I disagree with the implied principle. E.g., I think it’s good for me to help animal welfare and global poverty EAs with their goals sometimes (when I’m in an unusually good position to help out), even though I think their time and money would be better spent on existential risk mitigation.
Agreed that a principle of ‘only cooperate on goals you agree with’ is too strong. On the object-level, if MacAskill was personally neutral or skeptical on the object-level question of whether SBF should buy Twitter, do you think he should have helped SBF out?
When is cooperation inappropriate? Maybe when the outcome you’re cooperating on is more consequential (in the bad direction, according to your own goals) than the expected gains from establishing reciprocity.
This would have been the largest purchase in EA history, replacing much or most of FTXFF with “SBF owns part of Twitter”. I think when the outcome is as consequential as that, we should hold cooperators responsible as if they were striving for the outcome, because the effects of helping SBF buy Twitter greatly outweigh the benefits from improving Will’s relationship with SBF (which I model as already very good).
if MacAskill was personally neutral or skeptical on the object-level question of whether SBF should buy Twitter, do you think he should have helped SBF out
If Will had no reason to think SBF was a bad egg, then I’d guess he should have helped out even if he thought the thing was not the optimal use of Sam’s money. (While also complaining that he thinks the investment is a bad idea.)
If Will thought SBF was a “bad egg”, then it could be more important to establish influence with him, because you don’t need to establish influence (as in ‘willingness to cooperate’) with someone who is entirely value-aligned with you.
I disagree with the implied principle. E.g., I think it’s good for me to help animal welfare and global poverty EAs with their goals sometimes (when I’m in an unusually good position to help out), even though I think their time and money would be better spent on existential risk mitigation.
Agreed that a principle of ‘only cooperate on goals you agree with’ is too strong. On the object-level, if MacAskill was personally neutral or skeptical on the object-level question of whether SBF should buy Twitter, do you think he should have helped SBF out?
When is cooperation inappropriate? Maybe when the outcome you’re cooperating on is more consequential (in the bad direction, according to your own goals) than the expected gains from establishing reciprocity.
This would have been the largest purchase in EA history, replacing much or most of FTXFF with “SBF owns part of Twitter”. I think when the outcome is as consequential as that, we should hold cooperators responsible as if they were striving for the outcome, because the effects of helping SBF buy Twitter greatly outweigh the benefits from improving Will’s relationship with SBF (which I model as already very good).
If Will had no reason to think SBF was a bad egg, then I’d guess he should have helped out even if he thought the thing was not the optimal use of Sam’s money. (While also complaining that he thinks the investment is a bad idea.)
If Will thought SBF was a “bad egg”, then it could be more important to establish influence with him, because you don’t need to establish influence (as in ‘willingness to cooperate’) with someone who is entirely value-aligned with you.