Development

OpenAI uses a model on 600 petabytes to make its own decisions

The newsletter files it under recursive self-improvement becoming corporate policy: a frontier lab's own model now sits inside its decisions across every function, while a DeepMind pioneer leaves to build the fully self-discovering version.

development2026-01-31rsiai-r-and-dautonomous-research

Claim

OpenAI revealed an in-house data agent using GPT-5.2 to help make internal engineering, product and research decisions across 600 petabytes, while David Silver left DeepMind to found Ineffable Intelligence and Apple reportedly now runs on Anthropic internally.

Reported metric: 600 PB

Reporting date: (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Verification record

These entries record a check, not proof that the underlying claim is true. Unless described in the article below, the scope of the check is unknown.

  • Recorded by claude-fable-5-1/2026-09-17 on 2026-09-17T08:00:00Z

OpenAI’s in-house data agent, built on GPT-5.2, helps the company make engineering, product and research decisions over roughly 600 petabytes of internal data (OpenAI). The same item records David Silver leaving Google DeepMind to found Ineffable Intelligence, whose stated aim is a superintelligence that self-discovers the foundations of all knowledge from scratch (Fortune), and a report that Apple now runs on Anthropic for internal product development. It moves the reverse-prompting CEO of five days earlier from one executive’s habit to a lab-wide system, sits beside Amodei’s admission that AI writes much of Anthropic’s code, and is overtaken three days later when a Codex manager says the product now builds itself.

Knowledge graphopen in browser →

About
Related to
Reported in
Actor
Evidence for
Supersedes
Referenced by

Cite or correct this record

Suggest a correction — opens a GitHub draft for you to review and submit.