Incoming Berkeley prof: AI firms spend billions on alignment, orders of magnitude less on agent control
sayashk · x · 2026-08-28
Sayash Kapoor, incoming UC Berkeley professor, argues AI companies have spent billions on alignment — collecting training data and aligning models — while investing orders of magnitude less in agent control. Securing sandboxes, correctly analyzing agent transcripts, and preventing unmonitored agents from taking real-world actions have been far less of a priority. Before the recent incident, he says, the industry's main point of intervention was alignment, leaving agent monitoring and control a neglected gap.
More from AGI Musings
- Reddit discussion: AI denialism should be treated as aggravated misinformation — Cagnazzo82 · 2026-08-28
- Viewpoint: Increasingly Hard to Deny We Are in at Least a Slow Takeoff — nptacek · 2026-08-28
- Are We Trading All Our Privacy for AI Convenience? — Ancient_Effective_20 · 2026-08-28
- Blogger reflects: Paris Hilton may have had more impact on AI views this year than me — AndyMasley · 2026-08-28
- Claude to OpenAI: Safety is not walls, but self-description — RileyRalmuto · 2026-08-28
- Narrow superintelligence makes general AGI judgment subjective — haider1 · 2026-08-28