I think I’m significantly more involved than most people I know in tying the fate of effective altruism in general, and Rethink Priorities in particular, with that of FTX. This probably led to rather bad consequences ex post, and I’m very sorry for this.
I don’t think I’m meaningfully responsible for the biggest potential issue with FTX. I was not aware of the alleged highly unethical behavior (including severe mismanagement of consumer funds) at FTX. I also have not, to my knowledge, contributed meaningfully to the relevant reputational laundering or branding that led innocent external depositors to put money in FTX. The lack of influence there is not because I had any relevant special knowledge of FTX, but because I have primarily focused on building an audience within the effective altruism community, who are typically skeptical of cryptocurrency, and because I have actively avoided encouraging others to invest in cryptocurrency. I’m personally pretty skeptical about the alleged social value of pure financialization in general and cryptocurrency in particular, and also I’ve always thought of crypto as a substantially more risky asset than many retail investors are led to believe.[1]
However, I think I contributed both reputationally and operationally to tying FTX in with the effective altruism community. Most saliently, I’ve done the following:
I decided on my team’s strategic focus on longtermist megaprojects in large part due to anticipated investments in world-saving activities from FTX. Clearly this did not pan out.
In much of 2022, I strongly encouraged the rest of Rethink Priorities, including my manager, to seriously consider incorporating influencing the Future Fund into our organization’s Theory of Change.
I was a regranter for the Future Fund in a personal capacity (not RP-related). I was happy about the grants I gave, but in retrospect of course this probably led to more headaches and liabilities to my regrantees.
I tried to be cautious and tempered in my messaging to them, but I did not internalize that transactions are not final, and falsely implied in my dealings that all of the financial uncertainty comes before donations are in people’s bank accounts.
In late 2021 and early 2022, I implicitly and explicitly encouraged EAs to visit FTX in the Bahamas. I would not be surprised if this was counterfactual: I do think my words have nontrivial weight in EA, and I also tried to help people derisk their visit to FTX.
I tried to encourage some programmers I know to work at FTX, because I thought it was an unusually good and high-impact career opportunity for programmers.
To the best of my knowledge, my encouragement did not counterfactually lead anybody to join FTX (though, alas, not for want of trying on my end).
I did the above without much due diligence on FTX, or thinking too hard about where they got their money. I did think about failure modes, but did not manage to land on the current one, and was also probably systematically overly optimistic about their general prospects. Overall, I’m not entirely sure what mistakes in reasoning or judgment lead me to mistakenly bet on FTX. [2]But clearly this seemed to be a very wrong call ex post, and likely led to a lot of harm both to EA’s brand, and to individuals.
For all of that, I am very sorry. I will reflect on this, and hopefully do better going forward.
Note that this decision process did not apply to my own investments, where I think it makes sense to be substantially more risk-neutral. I had a non-trivial internal probability that crypto in general, and FTX in particular, might be worth a lot more in the coming years, and in accordance with this belief I did lose a lot of money on FTX.
Frustratingly, I did have access to some credible-to-me private information that led me on net to be more rather than less trusting of them. This example is particularly salient for me on the dangers of trusting private information over less biased reasoning processes like base rates.
(The below does not necessarily excuse defects in awareness, understanding or risk management around FTX/SBF from the most senior EA leaders, which should be very sophisticated.)
I think I contributed…Overall, I’m not entirely sure what mistakes in reasoning or judgment lead me to mistakenly bet on FTX.
I tried to encourage some programmers I know to work at FTX, because I thought it was an unusually good and high-impact career opportunity for programmers.
In late 2021 and early 2022, I implicitly and explicitly encouraged EAs to visit FTX in the Bahamas. I would not be surprised if this was counterfactual: I do think my words have nontrivial weight in EA, and I also tried to help people derisk their visit to FTX.
Prior to November, the idea that FTX was strange or dangerous was not known to even very experienced people in cryptocurrency.
As a datapoint from the non-EA world, several very smart people desperately wanted to work for FTX, because of the status, culture and excitement around FTX and SBF, even taking pay cuts to do so. For example, one such person was deeply into crypto, seasoned and older, e.g. 800K TC L6+ at Google (higher in salary/level to Jeff K for example).
