Hey gang, I’m realising need to effectively donate to all the people that SBF ripped off. Thinking about this logically we now have major attention on EA, more than ever. IF SBF has effectively crippled EA’s reputation -- We need to strike back and shift the narrative, otherwise this movement might be dead overnight. Think about how severely this will limit the amount of $ donated effectively in future
Oh wow, I didn’t understand this comment the first time I saw it and didn’t bother to click. In case there’s anyone else like me, the tweet says “Doing my part to get us refunded” with a screenshot of their comment.
The losses are likely in the billions. Even assuming we could come up with (say) $100MM, that might move the needle from a reported 9.4B shortfall to . . . 9.3B. That’s not going to change the narrative in any meaningful way.
One could argue for the EA community returning every dollar ever received from FTX-aligned sources . . . but I’m just not sure how much even that would move the needle of public opinion. Especially because the bad PR will be heaviest in the next few weeks to months, and it would seem exceedingly difficult to come up with that amount of money that fast.
Given SBF’s EA-ness and strong embeddedness within EA, I support the idea of making it an EA cause area to pay back people he effectively stole money from and used in EA (until that’s done), if he stole money from people and SBF + normal legal channels don’t suffice to return it.
It doesn’t strike me as appropriate for AMF to take money under the understanding it would go to combating malaria, and then divert that money to random crypto people instead.
But it strikes me as appropriate for the larger EA community to undo any harms like this that the community caused, at least if the conditions above hold, including if that means giving less to AMF, MIRI, etc. next year.
(MIRI didn’t receive any money from FTX AFAIK, and I have no idea whether AMF did; I’m just using them as examples of EA orgs. Diverting some donations from us to some pay-people-back fund would make sense if the above conditions hold, IMO, regardless of which specific EA orgs counterfactually lose money as a result. Because the point is to undo a harm caused by EAs, not to punish recipient orgs or undo every specific humanitarian benefit that occurred.)
Reimbursing people for the money spent within the EA ecosystem (if the above conditions hold) might take years, but it strikes me as doable, and much more obviously “our job” than trying to undo all harm FTX caused.
That said:
IF SBF has effectively crippled EA’s reputation -- We need to strike back and shift the narrative, otherwise this movement might be dead overnight. Think about how severely this will limit the amount of $ donated effectively in future
This strikes me as unreasonable and panicky, and also as weirdly manipulative. The reason we should help people we harmed is because it’s the right thing to do, not because we need to “shift the narrative”.
One could argue for the EA community returning every dollar ever received from FTX-aligned sources . . . but I’m just not sure how much even that would move the needle of public opinion.
I have the same objection here. If the facts shake out such that FTX indeed effectively stole money from people, and gave that money to EA projects, and no other channels suffice to reimburse the people who were stolen from, then as a community we should make it a priority to pay them back over time, up to the amount that was stolen from them and used by us.
I don’t know how likely it is that all those conditions will hold, but if they do hold, we should respond even if it’s useless for PR purposes and happens way too late to go viral on Twitter or whatever—because it’s the right thing to do. The reason to go fast would be because the victims will be more harmed if they’re parted from their money longer, not because going fast has better optics.
It does look increasingly to me like MIRI maybe couldn’t have plausibly benefited much from FTX, because the process whereby SBF selected EAs to work with was such as to exclude Bay Area people who knew about the early Alameda breakup, and this caused FTX money to go to Oxford-respectable longtermism and not to places like MIRI, not even dropping a mil there for old time’s sake while spending $135m on a stadium. I pinged SBF a couple of times to see if he wanted to have a conversation at some point; he never responded. I think there’s a legit sense in which SBF wasn’t one of ours, and it’s not clear to me that MIRI ought to be so easily lumped in with the groups that did potentially benefit.
This isn’t your fault, but you almost certainly “benefitted”—any (increased) funding from other EA funders is a counterfactual result of FTX generously funding other groups that otherwise would have competed for funds. And many regrantors certainly helped MIRI more indirectly by funding things you would have wanted that helped MIRI’s agenda in various ways.
This is not really a disagreement but rather nitpicking, but I noticed that according to https://intelligence.org/topcontributors/ MIRI did receive a donation from Alameda Research. Not a large one, but some money from the SBF ecosystem arrived at MIRI, apparently. But this does not really contradict the speculations you make about SBF avoiding certain Bay Area people.
Yep, well-picked nit, I was just told about that myself. Perfectly good substantive disagreement with the original thesis, imo, you don’t need to downplay it that much.
It also makes sense that the money would’ve come from the Alameda side (maybe in 2020 or early 2021 according to Wayback, somebody said) rather than the FTX side. Alameda would have had the Bay Areans, while FTX’s philanthropic side was constructed (exclusively?) out of Oxfordians.
