[T]hat’s not an interesting fact; we should be looking at how it changes and whether it’s socially maladaptive in important ways
It is interesting and relevant for the clause it was introducing—things that help the typical human are less likely to help you as you become less like the typical human. While, pessimistically, the degree of departure available in the positive direction may be sufficiently small that this is not relevant, it seems like a possibility worth being aware of as we strive use the resources available to become better.
things that help the typical human are less likely to help you as you become less like the typical human.
Yeah, that’s true, and I suppose I didn’t adequately address it in the grandparent. It doesn’t strike me as likely to cause serious problems, though, except perhaps in the social realm; the body of empirically supported self-improvement research that I’m currently aware of seems broad but shallow, not containing many long causal chains of the kind that could be disrupted by small changes in their early dependencies. It’s conceivable that completely eliminating a deeply rooted bias could render broad swathes of traditional self-improvement literature irrelevant at a stroke, but I’m not sure that’s plausible in the art’s current state.
In the future, perhaps, if LW or a similar community manages to come up with some seriously impressive results, but I don’t think that’s enough to merit a disclaimer.
It is interesting and relevant for the clause it was introducing—things that help the typical human are less likely to help you as you become less like the typical human. While, pessimistically, the degree of departure available in the positive direction may be sufficiently small that this is not relevant, it seems like a possibility worth being aware of as we strive use the resources available to become better.
Yeah, that’s true, and I suppose I didn’t adequately address it in the grandparent. It doesn’t strike me as likely to cause serious problems, though, except perhaps in the social realm; the body of empirically supported self-improvement research that I’m currently aware of seems broad but shallow, not containing many long causal chains of the kind that could be disrupted by small changes in their early dependencies. It’s conceivable that completely eliminating a deeply rooted bias could render broad swathes of traditional self-improvement literature irrelevant at a stroke, but I’m not sure that’s plausible in the art’s current state.
In the future, perhaps, if LW or a similar community manages to come up with some seriously impressive results, but I don’t think that’s enough to merit a disclaimer.