It seems significant if only $3B of customers were earning interest on cash or assets, and the other customers had the option to opt in to lending but explicitly did not. I think all the other customers would have an extremely reasonable expectation of their assets just sitting there. I’m not super convinced by the close reading of the terms of service but it seems like the common-sense case is strong.
I’m interested in understanding that $3B number and any relevant subtleties in the accounting. I feel like if that number had been $15B then this would plausibly just be a failure of risk management, in which case I guess that number is central to the clear-cut fraud vs willful negligence question. The $3B estimate seems plausible but all I’ve seen is an out of context screenshot which is not great.
I am piecing some of this together from this alleged report of someone who worked at in the two days before the end, from which I inferred that Caroline and Sam knew that they took an unprecedented step when they loaned that customer money that nobody knew about, which seems like it just wouldn’t really be the case if they had only used money from the people who had lending enabled: https://twitter.com/libevm/status/1592383746551910400
I am definitely very interested in a better estimate of how many customers had lending enabled (but even separately from that, given various other aspects of FTX’s finances, it seems very likely that the people who did deposit their money as non-lending in FTX will not get their money back, which seems like it would have required some kind of transfer of ownership from clients to FTX at some point, and therefore violating the ToS).
Many people have to make plans about the future of the EA community, as well as the future of their own organizations who might have received FTX funding.
Things like this seem to have a decent bearing on the total amount of clawbacks as well as the actual ethical lines that were crossed in the process of FTX, which seem highly relevant to me for learning from this mess.
I think a world where you loaned out a bunch of customer deposits who had explicitly said “yes, you can lend out my customer deposits” is very different than a world where you stole a bunch of people’s money. Both are bad, but I think they imply pretty different levels of wrongdoing.
I think given the speed of lawsuits in cases like this we might not have much clearer information for many years, and at least I have many decisions to make that cannot wait a year that relate to things like this.
(Just to be concrete, at least my personal continued involvement in the EA community is pretty contingent on the details of what happened here, e.g. the degree to which the EA community was partially responsible for this mess, the degree to which it seems likely to learn from its mistakes, the ways the rest of the world will relate to the EA community after this, so in some sense most of my career feels currently on hold, so finding out more stuff earlier has very high information value for me)
I’m really surprised to read this, thank you for sharing!
the actual ethical lines that were crossed in the process of FTX, which seem highly relevant to me for learning from this mess.
I think that whether fraud was committed or not, it seems to me now even clearer than before that we should never allow it, and we should make it even more of a priority to speak out against evil. Much like whether COVID actually came from a lab or not wouldn’t change too much our attitude towards biosecurity, given that a significant probability that it came from a lab already justifies being very careful about that risk in the future.
my personal continued involvement in the EA community is pretty contingent on the details of what happened here, e.g. the degree to which the EA community was partially responsible for this mess
The EA community is like 20k people, I find it surprising to judge whether e.g. ARC (I assume you’re more on the AI Safety side) would be more or less worthy of support based on whether <~20 people stole money. Even “EA leadership” is a lot of people, and before this scandal I don’t think many were including SBF (but indeed he was wrongly seen as a role model).
If it turned out that SBF indeed stole money, in what different ways would you try to do the most good?
I think that whether fraud was committed or not, it seems to me now even clearer than before that we should never allow it, and we should make it even more of a priority to speak out against evil.
Just to be clear, I think “never commit fraud” is not a good ethical guideline to take away from this (both in that it isn’t learning enough from this situation, and in the sense that there are a lot of situations where you want to do fraud-adjacent things that are actually the ethical thing to do), as I’ve tried to argue in various other places on the forum. I think I would be quite sad if that is the primary lesson we take away from this.
I do think there is something important in the “speak out against evil” direction, and that’s the direction I am most interested in exploring.
The EA community is like 20k people, I find it surprising to judge whether e.g. ARC (I assume you’re more on the AI Safety side) would be more or less worthy of support based on whether <~20 people stole money.
I think the situation with OpenAI is quite analogous to the situation with FTX in terms of its harms for the world and the EA community’s involvement, and sadly I do think Paul has contributed substantially to the role that OpenAI has played in the EA ecosystem, so that’s a concrete way in which I think the lessons we learn here have a direct relevance to how I relate to ARC. I also think I feel quite similar about the Anthropic situation.
My support for EA is not conditional on nobody in EA being blameworthy. I am part of EA in order to improve the world. If EA makes the world worse, I don’t want to invest in it, independently of whether any specific individual can clearly be blamed for anything bad. In as much as we give rise to institutions like FTX and OpenAI, it really seems like we should change how we operate, or cease existing, and I do think the whole EA thing seemed quite load-bearing for both OpenAI and FTX coming into existence.
and before this scandal I don’t think many were including SBF (but indeed he was wrongly seen as a role model).