Separate from the “Alameda did donate” point, I wouldn’t have predicted SBF to be excited about MIRI, because I modeled SBF and Future Fund as having very Will-MacAskill-y views of AI risk. (And of a constellation of other claims that are entangled with MIRI’s coolness, like “is causal decision theory good?”) That’s a very un-MIRI-ish set of heuristics and beliefs about the world, compared to the modal longtermist EA’s heuristics and beliefs.
I strongly think that those views are wrong, but they provide an alternative explanation for SBF liking Oxford more than the Bay, and I feel more wary of saying “the Bay shouldn’t help pay back money stolen by SBF and given to EA” insofar as SBF merely factually disagreed with rationalists a lot. (Also, I guess I lean toward the Bay helping out regardless.)
Apparently there was a $132K Alameda donation to MIRI in 2020 or early 2021. Didn’t actually know that.
Well, obviously they donated less to MIRI after they turned evil, and the stopping of MIRI donations was a huge red flag that we all should have noticed. Sage nod.
It’s worth pointing out that EA as a community does not directly generate revenue. If EA were to pay back funds already spent then those funds must come from new or existing donors, and that may be a particularly tough sell.
In addition, I think most of us are familiar with the Copenhagen Interpretation of Ethics. While returning donations from unethical sources would likely reflect well upon EA, soliciting new donations to pay back past donations may be perceived differently, especially in light of existing questions surrounding the knowledge and/or involvement of EA leadership.
soliciting new donations to pay back past donations may be perceived differently
I agree re soliciting money from non-EAs. But if the conditions hold for my proposal upthread, and the logic checks out upon further scrutiny, then I like the idea of committed large EA donors making this a priority.
And if they don’t cover the gap, I like the idea of committed small EA donors doing what we can here—not as the only thing we fund, but as part of our portfolio until the stolen money (if it was stolen) is returned.
I don’t have a lot to donate, but I would donate to a fund like that, if those conditions held. It seems to me like a “good citizenship” donation, and a donation in favor of integrity in a community I care about.
I don’t feel like I’m atoning for something bad I did, or for some sort of Communal Guilt—I don’t really see how that concept makes sense. But I do feel like I’d be contributing to EA being an honorable and principled thing, and also drawing a clear line in the sand that disincentivizes future “do evil things to give to EA” actions.
Seeking clarity: are you suggesting that the most appropriate course of action would be to return the present net balance of funds but not specifically plan to use future funds to offset losses because that would raise the difficulty bar on acquiring the new funds—“Why should I donate money to pay for someone else’s unethical actions”—or because repaying already spent funds rather than spending any future-donated money on more direct causes would be of lower utility overall?
Suppose that Alice is a committed EA in good standing with tons of social ties to other EAs, and Alice can achieve a lot more stuff in the world because her EA associations help establish her benevolence, integrity, etc.
Alice then goes on a burglary spree and steals $100 each from Bob, Carol, and Dan.
Alice burns Bob’s $100.
Alice donates Carol’s $100 to FHI.
Alice holds on to Dan’s $100, and ends up giving it back to Dan (or a court returns the money).
FHI spends $70 of the stolen money, before learning that it was stolen. Alice gets arrested, etc.
My proposal in this hypothetical is:
FHI keeps the $30 (if this is compatible with the law of the land), and doesn’t give $70 back either. (Because I think the generalized norm “the onus is on the donation recipient to give everything back” would destroy a lot of projects and hurt a lot of people who are innocent bystanders that had no reason to believe the money was stolen.)
The larger EA community makes it a priority to internally raise $100 and give it to Carol. (IRL, we could do this immediately because $100 is a small sum. If the sum is a lot larger, then we should try to pay it off over a few years, in a way that’s not thrashing a lot of EA careers and projects unnecessarily / keeps funding flows as stable as possible.)
EAs don’t make it a collective priority to raise money for Bob (or for Dan).
This is the norm I’m proposing in this super-simplified hypothetical. I’m pretty uncertain about how relevantly analogous it is to the FTX situation, however. So I’m separately interested in hearing disagreement with the proposed norm (if anyone disagrees), and disagreement with whether it applies here.
I snap my fingers to instantly teleport all unspent money in this category (totaling $n) back to its owners.
I snap my fingers to instantly start a conference call with the thirty largest EA donors that’s guaranteed to result in commitments to pay $n back to the same people ASAP.
