Some of us want unsafe ai so that we can train safety differently and only after the final iteration of between competing predictions.. A safe ai is no use in legal research, criminology, behavioral or social science. And I suspect the same in economics. I work extensively in…
Source date (UTC): 2023-09-30 04:08:56 UTC
Original post: https://twitter.com/i/web/status/1707970760084423032
Replying to: https://twitter.com/i/web/status/1707431004036649039
Leave a Reply