I think trying to align future AI's with our interest is as futile as...

April 17, 2023 at 2:06 AM  ·  Comment — Luke Stokes's post  ·  View on Facebook  ·  Markdown source

I think trying to align future AI's with our interest is as futile as a chimp trying to align human interests with those of chimps. We're not even going to understand our future descendants motivations, let alone shape them. IMO, the best we can do is to direct near term AI's to make all intelligences (human and otherwise) robust to attack via uploading, decentralized backups, waldos, etc.