I’d expect 1 to cause a lot of long-term damage to a big chunk of the world’s quality-adjusted philanthropic efforts. I don’t know what spread of things has received FTX money, but I’m imagining orgs and projects shuttering (orgs and projects that would have been fine if they’d had a year or two to secure new funding, rather than a few days or weeks), individuals suddenly scrambling to make rent or try to apply to other grant processes, and a cascade of EAs (and EA allies) being unable to keep formal and informal commitments they’ve already made (e.g., to employees).
This seems like a weird form of destruction to inflict on a bunch of bystanders who didn’t know anything about FTX’s dysfunctions. Money is fungible! If we think this is important, then it seems perverse to make things right in a way that messes up a bunch of other people’s lives or careers, when we have other options as a community.
Obviously not every project will be affected in a big way by giving back the money. But I’m still wary of orgs doing stuff that might contribute to a virtue-signaling equilibrium that destroys a lot of value (“I gave back X, which makes my friend look like an asshole if she doesn’t give back Y, which makes her friends look like assholes if...”), if there are in fact any better options.
And I feel broadly icky about putting the moral onus on recipient orgs and individuals to make things right (when they bear no more blame than any of the rest of us), as opposed to taking ownership of the responsibility at a community level. (Not the full responsibility for everything bad FTX-ish folks have ever done; but community-level responsibility for any wrongfully acquired funds the community has used.)
The nature of insolvency and fraud is that someone ends up getting hit with pain. The question is who should have to bear it. I don’t think it is correct to say that the community was the beneficiary of any ill-gotten funds; the true beneficiaries were the public. “The community” is even further removed from FTX than grantees, and the people who won’t be getting bednets (or whatever) because we diverted funds from effective charities are at an even further remove. I do not like the idea of diverting the pain from FTX depositors to bednet recipients unless that is morally obligatory. Without suggesting that the depositors are to blame, they are both better off and were more connected to the incident than the bednet recipients.
And I don’t see a moral obligation to return spent funds; the grantees were contracted to perform certain work that FTX-aligned people wanted done and they did it. We didn’t expect ordinary employees at Enron to give back their earned wages because their payor was a massive fraud.
However, the Enron employees were not morally entitled to unearned wages when that money could have gone to fraud victims. So too here. Because I think there generally is a moral obligation to return unused funds if those were the product of fraud, I am more inclined toward the idea of the community replacing those funds on behalf of grantees. But I would conceptually frame that a bit differently: we would be providing a grant to former FTX grantees to allow them to meet their obligation to refund unspent grant funds, so that their charitable work can continue. Maybe it’s just a matter of optics, but that feels subtly different than seeking monies to clean up the messes created by FTX.
The former, with the latter as its first order cause.
I think the most appropriate course of action won’t be discernable for a while. But returning the present net balance of funds is clearly the correct move, while the ethical decision regarding returning past funds is ambiguous.
Someone raised the point that if EAs try to offset the harm SBF caused, this creates a moral hazard of the form “people may be more willing to cause harm in the name of EA in the future, expecting other EAs to offset that harm”.
I think that’s a stronger objection to “offset all of SBF’s harms” (which I don’t endorse anyway) than to “collectively give back the amount EA received”, but maybe it will shift my view once I’ve chewed on it a bit more. At a glance, I don’t think I’d expect this concern to be a dominant factor?
I disagree with paying back being obviously the right thing to do. The implications of “pulling back” money whenever something large shady appears would be difficult to handle, and it would be costly. (If you are arguing that the current case is special and in future cases of alleged / proven financial crime we should evaluate case by case then I am very interested in what the specific argument is.)
I would look into options for vetting integrity of big donors in the future as the right thing to do though.
I agree with this. Actually, I think we could go further and initiate some form of productive public dialog with the wider world on this question. “Do you think that we ought to take money in the EA ecosystem and pay it back to people [potentially] defrauded by FTX, or should we put this money into the charities for which it was intended?”
That seems like responsible stewardship, and I’d expect people’s opinions would vary widely.
The question would be how we’d make such decisions, how we’d hold this dialog, and how much time and energy we’d want to put into that endeavor. One way might be to solicit input from groups that we think ought to have a say: charities we donate to, ethical thinkers, community leaders, and people who lost money in the FTX meltdown, to name a few. We could potentially make the decision by running some sort of vote, which could be as sophisticated as we like. We could vote on whether to return the money, but also how much of it should be returned.
Just brainstorming here, I don’t expect that these are the ideal way to deal with this. Just a starting point.
Do you think that we ought to take money in the EA ecosystem and pay it back to people [potentially] defrauded by FTX, or should we put this money into the charities for which it was intended?
I say give the money back (at the EA-community level, not the individual org/project level), and let the theft victim decide if they want to redonate their money to charity. It’s their money, after all!
