The OTHER AI Alignment Problem: Mesa-Optimizers and Inner Alignment
11:47
We Were Right! Real Inner Misalignment
11:32
There's No Rule That Says We'll Make It
12:23
El mayor defecto de la IA generativa - Computerphile
1:09:26
MIT Introduction to Deep Learning | 6.S191
30:38
Using Dangerous AI, But Safely?
17:52
Training AI Without Writing A Reward Function, with Reward Modelling
10:36
Why Would AI Want to do Bad Things? Instrumental Convergence
25:07