public.

IsaacKing avatar
Isaac King

This article seems to mostly be a rehash of what we had already pieced together months ago. However, their portrayal of the planning document does make it sound a little worse than we thought:

The document, which has not been previously reported, accuses Bankman-Fried of dismissing calls for stronger accounting and inflating the expected value of adding new exchanges, and said a majority of employees thought he was “negligent” and “unethical.” It also alleges he was “misreporting numbers” and “failing to update investors on poor performance.” The team “didn’t trust Sam to be in investor meetings alone,” colleagues wrote. “Sam will lie, and distort the truth for his own gain,” the document says.

Misrepresenting numbers and lying to investors sounds a lot like fraud.

Would be really nice if we could read the document itself rather than rely on TIME's summary.

finnhambly avatar
Finn Hamblyis predicting NO at 28%

@IsaacKing Would you be able to say how you'll decide whether "they had enough evidence to conclude there was fraud happening"?

It seems pretty unlikely that any EA leader is going to come out and say 'I put a pretty decent chance of fraud' more strongly than was given in this comment, so it's probably going to come down to your judgement on whether the evidence available to others was enough.

The Slack messages were deemed insufficient evidence on their own, and I'd expect that misrepresenting numbers and lying to investors would also not be enough for someone to have "a strong suspicion of the actual large financial crime that occurred" — but maybe you think differently?

(and please forgive me if quoting your words back to you looks antagonistic, I'm just attempting to keep track of what the criteria are without misrepresenting the previous discourse!)

finnhambly avatar
Finn Hamblyis predicting NO at 25%

@GeorgeVii I meant that EA leaders aren't precluded from seeing smoking guns of specific lawbreaking just because early Alameda employees didn't (as the article claims) — was just adding the caveat that they could have ended up seeing more

NathanpmYoung avatar
Nathan Youngbought Ṁ1,000 of NO

fwiw, I think it will hinge on this "The "unethical behavior" in question must have been related to fraudulent investiment strategies that involve spending other people's money without their permission."

I think people will turn out to have known that Sam was did shortcuts, but not that he did fraud or used other people's money. Hence I expect it to resolve negative.

NicoDelon avatar
Nico Dis predicting YES at 16%

@NathanpmYoung That seems optimistic. I hope you're right.

DavidMathers avatar
David Mathers

@NathanpmYoung What's a shortcut in this context? I think I agree with your general guess that people will have had good evidence Sam was shady, but this will probably resolve negatively because they didn't know about the actual fraud, and so it's unlikely (though certainly possible) that they had strong, evidence-backed suspicions that meet the 'involve spending other people's money without their permission' bit.

NathanpmYoung avatar
Nathan Youngis predicting NO at 16%

@DavidMathers I can imagine people knew he was sloppy in his accounting. That he didn't use normal processes if he didn't think he ought to.

johnleoks avatar
johnleoks
Comment hidden
KatjaGrace avatar
Katja Graceis predicting NO at 19%

@IsaacKing

""Cover up" refers to knowing about the unethical behavior and chosing to keep that knowledge private."

Sounds like it can't be fulfilled by if the knowledge is already not private—is that right? e.g. If a rumor reaches Bob that Alice might be doing something illegal, and Bob is told that this rumor is generally circulating, then can Bob be 'covering it up' if he guesses that Alice is committing fraud and doesn't announce that guess publicly?

IsaacKing avatar
Isaac King

@KatjaGrace I think it depends on how widespread the rumor is. If it's a rumor that only spreads among EA leadership and their friends/co-workers, then I think that still counts. It's it's a rumor that thousands of people have heard, gets mentioned on the EA forums, etc., then I don't think that should count.

KatjaGrace avatar
Katja Graceis predicting NO at 19%

@IsaacKing Thanks! In the Slack story, it sounds like it's 'some circles' plus some part of the government/law enforcement. Intuitively that sounds more like the former case (something like thousands of people have heard, and they are not EAs), so if one of the Slack users came to have a strong suspicion from this, I would not be inclined to say they covered it up. Would you?

IsaacKing avatar
Isaac King

@KatjaGrace Interesting, I have the opposite impression of the same information. I feel like if it was restricted to people in the government investigating the crime and people in a private Slack channel, that doesn't seem like public knowledge. There's no plausible way I could have found out about it, for example, even if I had gone looking.

KatjaGrace avatar
Katja Graceis predicting NO at 19%

@IsaacKing I read it as additionally spread among 'some circles', which sounded to me like probably many people, especially as the circles' view is described as 'a reputation'. But I don't know if it sounds like 'public' information, just not 'private'.

(Separately, it seems intuitively especially not like a 'cover up' if the people you received the information from were law enforcement of some kind, since they seem like the primary people you are meant to report crimes to.)

Fwiw, even if info were private, it would often seem strange to call failure to share it 'covering up' to me. e.g. if in general I hear a rumor about an acquaintance, am I morally required to post the rumor on Twitter? I don't think normal social norms say I should, even if I suspect that it is true. And I don't think most people would say that I 'covered it up' if I didn't actively spread it. (Seems different if someone asked me about the person and I misled them for instance.)

IsaacKing avatar
Isaac King

@KatjaGrace I agree my definition of "cover up" may be noncentral, but I think it's a little more reasonable than you make it out to be. Its usage in the resolution criteria isn't referring just to the spreading of unfounded rumors, but to anyone who has a serious belief that there's an issue and chooses not to say anything about it.