(If the thing that happened is basically theft, and if they don’t get their money back by some other channel. I’d be interested to hear counter-arguments on either of those fronts, or on the general policy I’m suggesting.)
Public discussions sound great too, but we can invite a public conversation without using that as a reason to put off making decisions about this.
(I do think we should think hard about the relevant factors here, before acting. There are a ton of things I expect to be confusing here; maybe the whole idea just doesn’t make sense because of some subtlety about comparing the counterfactuals. But I’m guessing a large public conversation wouldn’t give us much additional insight here, and would be useful for reasons other than improving the quality of our decision.)
Hi effectenator. I do appreciate the concern for EA’s reputation. (And, charitably, I assume you’re also factoring in that this might just be the right thing to do anyway regardless of the effects on EA’s image.)
But I also wonder: Do you think non-complicit FTX employees/freelancers should donate their salaries to these people too?
My guess is that the large majority of people would answer “No.” But I’m not sure that this case is all that different from grants. Future Fund wasn’t distributing gifts; it was distributing money to fund work. In fact, when I received an EA grant (not from FTX), I assumed that legally it would be classified as self-employment income and indeed that also turned out to be the main recommendation of the people giving me the grant, so it’s possible that in many cases these are in fact legally the same thing.
I think there is still a weak case for people paying back (especially unspent) salaries / severance pay / fees / grants—perhaps staff and grantees had a moral obligation to have done more due diligence before accepting money from FTX, and this is plausibly a much stronger consideration than the “prudential” obligations of people to do sufficient due diligence before trusting FTX with their money.
But still, in all of these scenarios you end up with a situation where many people who owned $XXX a week ago and were making financial decisions accordingly, now no longer have it and are potentially in financial difficulty. You’d just be changing who the victims are.
I think your metaphor is a good one, but doesnt entirely get you where you want to go.
Insolvency is all about distributing pain. I agree that non-executive employees (including indirect employees through grants) have a high priority moral claim for work already performed and should not be clawed back. There are also practical utilitarian reasons for this rule more generally.
However, I don’t think they have this moral superpriority status for work performed after the insolvency became known. Taking away the expectation for future compensation for future work that does not benefit those with high-priority moral claims on the insolvent entity is better than the alternative of foisting more losses on depositors than absolutely necessary. Depositors have actual losses, not losses of expectancy, and there are utilitarian reasons to grant them special protections over ordinary claimants.
Moreover, most individuals funded by grant work can find alternative employment in a fairly short period of time. So they are in a much better place to mitigate loses than depositors..which is another reason I think their claims are weaker.
EA is not an insurance fund. Beyond simply not being feasible, paying back all lost funds would create a significant moral hazard and attract bad actors who wish to appear trustworthy to the public.
If a con artist happens to give money to a beggar, you don’t go and try to get the money back off the beggar, it’s just gone.
It’s rough, but rough things happen in this world. I’m not saying anyone deserved to lose their money, but it’s not like people were unaware that investing in crypto is inherently risky.
This raises a good point. Charities such as AMF or GiveDirectly giving back money would be harming people who are mostly (much) worse off than the average FTX customer who’s lost money.
However, most of the FTX charitable donations seem to have been (via the Future Fund) to longtermist orgs that pay people in rich countries high salaries. It’s not unreasonable to suggest that these orgs/people take significant pay cuts to give money back, especially if they can still do their work living more frugally.
However, the difficulty lies with the amounts. The Future Fund has given out ~$200M. If the money lost by FTX customers amounts to ~$10B, that’s not going to make a dent. You can either pay everyone back 2%, which might be appreciated but mostly feel like nothing (or perhaps even a slap in the face?), or you could preferentially pay certain people back 100%, but then how do you choose who is in and who is out? (A lottery? Pay back smaller balances first? Means testing to pay back the poorest first?) Or should EA/Longtermism keep paying money back until everyone is whole? Seems like it would take years at best. It’s a mess :(
A lottery seems most reasonable to me. E.g.: keep selecting random individuals who haven’t gotten their money back; pay them (maybe up to some high cap?); repeat until you’ve paid out a total of ~$200M.
I’m not sure if it’s that simple. It seems quite likely that with a lottery there would be at least a few cases of relatively rich people (say people who had >$200k and lost $100k of it on FTX) getting paid back. This would inevitably lead to (justified) anger and complaints from people who were left much worse off (e.g. those who only had $10k and lost all of it). Maybe it needs to be means tested for hardship? It would take a lot more work, but perhaps there could be an application process and vetting by an independent (from EA) panel? That would still be open to manipulation though I guess (e.g. people shouting the loudest on social media).
