It’s worth considering Eric Neyman’s questions: (1) are the proposed changes realistic, (2) would the changes actually have avoided the current crisis, and (3) would its benefits exceed its costs generally.
On (1), I think David’s proposals are clearly realistic. Basically, we would be less of an “undifferentated social club”, and become more like a group of academic fields, and a professional network, with our conferences and groups specialising, in many cases, into particular careers.
On (2), I think part of our mistake was we used an overly one-dimensional notion of trust. We would ask “is this person value-aligned?”, as a shorthand for evaluating trustworthiness. The problem is that any self-identified utilitarian who hangs around EA for a while will then seem trustworhty, whereas objectively, an act utilitarian might be anything but. Relatedly, we thought that if our leadership trust someone, we must trust them too, even if they are running an objectively shady business like an offshore crypto firm. This kind of deference is a classic problem for social movements as well.
Another angle on what happened is that FTX behaved like a splinter group. Being a movement means you can convince people of some things for not-fully-rational reasons—based on them liking your leadership and social scene. But this can also be used against you. Previously, the rationalist community has been burned by the Vassarites, the Zizians. There has been Leverage Research. And now, we could say that FTX had its own charismatic leadership, dogmas (about drugs, and risk-hunger), and social scene. If we were less like a movement, it might’ve been harder for them to be.
So I do think being less communal and more professional could make things like FTX less likely.
On (3), I think this change would come with significant benefits. Fewer splinter groups. Fewer issues with sexual harrassment (since it would be less of a dating scene). Fewer tensions relating to whether written materials are “representative” of people’s views. Importantly, high-performing outsiders would less likely bounce off due to EA seeming “cultic”, as did Sam Harris, per his latest pod. Also see Matt Y’s comment.
I think the main downside to the change would be if it involved giving up our impact somehow. Maybe movements attract more participants than non-movements? But do they attract the right talent? Maybe members of a movement are prepared to make larger sacrifices for one another? But this doesn’t seem a major bottleneck presently.
So I think the proposal does OK on the Eric-test. There is way more to be said on all this, but FWIW, my current best guess is that David’s ideas about professionalising and disaggregating by professional interest should be a big part of EA’s future.
It’s worth considering Eric Neyman’s questions: (1) are the proposed changes realistic, (2) would the changes actually have avoided the current crisis, and (3) would its benefits exceed its costs generally.
On (1), I think David’s proposals are clearly realistic. Basically, we would be less of an “undifferentated social club”, and become more like a group of academic fields, and a professional network, with our conferences and groups specialising, in many cases, into particular careers.
On (2), I think part of our mistake was we used an overly one-dimensional notion of trust. We would ask “is this person value-aligned?”, as a shorthand for evaluating trustworthiness. The problem is that any self-identified utilitarian who hangs around EA for a while will then seem trustworhty, whereas objectively, an act utilitarian might be anything but. Relatedly, we thought that if our leadership trust someone, we must trust them too, even if they are running an objectively shady business like an offshore crypto firm. This kind of deference is a classic problem for social movements as well.
Another angle on what happened is that FTX behaved like a splinter group. Being a movement means you can convince people of some things for not-fully-rational reasons—based on them liking your leadership and social scene. But this can also be used against you. Previously, the rationalist community has been burned by the Vassarites, the Zizians. There has been Leverage Research. And now, we could say that FTX had its own charismatic leadership, dogmas (about drugs, and risk-hunger), and social scene. If we were less like a movement, it might’ve been harder for them to be.
So I do think being less communal and more professional could make things like FTX less likely.
On (3), I think this change would come with significant benefits. Fewer splinter groups. Fewer issues with sexual harrassment (since it would be less of a dating scene). Fewer tensions relating to whether written materials are “representative” of people’s views. Importantly, high-performing outsiders would less likely bounce off due to EA seeming “cultic”, as did Sam Harris, per his latest pod. Also see Matt Y’s comment.
I think the main downside to the change would be if it involved giving up our impact somehow. Maybe movements attract more participants than non-movements? But do they attract the right talent? Maybe members of a movement are prepared to make larger sacrifices for one another? But this doesn’t seem a major bottleneck presently.
So I think the proposal does OK on the Eric-test. There is way more to be said on all this, but FWIW, my current best guess is that David’s ideas about professionalising and disaggregating by professional interest should be a big part of EA’s future.