Eliezer—I strongly agree with this, and thanks for expressing these views in a clear and timely way.
I’ve seen some amount of EAs beating themselves up about ‘we should have known better than to trust SBF and FTX’.
Well, FWIW, I’ve been involved in crypto investing pretty intensively for a couple of years, and have followed crypto news fairly closely. Every crypto protocol, exchange, business, and ideas attracts its share of FUD (fear, uncertainty, & doubt—i.e. strategic slander from short-sellers and rival businesses). Lots of people were skeptical about Luna before it collapsed. Lots of people have been skeptical about the Tether stablecoin for years, and it hasn’t collapsed yet. People were skeptical about Solana being quite centralized and having VC backing—but I don’t recall much concern about Solana having SBF’s backing.
I honestly didn’t see big red flags around FTX being waved by crypto journalists or influencers. FTX seemed to attract a lot less FUD and skepticism than most other crypto businesses. (I could be wrong, and this is an empirical question—one would need to quantify relative FUD amounts across projects, maybe in proportion to their market cap or trading volume.)
I didn’t happen to use FTX myself, because I preferred using exchanges that were registered in the US for legal purposes, and that used ‘proof of reserve’ audits to verify user funds. But I could easily imagine a counterfactual world in which I transferred a lot of crypto to FTX—trusting partly in its EA bona fides—and lost everything.
Long story short, I agree with one of Eliezer’s key points: if the smartest VCs in crypto and Silicon Valley thought FTX was a great investment and an honorable company, and they have access to all the due diligence and internal data and audits that EA organizations didn’t, and if crypto journalists were NOT routinely warning that FTX was a shady company, then maybe we shouldn’t be beating ourselves up too hard about having been duped.
(Epistemic status: moderate at best; subject to updating given new information about a complex, ongoing crisis).
I’m mostly uninformed on this topic, but this all sounds very plausible to me — more plausible than very hindsight-bias-y accounts along the lines of “I didn’t guess this in advance, but of course it should have been obvious”.
I do want to note that “we should be wary of hindsight, and not assume this was obvious” is compatible with “we should give extra epistemic credit to people who did in fact correctly call this, especially insofar as they gave good supporting arguments at the time”.
Even if this was difficult to predict, we should expect there to have been some hints out there that people could pick up on, and to some degree we should update in favor of the heuristics that paid attention to those hints.
Eliezer—I strongly agree with this, and thanks for expressing these views in a clear and timely way.
I’ve seen some amount of EAs beating themselves up about ‘we should have known better than to trust SBF and FTX’.
Well, FWIW, I’ve been involved in crypto investing pretty intensively for a couple of years, and have followed crypto news fairly closely. Every crypto protocol, exchange, business, and ideas attracts its share of FUD (fear, uncertainty, & doubt—i.e. strategic slander from short-sellers and rival businesses). Lots of people were skeptical about Luna before it collapsed. Lots of people have been skeptical about the Tether stablecoin for years, and it hasn’t collapsed yet. People were skeptical about Solana being quite centralized and having VC backing—but I don’t recall much concern about Solana having SBF’s backing.
I honestly didn’t see big red flags around FTX being waved by crypto journalists or influencers. FTX seemed to attract a lot less FUD and skepticism than most other crypto businesses. (I could be wrong, and this is an empirical question—one would need to quantify relative FUD amounts across projects, maybe in proportion to their market cap or trading volume.)
I didn’t happen to use FTX myself, because I preferred using exchanges that were registered in the US for legal purposes, and that used ‘proof of reserve’ audits to verify user funds. But I could easily imagine a counterfactual world in which I transferred a lot of crypto to FTX—trusting partly in its EA bona fides—and lost everything.
Long story short, I agree with one of Eliezer’s key points: if the smartest VCs in crypto and Silicon Valley thought FTX was a great investment and an honorable company, and they have access to all the due diligence and internal data and audits that EA organizations didn’t, and if crypto journalists were NOT routinely warning that FTX was a shady company, then maybe we shouldn’t be beating ourselves up too hard about having been duped.
(Epistemic status: moderate at best; subject to updating given new information about a complex, ongoing crisis).
I’m mostly uninformed on this topic, but this all sounds very plausible to me — more plausible than very hindsight-bias-y accounts along the lines of “I didn’t guess this in advance, but of course it should have been obvious”.
I do want to note that “we should be wary of hindsight, and not assume this was obvious” is compatible with “we should give extra epistemic credit to people who did in fact correctly call this, especially insofar as they gave good supporting arguments at the time”.
Even if this was difficult to predict, we should expect there to have been some hints out there that people could pick up on, and to some degree we should update in favor of the heuristics that paid attention to those hints.