I’d vote for an algorithm that is transparent and has limited if any discretion. Maybe the end result would be somewhat suboptimal to discretionary independent-panel vetting, but it would be critical that the process be seen as fair.
Plus, a complicated procedure will eat up a lot of the funds intended for victims, unless you think the EA community should both return any FTX money and pay for an involved claims process. That goes beyond mere disgorgement.
Hey gang, I’m realising need to effectively donate to all the people that SBF ripped off. Thinking about this logically we now have major attention on EA, more than ever. IF SBF has effectively crippled EA’s reputation -- We need to strike back and shift the narrative, otherwise this movement might be dead overnight. Think about how severely this will limit the amount of $ donated effectively in future
To be clear, this is an account that joined from Twitter to post this comment (link).
Oh wow, I didn’t understand this comment the first time I saw it and didn’t bother to click. In case there’s anyone else like me, the tweet says “Doing my part to get us refunded” with a screenshot of their comment.
Yep; I interpret this comment as an attempt to manipulate EAs, not as an honest good-faith proposal.
Even if Mitch is a committed EA (which seems unlikely), it fails to mention the obvious conflict of interest.
The losses are likely in the billions. Even assuming we could come up with (say) $100MM, that might move the needle from a reported 9.4B shortfall to . . . 9.3B. That’s not going to change the narrative in any meaningful way.
One could argue for the EA community returning every dollar ever received from FTX-aligned sources . . . but I’m just not sure how much even that would move the needle of public opinion. Especially because the bad PR will be heaviest in the next few weeks to months, and it would seem exceedingly difficult to come up with that amount of money that fast.
A couple of hours ago, I tweeted:
Reimbursing people for the money spent within the EA ecosystem (if the above conditions hold) might take years, but it strikes me as doable, and much more obviously “our job” than trying to undo all harm FTX caused.
That said:
This strikes me as unreasonable and panicky, and also as weirdly manipulative. The reason we should help people we harmed is because it’s the right thing to do, not because we need to “shift the narrative”.
I have the same objection here. If the facts shake out such that FTX indeed effectively stole money from people, and gave that money to EA projects, and no other channels suffice to reimburse the people who were stolen from, then as a community we should make it a priority to pay them back over time, up to the amount that was stolen from them and used by us.
I don’t know how likely it is that all those conditions will hold, but if they do hold, we should respond even if it’s useless for PR purposes and happens way too late to go viral on Twitter or whatever—because it’s the right thing to do. The reason to go fast would be because the victims will be more harmed if they’re parted from their money longer, not because going fast has better optics.
It does look increasingly to me like MIRI maybe couldn’t have plausibly benefited much from FTX, because the process whereby SBF selected EAs to work with was such as to exclude Bay Area people who knew about the early Alameda breakup, and this caused FTX money to go to Oxford-respectable longtermism and not to places like MIRI, not even dropping a mil there for old time’s sake while spending $135m on a stadium. I pinged SBF a couple of times to see if he wanted to have a conversation at some point; he never responded. I think there’s a legit sense in which SBF wasn’t one of ours, and it’s not clear to me that MIRI ought to be so easily lumped in with the groups that did potentially benefit.
This isn’t your fault, but you almost certainly “benefitted”—any (increased) funding from other EA funders is a counterfactual result of FTX generously funding other groups that otherwise would have competed for funds. And many regrantors certainly helped MIRI more indirectly by funding things you would have wanted that helped MIRI’s agenda in various ways.
This is not really a disagreement but rather nitpicking, but I noticed that according to https://intelligence.org/topcontributors/ MIRI did receive a donation from Alameda Research. Not a large one, but some money from the SBF ecosystem arrived at MIRI, apparently. But this does not really contradict the speculations you make about SBF avoiding certain Bay Area people.
Yep, well-picked nit, I was just told about that myself. Perfectly good substantive disagreement with the original thesis, imo, you don’t need to downplay it that much.
It also makes sense that the money would’ve come from the Alameda side (maybe in 2020 or early 2021 according to Wayback, somebody said) rather than the FTX side. Alameda would have had the Bay Areans, while FTX’s philanthropic side was constructed (exclusively?) out of Oxfordians.
Separate from the “Alameda did donate” point, I wouldn’t have predicted SBF to be excited about MIRI, because I modeled SBF and Future Fund as having very Will-MacAskill-y views of AI risk. (And of a constellation of other claims that are entangled with MIRI’s coolness, like “is causal decision theory good?”) That’s a very un-MIRI-ish set of heuristics and beliefs about the world, compared to the modal longtermist EA’s heuristics and beliefs.
