Non-alignment project ideas for making transformative AI go well

post by Lukas Finnveden (Lanrian) · 2024-01-04T07:23:13.658Z · LW · GW · 1 comments

This is a link post for https://lukasfinnveden.substack.com/p/projects-for-making-transformative

1 comments

Comments sorted by top scores.

comment by RogerDearnaley (roger-d-1) · 2024-01-04T09:37:27.626Z · LW(p) · GW(p)

It’s plausible that there will soon be digital minds that are sentient and deserving of rights. This raises several important issues that we don’t know how to deal with.

See the first five posts of my sequence AI, Alignment, and Ethics [? · GW] for a detailed analysis of exactly this, including:

  • how to make decisions like this
  • under what circumstance we should, and should not, give rights to different sorts of digital minds, how much moral weight to give them, and how this interacts with things like their ease of duplication
  • which rights to give them and why
  • whether the criterion for moral rights should be sapience, or sentience

(Some of the results were quite counterintuitive to me when I first showed them.)