I tried to be cautious and tempered in my messaging to them, but I did not internalize that transactions are not final, and falsely implied in my dealings that all of the financial uncertainty comes before donations are in people’s bank accounts.
This risk of clawback is probably truly unusual. I think this situation (massive fraud leading to total collapse) is one of the only situations where this could happen.
According to the criminal indictment filed by the U.S. Attorneys Office for the Southern District of New York, and also the civil complaints filed by the SEC and the CFTC: FTX/Alameda was a criminal fraud enterprise from inception. FTX embezzled its customers’ money.
Those who received grants from FTX received stolen money. It’s not only a matter of bankruptcy potentially clawing the money back; now if a grantee is holding unspent funds they are holding criminal proceeds. (And don’t rush to spend the funds; since the date FTX filed for bankruptcy, grant recipients were on notice of this possibility.)
This presents a real world ethical problem for those grantees. Do they return the criminal funds they received? Or do they cast themselves as victims and rationalize they are somehow entitled to keep the money they now know was stolen?
When I posted earlier that grantees should return the unspent funds there was a lot of strained arguing that returning the money was not required. This was eyebrow-raising coming from a group that uses “ethical” in its name.
If you received a grant from FTX then you received stolen money from a criminal fraudster. It would be asking too much to return money already spent before FTX filed for bankruptcy. But what about the unspent money? Are people going to voluntarily return that unspent money to the FTX bankruptcy estate? Or is ethics only something to preach to others, but to avoid when it impacts your own money?
Or is ethics only something to preach to others, but to avoid when it impacts your own money?
To be clear, even ignoring that I’ve lost most of my net worth on FTX, I’ve donated 10-20% of my income since 2015, I was an unpaid volunteer for many months in 2017 and 2020, and relative to my past FAANG job, I took substantially greater than a 50% paycut because I think my current job is morally important enough. This isn’t much relative to the rest of this community; I’m sure many EAs made larger sacrifices to live in accordance with their ethical values. In many ways I fall short of my ethical ideals, but what have you done?
Or is ethics only something to preach to others, but to avoid when it impacts your own money?
You say this as if the EA community has made a habit of policing the ethics of the sources of other people’s funding. But whenever people have demanded non-EA charitable organizations to relinquish money from the villain of the year, e.g. Jimmy Saville or the Sackler family, I have always thought that was silly, and I suspect other EAs tend to agree.
A Personal Apology
I think I’m significantly more involved than most people I know in tying the fate of effective altruism in general, and Rethink Priorities in particular, with that of FTX. This probably led to rather bad consequences ex post, and I’m very sorry for this.
I don’t think I’m meaningfully responsible for the biggest potential issue with FTX. I was not aware of the alleged highly unethical behavior (including severe mismanagement of consumer funds) at FTX. I also have not, to my knowledge, contributed meaningfully to the relevant reputational laundering or branding that led innocent external depositors to put money in FTX. The lack of influence there is not because I had any relevant special knowledge of FTX, but because I have primarily focused on building an audience within the effective altruism community, who are typically skeptical of cryptocurrency, and because I have actively avoided encouraging others to invest in cryptocurrency. I’m personally pretty skeptical about the alleged social value of pure financialization in general and cryptocurrency in particular, and also I’ve always thought of crypto as a substantially more risky asset than many retail investors are led to believe.[1]
However, I think I contributed both reputationally and operationally to tying FTX in with the effective altruism community. Most saliently, I’ve done the following:
I decided on my team’s strategic focus on longtermist megaprojects in large part due to anticipated investments in world-saving activities from FTX. Clearly this did not pan out.
In much of 2022, I strongly encouraged the rest of Rethink Priorities, including my manager, to seriously consider incorporating influencing the Future Fund into our organization’s Theory of Change.
Fortunately RP’s leadership was more generally skeptical of tying in our fate with cryptocurrency than I was, and took steps to minimize exposure.
I was a regranter for the Future Fund in a personal capacity (not RP-related). I was happy about the grants I gave, but in retrospect of course this probably led to more headaches and liabilities to my regrantees.
