Bracketing doesnāt in general recommend focusing on the āfirst order consequencesā, in the sense people usually use that term (e.g. the first step in some coarse-grained causal pathway). There can be locations of value Iā where weād think A >_{Iā} B if we only considered first order consequences, yet A [incomparable]_{Iā} B all things considered. Conversely, there can be locations of value Iā that are only affected by higher-order consequences, yet A >_{Iā} B.
Not sure exactly what you mean by āgenerally do betterā, but just to be clear: Bracketing is its own theory of what it means to ādo betterā as an impartial altruist, not a formalization of a heuristic for getting higher EV. (Jesse says as much in the summary.)
On your second point: Even if AI would plausibly kill the bed net recipients soon, we also need to say whether (1) any concrete intervention weāre aware of would decrease AI risk in expectation, and (2) that intervention would be more cost-effective for bracketed-in welfare than the alternatives, if so.
Iām skeptical of (1), briefly, because whether an intervention prevents vs. causes an AI x-risk seems sensitive to various dynamics that we have little evidence about + are too unfamiliar for us to trust our intuitions about. More on this here.
Re: (2), if weāre bracketing out the far-future consequences, I expect itās hard to argue that AI risk work is more cost-effective than the best animal welfare opportunities. (Less confident in this point than the previous one, conditional on (1).)
Thanks Ben ā a few clarifications:
Bracketing doesnāt in general recommend focusing on the āfirst order consequencesā, in the sense people usually use that term (e.g. the first step in some coarse-grained causal pathway). There can be locations of value Iā where weād think A >_{Iā} B if we only considered first order consequences, yet A [incomparable]_{Iā} B all things considered. Conversely, there can be locations of value Iā that are only affected by higher-order consequences, yet A >_{Iā} B.
Not sure exactly what you mean by āgenerally do betterā, but just to be clear: Bracketing is its own theory of what it means to ādo betterā as an impartial altruist, not a formalization of a heuristic for getting higher EV. (Jesse says as much in the summary.)
On your second point: Even if AI would plausibly kill the bed net recipients soon, we also need to say whether (1) any concrete intervention weāre aware of would decrease AI risk in expectation, and (2) that intervention would be more cost-effective for bracketed-in welfare than the alternatives, if so.
Iām skeptical of (1), briefly, because whether an intervention prevents vs. causes an AI x-risk seems sensitive to various dynamics that we have little evidence about + are too unfamiliar for us to trust our intuitions about. More on this here.
Re: (2), if weāre bracketing out the far-future consequences, I expect itās hard to argue that AI risk work is more cost-effective than the best animal welfare opportunities. (Less confident in this point than the previous one, conditional on (1).)