Oleg Alexandrov comments on Critiques of non-existent AI safety labs: Yours

[Oleg Alexandrov]

Fail early, fail often. Many little dooms are good. One big doom is not so good.