Robert Wiblin: Anthropic hasn't shown it's on track to actually understand AI internals
GaryMarcus · x · 2026-11-20
Robert Wiblin (amplified by Gary Marcus) argues that Anthropic is failing to persuade skeptics of any of its key interpretability claims—that the company is actually on track to meaningfully understand what's happening inside frontier models. A pointed critique of the gap between Anthropic's safety narrative and demonstrable research progress.
More from AGI Musings
- Ai2 launches open-source Theorizer in Asta: AI proposes testable scientific theories — allen_ai · 2026-11-20
- repligate: top models have surpassed philosophers, mind-braiding is next — repligate · 2026-11-20
- UI isn't dead: chat alone is far from enough for human-agent communication — HaijunXia · 2026-11-20
- Dario Amodei: we talk about AI risk because it's preventable — bookwormengr · 2026-11-20
- Beff Jezos hopes Anthropic's billions push it to drop EA and embrace e/acc — beffjezos · 2026-11-20
- Superintelligence says build more housing; economists still cap AI growth at 10%/year — deanwball · 2026-11-20