How to Build Safe AI (Without Making the AI Safe)

3 pointsposted 10 hours ago
by stefie10

1 Comments

ibrahimuyaradua

10 hours ago

Interesting. Even tho I think there is another path that exists but it’s being ignored. Accuracy cannot be expected to result from generative and probabilistic machines especially when they programmed to retain attention instead of accuracy.