I don’t think the parody works in its current form. The climate scientist claims expertise on climate science x-risk through being a climate-science expert, not through being an expert on x-risk more generally. So him being wrong on other x-risks doesn’t update my assessment of his views on climate x-risk that much. In contrast, if the climate scientist’s organization built its headquarters in a flood plain and didn’t buy insurance, the resulting flood which destroyed the HQ would reduce my confidence in their ability to assess climate x-risk because they have shown themselves incompetent at least once in at assessing climate risks chose to them.
In contrast, EA (and the FF in particular) asserts/ed expertise in x-risk more generally. For someone claiming this kind of experience, the events that would cause me to downgrade are different than for a subject-matter expert. Missing an x-risk under one’s nose would count. While I don’t think “existential risk in one context equals existential risk in another context,” I don’t think the past performance has no bearing on estimates of future performance either.
I think assessing the extent to which the “miss” on FTX should cause a reasonable observer to downgrade EA’s x-risk credentials has been made difficult by the silence-on-advise-of-legal-counsel approach. To the extent that the possibility of FTX drying up wasn’t even on the radar of top leadership people, that would be a very serious downgrade for me. (Actually, it would be a significant downgrade in general confidence for any similarly-sized movement that lacked awareness that promised billions from a three-year old crypto company had a good chance of not materializing.) A failure to specifically recognize the risk of very shady business practices (even if not Madoff 2.0) would be a significant demerit in light of the well-known history of such things in the crypto space. To the extent that there was clear awareness and the probabilities were just wrong in hindsight, that is only a minor demerit for me.
To perhaps make it clearer: I think EA is trying to be expert in “existential risks to humanity”, and that really does have almost no overlap with “existential risks to individual firms or organizations”.
Or to sharpen the parody: if it was a climate-risk org that had got in trouble because it was funded by FTX, would that downgrade your expectation of their ability to assess climate risks?
But on mainstream EA assumptions about x-risk, the failure of the Future Fund materially increased existential risk to humanity. You’d need to find a similar event that materially changed the risk of catastrophic climate change for the analogy to potentially hold—the death of a single researcher or the loss of a non-critical funding source for climate-mitigation efforts doesn’t work for me.
More generally, I think it’s probably reasonable to downgrade for missing FTX on “general competence” and “ability to predict and manage risk” as well. I think both of those attributes are correlated with “ability to predict and manage existential risk,” the latter more so than the former. Given that existential-risk expertise is a difficult attribute to measure, it’s reasonable to downgrade when downgrading one’s assessment of more measureable attributes. Although that effect would also apply to the climate-mitigation movement if it suffered an FTX-level setback event involving insiders, the justification for listening to climate scientists isn’t nearly as heavily loaded on “ability to predict and manage existential risk.” It’s primarily loaded on domain-specific expertise in climate science, and missing FTX wouldn’t make me think materially less of the relevant people as scientists.
To be clear, I’m not endorsing the narrative that EA is near-useless on x-risk because it missed FTX. My own assumption is that people recognized a risk that FTX funding wouldn’t come through, and that the leaders recognized a risk that SBF was doing shady stuff (cf. the leaked leader chat) although perhaps not a Madoff 2.0. I think those risks were likely underestimated, which leads me to a downgrade but not a massive one.
I don’t think the parody works in its current form. The climate scientist claims expertise on climate science x-risk through being a climate-science expert, not through being an expert on x-risk more generally. So him being wrong on other x-risks doesn’t update my assessment of his views on climate x-risk that much. In contrast, if the climate scientist’s organization built its headquarters in a flood plain and didn’t buy insurance, the resulting flood which destroyed the HQ would reduce my confidence in their ability to assess climate x-risk because they have shown themselves incompetent at least once in at assessing climate risks chose to them.
In contrast, EA (and the FF in particular) asserts/ed expertise in x-risk more generally. For someone claiming this kind of experience, the events that would cause me to downgrade are different than for a subject-matter expert. Missing an x-risk under one’s nose would count. While I don’t think “existential risk in one context equals existential risk in another context,” I don’t think the past performance has no bearing on estimates of future performance either.
I think assessing the extent to which the “miss” on FTX should cause a reasonable observer to downgrade EA’s x-risk credentials has been made difficult by the silence-on-advise-of-legal-counsel approach. To the extent that the possibility of FTX drying up wasn’t even on the radar of top leadership people, that would be a very serious downgrade for me. (Actually, it would be a significant downgrade in general confidence for any similarly-sized movement that lacked awareness that promised billions from a three-year old crypto company had a good chance of not materializing.) A failure to specifically recognize the risk of very shady business practices (even if not Madoff 2.0) would be a significant demerit in light of the well-known history of such things in the crypto space. To the extent that there was clear awareness and the probabilities were just wrong in hindsight, that is only a minor demerit for me.
To perhaps make it clearer: I think EA is trying to be expert in “existential risks to humanity”, and that really does have almost no overlap with “existential risks to individual firms or organizations”.
Or to sharpen the parody: if it was a climate-risk org that had got in trouble because it was funded by FTX, would that downgrade your expectation of their ability to assess climate risks?
But on mainstream EA assumptions about x-risk, the failure of the Future Fund materially increased existential risk to humanity. You’d need to find a similar event that materially changed the risk of catastrophic climate change for the analogy to potentially hold—the death of a single researcher or the loss of a non-critical funding source for climate-mitigation efforts doesn’t work for me.
More generally, I think it’s probably reasonable to downgrade for missing FTX on “general competence” and “ability to predict and manage risk” as well. I think both of those attributes are correlated with “ability to predict and manage existential risk,” the latter more so than the former. Given that existential-risk expertise is a difficult attribute to measure, it’s reasonable to downgrade when downgrading one’s assessment of more measureable attributes. Although that effect would also apply to the climate-mitigation movement if it suffered an FTX-level setback event involving insiders, the justification for listening to climate scientists isn’t nearly as heavily loaded on “ability to predict and manage existential risk.” It’s primarily loaded on domain-specific expertise in climate science, and missing FTX wouldn’t make me think materially less of the relevant people as scientists.
To be clear, I’m not endorsing the narrative that EA is near-useless on x-risk because it missed FTX. My own assumption is that people recognized a risk that FTX funding wouldn’t come through, and that the leaders recognized a risk that SBF was doing shady stuff (cf. the leaked leader chat) although perhaps not a Madoff 2.0. I think those risks were likely underestimated, which leads me to a downgrade but not a massive one.