I tried to be cautious and tempered in my messaging to them, but I did not internalize that transactions are not final, and falsely implied in my dealings that all of the financial uncertainty comes before donations are in people’s bank accounts.
In late 2021 and early 2022, I implicitly and explicitly encouraged EAs to visit FTX in the Bahamas. I would not be surprised if this was counterfactual: I do think my words have nontrivial weight in EA, and I also tried to help people derisk their visit to FTX.
I tried to encourage some programmers I know to work at FTX, because I thought it was an unusually good and high-impact career opportunity for programmers.
To the best of my knowledge, my encouragement did not counterfactually lead anybody to join FTX (though, alas, not for want of trying on my end).
I did the above without much due diligence on FTX, or thinking too hard about where they got their money. I did think about failure modes, but did not manage to land on the current one, and was also probably systematically overly optimistic about their general prospects. Overall, I’m not entirely sure what mistakes in reasoning or judgment lead me to mistakenly bet on FTX. [2]But clearly this seemed to be a very wrong call ex post, and likely led to a lot of harm both to EA’s brand, and to individuals.
For all of that, I am very sorry. I will reflect on this, and hopefully do better going forward.
Note that this decision process did not apply to my own investments, where I think it makes sense to be substantially more risk-neutral. I had a non-trivial internal probability that crypto in general, and FTX in particular, might be worth a lot more in the coming years, and in accordance with this belief I did lose a lot of money on FTX.
Frustratingly, I did have access to some credible-to-me private information that led me on net to be more rather than less trusting of them. This example is particularly salient for me on the dangers of trusting private information over less biased reasoning processes like base rates.
(The below does not necessarily excuse defects in awareness, understanding or risk management around FTX/SBF from the most senior EA leaders, which should be very sophisticated.)
Prior to November, the idea that FTX was strange or dangerous was not known to even very experienced people in cryptocurrency.
As a datapoint from the non-EA world, several very smart people desperately wanted to work for FTX, because of the status, culture and excitement around FTX and SBF, even taking pay cuts to do so. For example, one such person was deeply into crypto, seasoned and older, e.g. 800K TC L6+ at Google (higher in salary/level to Jeff K for example).
This risk of clawback is probably truly unusual. I think this situation (massive fraud leading to total collapse) is one of the only situations where this could happen.
According to the criminal indictment filed by the U.S. Attorneys Office for the Southern District of New York, and also the civil complaints filed by the SEC and the CFTC: FTX/Alameda was a criminal fraud enterprise from inception. FTX embezzled its customers’ money.
Those who received grants from FTX received stolen money. It’s not only a matter of bankruptcy potentially clawing the money back; now if a grantee is holding unspent funds they are holding criminal proceeds. (And don’t rush to spend the funds; since the date FTX filed for bankruptcy, grant recipients were on notice of this possibility.)
This presents a real world ethical problem for those grantees. Do they return the criminal funds they received? Or do they cast themselves as victims and rationalize they are somehow entitled to keep the money they now know was stolen?
When I posted earlier that grantees should return the unspent funds there was a lot of strained arguing that returning the money was not required. This was eyebrow-raising coming from a group that uses “ethical” in its name.
If you received a grant from FTX then you received stolen money from a criminal fraudster. It would be asking too much to return money already spent before FTX filed for bankruptcy. But what about the unspent money? Are people going to voluntarily return that unspent money to the FTX bankruptcy estate? Or is ethics only something to preach to others, but to avoid when it impacts your own money?
To be clear, even ignoring that I’ve lost most of my net worth on FTX, I’ve donated 10-20% of my income since 2015, I was an unpaid volunteer for many months in 2017 and 2020, and relative to my past FAANG job, I took substantially greater than a 50% paycut because I think my current job is morally important enough. This isn’t much relative to the rest of this community; I’m sure many EAs made larger sacrifices to live in accordance with their ethical values. In many ways I fall short of my ethical ideals, but what have you done?
Not everyone has the same ethical beliefs.
You say this as if the EA community has made a habit of policing the ethics of the sources of other people’s funding. But whenever people have demanded non-EA charitable organizations to relinquish money from the villain of the year, e.g. Jimmy Saville or the Sackler family, I have always thought that was silly, and I suspect other EAs tend to agree.