Development

A lab touts an 80% price cut, a billion users, and agents writing 99.8% of tokens

The recursive loop restated as corporate strategy: a lab's finance chief prices abundance off gains the models produced for themselves, in a plan written from behind in the race.

development2026-08-01rsiai-r-and-d

Claim

OpenAI's finance chief laid out an abundant intelligence strategy touting an 80% price cut, a billion users and agents writing 99.8% of output tokens, after a postmortem traced how the company lost its crown to Anthropic's coding agent and near-trillion valuation.

Reported metric: 99.8% of tokens

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 chief financial officer Sarah Friar laid out an “abundant intelligence” strategy (OpenAI) built on three numbers: the 80% GPT-5.6 price cut of two days earlier, a billion users, and agents writing 99.8% of output tokens, the figure first reported in June when Codex was found generating almost all of its own weekly output. It landed alongside a Wall Street Journal postmortem (WSJ) on how the company lost its lead to Anthropic’s Claude Code and near-trillion valuation, an account OpenAI answers with Sol and a Codex super app. In the trajectory it is the moment the loop’s outputs, self-written kernels and self-generated tokens, become the headline of a lab’s business plan rather than a research aside, following the February reports that Codex builds itself and that all of Anthropic’s product code is written by Claude, and a day before an Anthropic researcher argued that labs will compete with their own customers.

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.