If I’m understanding your question correctly, that part of my expectation is almost entirely conditional on being in a post-ASI world. Before then, if interest in (effectively) reducing animal suffering stays roughly the size of “EA”, then I don’t particularly expect it to be become cheap enough to subsize people farming animals to raise them in humane conditions. (This expectation becomes weaker with longer AI timelines, but I haven’t though that hard about what the world looks like in 20+ years without strong AI, and how that affects the marginal cost of various farmed animal welfare interventions.)
So my timelines on that are pretty much just my AI timelines, conditioned on “we don’t all die” (which are shifted a bit longer than my overall AI timelines, but not by that much).
If I’m understanding your question correctly, that part of my expectation is almost entirely conditional on being in a post-ASI world. Before then, if interest in (effectively) reducing animal suffering stays roughly the size of “EA”, then I don’t particularly expect it to be become cheap enough to subsize people farming animals to raise them in humane conditions. (This expectation becomes weaker with longer AI timelines, but I haven’t though that hard about what the world looks like in 20+ years without strong AI, and how that affects the marginal cost of various farmed animal welfare interventions.)
So my timelines on that are pretty much just my AI timelines, conditioned on “we don’t all die” (which are shifted a bit longer than my overall AI timelines, but not by that much).