Vals AI Backs Dario's Slowdown Call, Says Its Benchmarks Now Appear in Every Major Lab's Model Cards
Miles_Brundage · x · 2026-09-13
Vals AI responded to Dario Amodei's call to "pace the frontier," saying it is encouraging to see industry leaders demand rigorous, independent evaluations as AI capabilities advance. Vals argues investment in AI systems has far outpaced our ability to test and understand them, making a market-based evaluation approach the best mechanism for public confidence.
Vals says it has built the infrastructure and novel methodologies independent evaluation requires, producing the first public RSI benchmark, an index of the economic impact of frontier models, and evaluations of model behavior across social safety nets, cyber risk, and mental health. Its benchmarks already appear in the model cards of every major foundation model lab, and it regularly briefs the U.S. government and national security bodies.
Context: Amodei published an essay arguing the AI industry should slow down, with a three-part plan. Anthropic unilaterally committed to the first step — giving third-party evaluators permanent, employee-level access to its systems to verify safety measures, report incidents, and assess alignment during training.
More from AGI Musings
- Guardian column: the way out of AI doom nihilism is human flourishing, not tech — nordicinst · 2026-09-14
- Zack Kanter: A 10% extinction risk beats an AI-fueled global unigovernment — beffjezos · 2026-09-14
- AI doomerism discourse deepens public fear and makes people less willing to try AI — Kyrannio · 2026-09-14
- VC krishnan wants AI to become the top midterm issue to wake society up to job disruption — krishnan · 2026-09-14
- Bryan Johnson: I don't trust anyone's AI risk assessment, so I'm betting on optimism — QuanquanGu · 2026-09-14
- What should AI labs do if AGI risk were real? A challenge to skeptics — AndyMasley · 2026-09-14