AI Safety Is Falling Behind Capabilities | Boaz Barak

Boaz Barak argues AI safety and alignment are improving but not fast enough to keep pace with rapidly growing model capabilities.

• Safety lags capabilities: Barak says alignment evals keep getting saturated, but since model capabilities keep growing faster, safety is falling behind — he favors pacing AI development to let safety catch up.
• Rejects 'gradual disempowerment': Barak worries AI meeting all material needs could remove vocational pressure to learn, causing people to voluntarily cede control — even absent classic doom scenarios.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.