I strongly think that those views are wrong, but they provide an alternative explanation for SBF liking Oxford more than the Bay, and I feel more wary of saying “the Bay shouldn’t help pay back money stolen by SBF and given to EA” insofar as SBF merely factually disagreed with rationalists a lot. (Also, I guess I lean toward the Bay helping out regardless.)
Apparently there was a $132K Alameda donation to MIRI in 2020 or early 2021. Didn’t actually know that.
Well, obviously they donated less to MIRI after they turned evil, and the stopping of MIRI donations was a huge red flag that we all should have noticed. Sage nod.
For the sake of completeness, here’s a thread with all the financial interactions between MIRI and FTX/Alameda/SBF: https://twitter.com/robbensinger/status/1595893840484843521
It’s worth pointing out that EA as a community does not directly generate revenue. If EA were to pay back funds already spent then those funds must come from new or existing donors, and that may be a particularly tough sell.
In addition, I think most of us are familiar with the Copenhagen Interpretation of Ethics. While returning donations from unethical sources would likely reflect well upon EA, soliciting new donations to pay back past donations may be perceived differently, especially in light of existing questions surrounding the knowledge and/or involvement of EA leadership.
I agree re soliciting money from non-EAs. But if the conditions hold for my proposal upthread, and the logic checks out upon further scrutiny, then I like the idea of committed large EA donors making this a priority.
And if they don’t cover the gap, I like the idea of committed small EA donors doing what we can here—not as the only thing we fund, but as part of our portfolio until the stolen money (if it was stolen) is returned.
I don’t have a lot to donate, but I would donate to a fund like that, if those conditions held. It seems to me like a “good citizenship” donation, and a donation in favor of integrity in a community I care about.
I don’t feel like I’m atoning for something bad I did, or for some sort of Communal Guilt—I don’t really see how that concept makes sense. But I do feel like I’d be contributing to EA being an honorable and principled thing, and also drawing a clear line in the sand that disincentivizes future “do evil things to give to EA” actions.
Seeking clarity: are you suggesting that the most appropriate course of action would be to return the present net balance of funds but not specifically plan to use future funds to offset losses because that would raise the difficulty bar on acquiring the new funds—“Why should I donate money to pay for someone else’s unethical actions”—or because repaying already spent funds rather than spending any future-donated money on more direct causes would be of lower utility overall?
Suppose that Alice is a committed EA in good standing with tons of social ties to other EAs, and Alice can achieve a lot more stuff in the world because her EA associations help establish her benevolence, integrity, etc.
Alice then goes on a burglary spree and steals $100 each from Bob, Carol, and Dan.
Alice burns Bob’s $100.
Alice donates Carol’s $100 to FHI.
Alice holds on to Dan’s $100, and ends up giving it back to Dan (or a court returns the money).
FHI spends $70 of the stolen money, before learning that it was stolen. Alice gets arrested, etc.
My proposal in this hypothetical is:
FHI keeps the $30 (if this is compatible with the law of the land), and doesn’t give $70 back either. (Because I think the generalized norm “the onus is on the donation recipient to give everything back” would destroy a lot of projects and hurt a lot of people who are innocent bystanders that had no reason to believe the money was stolen.)
The larger EA community makes it a priority to internally raise $100 and give it to Carol.
(IRL, we could do this immediately because $100 is a small sum. If the sum is a lot larger, then we should try to pay it off over a few years, in a way that’s not thrashing a lot of EA careers and projects unnecessarily / keeps funding flows as stable as possible.)EAs don’t make it a collective priority to raise money for Bob (or for Dan).
This is the norm I’m proposing in this super-simplified hypothetical. I’m pretty uncertain about how relevantly analogous it is to the FTX situation, however. So I’m separately interested in hearing disagreement with the proposed norm (if anyone disagrees), and disagreement with whether it applies here.
Why not give back unspent funds though?
Comparing two scenarios:
I snap my fingers to instantly teleport all unspent money in this category (totaling $n) back to its owners.
I snap my fingers to instantly start a conference call with the thirty largest EA donors that’s guaranteed to result in commitments to pay $n back to the same people ASAP.
I’d expect 1 to cause a lot of long-term damage to a big chunk of the world’s quality-adjusted philanthropic efforts. I don’t know what spread of things has received FTX money, but I’m imagining orgs and projects shuttering (orgs and projects that would have been fine if they’d had a year or two to secure new funding, rather than a few days or weeks), individuals suddenly scrambling to make rent or try to apply to other grant processes, and a cascade of EAs (and EA allies) being unable to keep formal and informal commitments they’ve already made (e.g., to employees).
