Blogger: OpenAI's coordinated AI safety talks with Anthropic and DeepMind mask a capability plateau

DavidLinthicum · x · 2026-09-18

David Linthicum argues that OpenAI's confirmation of weeks-long behind-the-scenes safety coordination with Anthropic and Google DeepMind shows the doom narrative was orchestrated PR, not independent conclusions. His take: frontier labs face drying training data, shrinking compute gains, and a flattening capability curve — and the safety framing deflects from slowing progress, ultimately protecting stock prices rather than users.

Original post →

More from AGI Musings

AGI Musings channel →