Anthropic unveils three metrics to track the pace of AI development inside frontier labs
rohanpaul_ai · x · 2026-09-18
Anthropic has published a new post laying out three measurement tools for the public to track AI progress inside frontier labs:
- Share of AI R&D performed by AI itself: how much of the next AI generation is built by AI rather than humans
- Oversight of agent actions: the lab's ability to oversee and intervene in actions AI agents take on its systems
- Compute allocation: the resources powering development of more capable models
The post also includes a first snapshot of these metrics from inside Anthropic. The company notes the numbers would shift if the frontier were paced in a coordinated way, as CEO Dario Amodei has called for. Anthropic also plans to embed independent third-party evaluators from multiple organizations with access to internal processes, systems and data comparable to internal risk teams, so they can verify safety practices, report incidents and monitor these key metrics.
Related event: Anthropic Discloses Internal Metrics: Claude Now Drives 26% of Its R&D(41 posts)→
More from AGI Musings
- Avoiding AI brain rot: Miles Cranmer's lecture on learning with AI — _ScottCondron · 2026-11-02
- Yacine mocks "autonomous research" hype as Naam argues RSI loop needs 5-10x boost to run away — yacineMTB · 2026-11-02
- Vitalik's "post-AI cryptography": pseudonyms are dead, per-action anonymity is in — devanshmehta · 2026-11-02
- Can law be formalized as deterministic rules? A French thread argues LLMs unlock the soft cases — IgorCarron · 2026-11-02
- Nobel economist William Nordhaus dies; essays revisit his growth, tech and climate work — Afinetheorem · 2026-11-02
- dhh: agents' great gift is routing around every gatekeeper in software — Stack Overflow proved it — IgorCarron · 2026-11-02