This seems like a weird form of destruction to inflict on a bunch of bystanders who didn’t know anything about FTX’s dysfunctions. Money is fungible! If we think this is important, then it seems perverse to make things right in a way that messes up a bunch of other people’s lives or careers, when we have other options as a community.
Obviously not every project will be affected in a big way by giving back the money. But I’m still wary of orgs doing stuff that might contribute to a virtue-signaling equilibrium that destroys a lot of value (“I gave back X, which makes my friend look like an asshole if she doesn’t give back Y, which makes her friends look like assholes if...”), if there are in fact any better options.
And I feel broadly icky about putting the moral onus on recipient orgs and individuals to make things right (when they bear no more blame than any of the rest of us), as opposed to taking ownership of the responsibility at a community level. (Not the full responsibility for everything bad FTX-ish folks have ever done; but community-level responsibility for any wrongfully acquired funds the community has used.)
The nature of insolvency and fraud is that someone ends up getting hit with pain. The question is who should have to bear it. I don’t think it is correct to say that the community was the beneficiary of any ill-gotten funds; the true beneficiaries were the public. “The community” is even further removed from FTX than grantees, and the people who won’t be getting bednets (or whatever) because we diverted funds from effective charities are at an even further remove. I do not like the idea of diverting the pain from FTX depositors to bednet recipients unless that is morally obligatory. Without suggesting that the depositors are to blame, they are both better off and were more connected to the incident than the bednet recipients.
And I don’t see a moral obligation to return spent funds; the grantees were contracted to perform certain work that FTX-aligned people wanted done and they did it. We didn’t expect ordinary employees at Enron to give back their earned wages because their payor was a massive fraud.
However, the Enron employees were not morally entitled to unearned wages when that money could have gone to fraud victims. So too here. Because I think there generally is a moral obligation to return unused funds if those were the product of fraud, I am more inclined toward the idea of the community replacing those funds on behalf of grantees. But I would conceptually frame that a bit differently: we would be providing a grant to former FTX grantees to allow them to meet their obligation to refund unspent grant funds, so that their charitable work can continue. Maybe it’s just a matter of optics, but that feels subtly different than seeking monies to clean up the messes created by FTX.
The former, with the latter as its first order cause.
I think the most appropriate course of action won’t be discernable for a while. But returning the present net balance of funds is clearly the correct move, while the ethical decision regarding returning past funds is ambiguous.
Someone raised the point that if EAs try to offset the harm SBF caused, this creates a moral hazard of the form “people may be more willing to cause harm in the name of EA in the future, expecting other EAs to offset that harm”.
I think that’s a stronger objection to “offset all of SBF’s harms” (which I don’t endorse anyway) than to “collectively give back the amount EA received”, but maybe it will shift my view once I’ve chewed on it a bit more. At a glance, I don’t think I’d expect this concern to be a dominant factor?
I disagree with paying back being obviously the right thing to do. The implications of “pulling back” money whenever something large shady appears would be difficult to handle, and it would be costly. (If you are arguing that the current case is special and in future cases of alleged / proven financial crime we should evaluate case by case then I am very interested in what the specific argument is.)
I would look into options for vetting integrity of big donors in the future as the right thing to do though.
I agree with this. Actually, I think we could go further and initiate some form of productive public dialog with the wider world on this question. “Do you think that we ought to take money in the EA ecosystem and pay it back to people [potentially] defrauded by FTX, or should we put this money into the charities for which it was intended?”
That seems like responsible stewardship, and I’d expect people’s opinions would vary widely.
The question would be how we’d make such decisions, how we’d hold this dialog, and how much time and energy we’d want to put into that endeavor. One way might be to solicit input from groups that we think ought to have a say: charities we donate to, ethical thinkers, community leaders, and people who lost money in the FTX meltdown, to name a few. We could potentially make the decision by running some sort of vote, which could be as sophisticated as we like. We could vote on whether to return the money, but also how much of it should be returned.
Just brainstorming here, I don’t expect that these are the ideal way to deal with this. Just a starting point.
I say give the money back (at the EA-community level, not the individual org/project level), and let the theft victim decide if they want to redonate their money to charity. It’s their money, after all!
(If the thing that happened is basically theft, and if they don’t get their money back by some other channel. I’d be interested to hear counter-arguments on either of those fronts, or on the general policy I’m suggesting.)
Public discussions sound great too, but we can invite a public conversation without using that as a reason to put off making decisions about this.
(I do think we should think hard about the relevant factors here, before acting. There are a ton of things I expect to be confusing here; maybe the whole idea just doesn’t make sense because of some subtlety about comparing the counterfactuals. But I’m guessing a large public conversation wouldn’t give us much additional insight here, and would be useful for reasons other than improving the quality of our decision.)
