Comment by roywiggins

2 years ago

Fortunately, so far we don't seem to know how to make an AI at all. Unfortunately we also don't know how to define "safe" either.

0 comments