But arguments against unilateralism can’t be applied as arguments against investigation. Investigation grows the intellectual commons. Empirically it’s clear there is much to investigate. EAs generally agree that AI risk is the most important problem yet there is no plan to move forward (aside from help out OpenAI and hope that this somehow turns out to be a good idea instead of an apocalyptic one).
I like that you contrast deference with investigation, rather than unilaterism. So many discussions and posts about deference devolve into discussion about unilateralism. Example: https://forum.effectivealtruism.org/posts/Jx6ncakmergiC74kG/deference-culture-in-ea?commentId=epR5HxT6nkdSCtMCf
But arguments against unilateralism can’t be applied as arguments against investigation. Investigation grows the intellectual commons. Empirically it’s clear there is much to investigate. EAs generally agree that AI risk is the most important problem yet there is no plan to move forward (aside from help out OpenAI and hope that this somehow turns out to be a good idea instead of an apocalyptic one).