Hi effectenator. I do appreciate the concern for EA’s reputation. (And, charitably, I assume you’re also factoring in that this might just be the right thing to do anyway regardless of the effects on EA’s image.)
But I also wonder: Do you think non-complicit FTX employees/freelancers should donate their salaries to these people too?
My guess is that the large majority of people would answer “No.” But I’m not sure that this case is all that different from grants. Future Fund wasn’t distributing gifts; it was distributing money to fund work. In fact, when I received an EA grant (not from FTX), I assumed that legally it would be classified as self-employment income and indeed that also turned out to be the main recommendation of the people giving me the grant, so it’s possible that in many cases these are in fact legally the same thing.
I think there is still a weak case for people paying back (especially unspent) salaries / severance pay / fees / grants—perhaps staff and grantees had a moral obligation to have done more due diligence before accepting money from FTX, and this is plausibly a much stronger consideration than the “prudential” obligations of people to do sufficient due diligence before trusting FTX with their money.
But still, in all of these scenarios you end up with a situation where many people who owned $XXX a week ago and were making financial decisions accordingly, now no longer have it and are potentially in financial difficulty. You’d just be changing who the victims are.
I think your metaphor is a good one, but doesnt entirely get you where you want to go.
Insolvency is all about distributing pain. I agree that non-executive employees (including indirect employees through grants) have a high priority moral claim for work already performed and should not be clawed back. There are also practical utilitarian reasons for this rule more generally.
However, I don’t think they have this moral superpriority status for work performed after the insolvency became known. Taking away the expectation for future compensation for future work that does not benefit those with high-priority moral claims on the insolvent entity is better than the alternative of foisting more losses on depositors than absolutely necessary. Depositors have actual losses, not losses of expectancy, and there are utilitarian reasons to grant them special protections over ordinary claimants.
Moreover, most individuals funded by grant work can find alternative employment in a fairly short period of time. So they are in a much better place to mitigate loses than depositors..which is another reason I think their claims are weaker.
Yes I think I agree on both counts.
[Edited to add: Maybe after including severance pay or equivalent? Sudden loss of income seems much worse than sudden loss of savings.]
EA is not an insurance fund. Beyond simply not being feasible, paying back all lost funds would create a significant moral hazard and attract bad actors who wish to appear trustworthy to the public.
Dear God no.
If a con artist happens to give money to a beggar, you don’t go and try to get the money back off the beggar, it’s just gone.
It’s rough, but rough things happen in this world. I’m not saying anyone deserved to lose their money, but it’s not like people were unaware that investing in crypto is inherently risky.
This raises a good point. Charities such as AMF or GiveDirectly giving back money would be harming people who are mostly (much) worse off than the average FTX customer who’s lost money.
However, most of the FTX charitable donations seem to have been (via the Future Fund) to longtermist orgs that pay people in rich countries high salaries. It’s not unreasonable to suggest that these orgs/people take significant pay cuts to give money back, especially if they can still do their work living more frugally.
However, the difficulty lies with the amounts. The Future Fund has given out ~$200M. If the money lost by FTX customers amounts to ~$10B, that’s not going to make a dent. You can either pay everyone back 2%, which might be appreciated but mostly feel like nothing (or perhaps even a slap in the face?), or you could preferentially pay certain people back 100%, but then how do you choose who is in and who is out? (A lottery? Pay back smaller balances first? Means testing to pay back the poorest first?) Or should EA/Longtermism keep paying money back until everyone is whole? Seems like it would take years at best. It’s a mess :(
A lottery seems most reasonable to me. E.g.: keep selecting random individuals who haven’t gotten their money back; pay them (maybe up to some high cap?); repeat until you’ve paid out a total of ~$200M.
I’m not sure if it’s that simple. It seems quite likely that with a lottery there would be at least a few cases of relatively rich people (say people who had >$200k and lost $100k of it on FTX) getting paid back. This would inevitably lead to (justified) anger and complaints from people who were left much worse off (e.g. those who only had $10k and lost all of it). Maybe it needs to be means tested for hardship? It would take a lot more work, but perhaps there could be an application process and vetting by an independent (from EA) panel? That would still be open to manipulation though I guess (e.g. people shouting the loudest on social media).
I’d vote for an algorithm that is transparent and has limited if any discretion. Maybe the end result would be somewhat suboptimal to discretionary independent-panel vetting, but it would be critical that the process be seen as fair.
Plus, a complicated procedure will eat up a lot of the funds intended for victims, unless you think the EA community should both return any FTX money and pay for an involved claims process. That goes beyond mere disgorgement.
Nice try—I like your on-the-nose username