FrontierAugust 13, 2026via The Decoder

Top AI lab researchers warned about automated AI research, and several of their predicted milestones have already fallen

Why it matters

AI labs' own researchers forecast automated model development timelines; the prediction track record matters for capability forecasting and safety planning in the race.

Key signals

  • 25 researchers interviewed from OpenAI, Anthropic, Google DeepMind, Meta, and US universities
  • Topic: recursive self-improvement and automated AI research
  • Several predicted milestones have already been reached
  • Source: IAPS fellow Severin Field blog post
  • Published August 2026

The hook

25 frontier researchers predicted recursive self-improvement milestones. Several have already hit.

IAPS fellow Severin Field interviewed 25 researchers from OpenAI, Anthropic, Google Deepmind, Meta, and US universities about recursive self-improvement. In a new blog post, he takes stock. Several of the milestones those researchers named have already been hit.

The week's key stories, every Friday.

For practitioners and enthusiasts — free, in your inbox.

Free forever. No spam.