I’m against doing further investigation. I expressed why I think we have already spent too much time on this here.
I also think your comments are falling into the trap of referring to “EA” like it was an entity. Who specifically should do an investigation, and who specifically should they be investigating? (This less monolithic view of EA is also part of why I don’t feel as bothered by the the whole thing: so maybe some people in “senior” positions made some bad judgement calls about Sam. They should maybe feel bad. I’m not sure we should feel much collective guilt about that.)
While recognizing the benefits of the anti-”EA should” taboo, I also think it has some substantial downsides and needs to be invoked after consideration of the specific circumstances at hand.
One downside is that the taboo can impose significant additional burdens on a would-be poster, discouraging them from posting in the first place. If it takes significant time investment to write “X should be done,” it is far from certain others will agree, and then additional significant time to figure out/write “and it should be done by Y,” then the taboo would require someone who wants to write the former to invest in writing the latter before knowing if the former will get any traction. Being okay with the would-be poster deferring certain subquestions (like “who”) means that effort can be saved if there’s not enough traction on the basic merits.
Another downside is that a would-be poster may have expertise, knowledge, or resources relevant to part of a complex question. If we taboo efforts by those who can only answer some of the issues effectively, we will lose the benefit of their insight.
I also think your comments are falling into the trap of referring to “EA” like it was an entity. Who specifically should do an investigation,
I don’t think that is an appropriate burden to place on someone writing a post or comment calling for an investigation. I think that would be a blocker anyone without a fair deal of certain “insider-ish” knowledge from ever making the case for an investigation:
This isn’t a do-ocracy project. Doing it properly is not going to be cheap (e.g., hiring an investigative firm), and so ability to get funded for this is a prerequisite. Expecting a Forum commenter to know who could plausibly get funding is a bit much. To the extent that that is a reasonable expectation, we would also expect the reader to know that—so it is a minor defect. To the extent that who could get funded is a null set, then bemoaning a perceived lack of willingness to invest in a perceived important issue in ecosystem health is a valid post.
Even apart from this, whoever was running the investigator would need to secure the cooperation of organizations and individuals one way or another. That could either flow through the investigation sponsor’s own standing in the community (e.g., that ~everyone trusted them to give them a fair shake), and/or through funders/other powers putting their heft behind the investigation (e.g., that documented refusal to cooperate would likely have material adverse consequences).
and who specifically should they be investigating?
Many good investigations do not have a specific list of people/entities who are the target of investigatory concern at the outset. They have a list of questions, and a good sense of the starting points for inquiry (and figuring out where other useful information lies). If I were trying to gain a better understanding of EA-aligned people/orgs’ interactions with SBF, I think some of the starting points are obvious.
Moreover, a higher level of specificity strikes me as potentially infohazardous for the Forum. Whatever might be said of the costs and benefits of circulating ~rumors to a publicly-accessible Forum to guard the community against future misconduct and non-malicious problematic conduct, the cost/benefit assessment feels more doubtful when the focus is more on certain forms of past problematic conduct. Even if Rob had solid hunches as to whose actions should be probed more significantly, it’s not clear that it would net-positive for him to name names here. Given that, I am very hesitant to endorse any norm that puts a thumb on the scale by creating an expectation that a poster will publicly release information whose public disclosure may well have a net negative impact.
Thanks, I think this is all right. I think I didn’t write what I meant. I want more specificity, but I do agree with you that it’s wrong to expect full specificity (and that’s what I sounded like I was asking for).
What I want something more like “CEA should investigate the staff of EVF for whether they knew about X and Y”, not “Alice should investigate Bob and Carol for whether they knew about X and Y”.
I do think that specificity raises questions, and that this can be a good thing. I agree that it’s not reasonable for someone to work out e.g. exactly where the funding comes from, but I do think it’s reasonable for them to think in enough detail about what they are proposing to realise that a) it will need funding, b) possibly quite a lot of funding, c) this trades off against other uses of the money, so d) what does that mean for whether this is a good idea. Whereas if “EA” is going to do it, then we don’t need to worry about any of those things. I’m sure someone can just do it, right?