I think it would have been quite weird to not include SBF in “EA Leadership” last year. It was pretty clear he was doing a lot of leading, and he was invited to all the relevant events I can think of.
It seems significant if only $3B of customers were earning interest on cash or assets, and the other customers had the option to opt in to lending but explicitly did not. I think all the other customers would have an extremely reasonable expectation of their assets just sitting there. I’m not super convinced by the close reading of the terms of service but it seems like the common-sense case is strong.
I’m interested in understanding that $3B number and any relevant subtleties in the accounting. I feel like if that number had been $15B then this would plausibly just be a failure of risk management, in which case I guess that number is central to the clear-cut fraud vs willful negligence question. The $3B estimate seems plausible but all I’ve seen is an out of context screenshot which is not great.
Yeah, I also don’t have anything else better.
I am piecing some of this together from this alleged report of someone who worked at in the two days before the end, from which I inferred that Caroline and Sam knew that they took an unprecedented step when they loaned that customer money that nobody knew about, which seems like it just wouldn’t really be the case if they had only used money from the people who had lending enabled: https://twitter.com/libevm/status/1592383746551910400
I am definitely very interested in a better estimate of how many customers had lending enabled (but even separately from that, given various other aspects of FTX’s finances, it seems very likely that the people who did deposit their money as non-lending in FTX will not get their money back, which seems like it would have required some kind of transfer of ownership from clients to FTX at some point, and therefore violating the ToS).
I am curious as to how this would be decision-relevant at this moment.
It seems to me that there’s a lot of information that will surface in the next months and years (random example: context behind the FTX US President stepping down the day before a suspicious (in hindsight) transaction).
To me, the best thing to do seems to be to just wait for the judicial proceedings and for information to surface.
Many people have to make plans about the future of the EA community, as well as the future of their own organizations who might have received FTX funding.
Things like this seem to have a decent bearing on the total amount of clawbacks as well as the actual ethical lines that were crossed in the process of FTX, which seem highly relevant to me for learning from this mess.
I think a world where you loaned out a bunch of customer deposits who had explicitly said “yes, you can lend out my customer deposits” is very different than a world where you stole a bunch of people’s money. Both are bad, but I think they imply pretty different levels of wrongdoing.
I think given the speed of lawsuits in cases like this we might not have much clearer information for many years, and at least I have many decisions to make that cannot wait a year that relate to things like this.
(Just to be concrete, at least my personal continued involvement in the EA community is pretty contingent on the details of what happened here, e.g. the degree to which the EA community was partially responsible for this mess, the degree to which it seems likely to learn from its mistakes, the ways the rest of the world will relate to the EA community after this, so in some sense most of my career feels currently on hold, so finding out more stuff earlier has very high information value for me)
I’m really surprised to read this, thank you for sharing!
I think that whether fraud was committed or not, it seems to me now even clearer than before that we should never allow it, and we should make it even more of a priority to speak out against evil.
Much like whether COVID actually came from a lab or not wouldn’t change too much our attitude towards biosecurity, given that a significant probability that it came from a lab already justifies being very careful about that risk in the future.
The EA community is like 20k people, I find it surprising to judge whether e.g. ARC (I assume you’re more on the AI Safety side) would be more or less worthy of support based on whether <~20 people stole money. Even “EA leadership” is a lot of people, and before this scandal I don’t think many were including SBF (but indeed he was wrongly seen as a role model).
If it turned out that SBF indeed stole money, in what different ways would you try to do the most good?
Just to be clear, I think “never commit fraud” is not a good ethical guideline to take away from this (both in that it isn’t learning enough from this situation, and in the sense that there are a lot of situations where you want to do fraud-adjacent things that are actually the ethical thing to do), as I’ve tried to argue in various other places on the forum. I think I would be quite sad if that is the primary lesson we take away from this.
I do think there is something important in the “speak out against evil” direction, and that’s the direction I am most interested in exploring.
I think the situation with OpenAI is quite analogous to the situation with FTX in terms of its harms for the world and the EA community’s involvement, and sadly I do think Paul has contributed substantially to the role that OpenAI has played in the EA ecosystem, so that’s a concrete way in which I think the lessons we learn here have a direct relevance to how I relate to ARC. I also think I feel quite similar about the Anthropic situation.
My support for EA is not conditional on nobody in EA being blameworthy. I am part of EA in order to improve the world. If EA makes the world worse, I don’t want to invest in it, independently of whether any specific individual can clearly be blamed for anything bad. In as much as we give rise to institutions like FTX and OpenAI, it really seems like we should change how we operate, or cease existing, and I do think the whole EA thing seemed quite load-bearing for both OpenAI and FTX coming into existence.
I think it would have been quite weird to not include SBF in “EA Leadership” last year. It was pretty clear he was doing a lot of leading, and he was invited to all the relevant events I can think of.
None of this is decision-relevant for me and waiting a few months or years makes complete sense, but alas “interesting” != “decision-relevant.”