KatjaGrace avatar
Katja Gracebought Ṁ1,000 of NO

@IsaacKing I thought the situation here that might trigger resolution yes was that someone in the Slack discussion got a serious belief in high P(fraud) as a result of hearing the rumor

IsaacKing avatar
Isaac King

Why is this so low? While none of the public evidence seems conclusive, the fact that we have people saying things in an EA leadership Slack channel like:

Just FYSA, said to me yesterday in DC by somebody in gov’t: ‘Hey I was investigating someone for [x type of crime] and realized they’re on the board of CEA or run EA or something? Crazy! I didn’t realize you could be an EA and also commit a lot of crime.

[...]

I think in some circles SBF has a reputation as someone who regularly breaks laws to make money

[...]

I guess my point in sharing this is to raise awareness that a) in some circles SBF’s reputation is very bad b) in some circles SBF’s reputation is closely tied to EA, and c) there’s some chance SBF’s reputation gets much, much worse.

certainly seems to imply that somebody could have had reasonable suspicion of serious fraud.

What we should be considering is not the evidence that has been made public, but all evidence that exists. The fact that nothing completely damning has come to light implies that it doesn't exist, since if it did exist, it would be more likely to be shared by someone than the wishy washy borderline evidence we're debating here. However, the existence of this amount of borderline evidence tells me there's probably a lot more where that came from, and the only reason we don't know about it was that it was in more private locations. If we look at the trend line and update to the end, it certainly looks to me like there's a greater than 16% chance somebody had enough data to make the connections and have strong suspicions that a significant fraction of FTX/Alameda was built on fraud.

Jason avatar
Jasonis predicting NO at 23%

@IsaacKing "fraudulent investiment strategies that involve spending other people's money without their permission" is pretty narrow. Reports are that as few as 4 people at FTX/Alameda actually knew about the mechanism. Would probably be higher for "fraud" or "illegal conduct that is malum in se"

IsaacKing avatar
Isaac King

Alright, I'm going back through my markets and making sure I addressed everything that needed addressing. (Apologies for how long it took.) Are there still any questions or disagreements about this market's resolution criteria, or are we just waiting to see what comes out in the trial?

finnhambly avatar
Finn Hamblyis predicting YES at 19%

@IsaacKing There are a bunch of outstanding questions about the resolution criteria in the comments IIRC, and I don't think they're easily summarisable unfortunately

DeanValentine avatar
Dean Valentineis predicting NO at 23%

When I bet NO on this market a while ago, I did so based on the understanding that it would have been crazy and unusual for SBF or his coconspirators to go leak info about their Ponzi scheme to the charities they donate to.

That seems basically to have been a sound conclusion. Caroline is testifying in court now that, quite reasonably from their perspective, only four people knew about the "loans" to Alameda, even within FTX - Her, Gary Wang, Nishad Singh, and Sam. The NYPost article showed that EA leadership may have heard through the grapevine that the DoJ was investigating Sam for something, but it sounds unlikely so far that they were investigating them for "fraudulent investiment strategies that involve spending other people's money without their permission" or that EA leadership concluded that.

I am eager to find out whether or not Sam takes a guilty plea. If he does and he confirms that the knowledge about Alameda's gambling house funds was limited to those four coconspirators, I see no reason not to take him at his word, here, and would think this should be at <5%. I would vote it down to <5% now, frankly, if I weren't worried that Isaac King somehow disagreed with me about resolution criteria.

NeelNanda avatar
Neel Nandais predicting NO at 22%

"If he does and he confirms that the knowledge about Alameda's gambling house funds was limited to those four coconspirators, I see no reason not to take him at his word, here, and would think this should be at <5%"

Why do you believe this? Seems to me that if SBF has friends who also knew, he would prefer them to not be thrown under the bus, all things being equal.

firstuserhere avatar
firstuserhereis predicting YES at 22%

@NeelNanda prisoner's dilemma iterated basically

BetsByAnon avatar
Bets By Anonis predicting YES at 25%

@DeanValentine I've recently seen how hard it was for Coffeezilla to get SBF to admit commingling of funds. I think his overall strategy might have been conversational aikido and sleight of hand. Counting on conversational partners not having enough working memory to keep track, but not not leaking

Caroline on the other hand didn't seem to have employed obfuscation to the same degree. Surface impression is less message discipline, anyway.

So I think there's a high chance someone might have figured it out.

DeanValentine avatar
Dean Valentineis predicting NO at 25%

@NeelNanda Because he would be in violation of his plea agreement for lying. As would Caroline.

DeanValentine avatar
Dean Valentineis predicting NO at 25%

@BetsByAnon "Keep track" of what? Just don't tell people you're committing massive fraud? It's harder when you've already been caught doing it, but I don't understand what hard questions you expect he would have had to answer in September of this year.

BetsByAnon avatar
Bets By Anonis predicting YES at 23%

@DeanValentine Keep track of moving parts in the explanation. For example, when pressed on the FTX TOS, he always started talking about margin trading.

Questions like "did you, or did you not knowingly lend out user funds?" seems pretty hard for him to answer straight.

DeanValentine avatar
Dean Valentineis predicting NO at 23%

@BetsByAnon But under what circumstances would he have had to explain any such things before FTX's balance sheet was leaked in November?

BetsByAnon avatar
Bets By Anonis predicting YES at 15%

@DeanValentine Right about the moment customers are asked to send their holdings to Alameda in order to get it onto FTX?