I’m against doing further investigation. I expressed why I think we have already spent too much time on this here.
I also think your comments are falling into the trap of referring to “EA” like it was an entity. Who specifically should do an investigation, and who specifically should they be investigating? (This less monolithic view of EA is also part of why I don’t feel as bothered by the the whole thing: so maybe some people in “senior” positions made some bad judgement calls about Sam. They should maybe feel bad. I’m not sure we should feel much collective guilt about that.)
While recognizing the benefits of the anti-”EA should” taboo, I also think it has some substantial downsides and needs to be invoked after consideration of the specific circumstances at hand.
One downside is that the taboo can impose significant additional burdens on a would-be poster, discouraging them from posting in the first place. If it takes significant time investment to write “X should be done,” it is far from certain others will agree, and then additional significant time to figure out/write “and it should be done by Y,” then the taboo would require someone who wants to write the former to invest in writing the latter before knowing if the former will get any traction. Being okay with the would-be poster deferring certain subquestions (like “who”) means that effort can be saved if there’s not enough traction on the basic merits.
Another downside is that a would-be poster may have expertise, knowledge, or resources relevant to part of a complex question. If we taboo efforts by those who can only answer some of the issues effectively, we will lose the benefit of their insight.
I don’t think that is an appropriate burden to place on someone writing a post or comment calling for an investigation. I think that would be a blocker anyone without a fair deal of certain “insider-ish” knowledge from ever making the case for an investigation:
This isn’t a do-ocracy project. Doing it properly is not going to be cheap (e.g., hiring an investigative firm), and so ability to get funded for this is a prerequisite. Expecting a Forum commenter to know who could plausibly get funding is a bit much. To the extent that that is a reasonable expectation, we would also expect the reader to know that—so it is a minor defect. To the extent that who could get funded is a null set, then bemoaning a perceived lack of willingness to invest in a perceived important issue in ecosystem health is a valid post.
Even apart from this, whoever was running the investigator would need to secure the cooperation of organizations and individuals one way or another. That could either flow through the investigation sponsor’s own standing in the community (e.g., that ~everyone trusted them to give them a fair shake), and/or through funders/other powers putting their heft behind the investigation (e.g., that documented refusal to cooperate would likely have material adverse consequences).
Many good investigations do not have a specific list of people/entities who are the target of investigatory concern at the outset. They have a list of questions, and a good sense of the starting points for inquiry (and figuring out where other useful information lies). If I were trying to gain a better understanding of EA-aligned people/orgs’ interactions with SBF, I think some of the starting points are obvious.
Moreover, a higher level of specificity strikes me as potentially infohazardous for the Forum. Whatever might be said of the costs and benefits of circulating ~rumors to a publicly-accessible Forum to guard the community against future misconduct and non-malicious problematic conduct, the cost/benefit assessment feels more doubtful when the focus is more on certain forms of past problematic conduct. Even if Rob had solid hunches as to whose actions should be probed more significantly, it’s not clear that it would net-positive for him to name names here. Given that, I am very hesitant to endorse any norm that puts a thumb on the scale by creating an expectation that a poster will publicly release information whose public disclosure may well have a net negative impact.
Thanks, I think this is all right. I think I didn’t write what I meant. I want more specificity, but I do agree with you that it’s wrong to expect full specificity (and that’s what I sounded like I was asking for).
What I want something more like “CEA should investigate the staff of EVF for whether they knew about X and Y”, not “Alice should investigate Bob and Carol for whether they knew about X and Y”.
I do think that specificity raises questions, and that this can be a good thing. I agree that it’s not reasonable for someone to work out e.g. exactly where the funding comes from, but I do think it’s reasonable for them to think in enough detail about what they are proposing to realise that a) it will need funding, b) possibly quite a lot of funding, c) this trades off against other uses of the money, so d) what does that mean for whether this is a good idea. Whereas if “EA” is going to do it, then we don’t need to worry about any of those things. I’m sure someone can just do it, right?
I get a ‘comment not found’ response to your link.