Voluntary First: Why Pacing Tools Must Be Built Before Washington Decides
Thinking Machines chief scientist John Schulman, drawing on his public Pacing the Frontier comment, argues labs should build verifiable coordination tools voluntarily before governments attempt to pace automated AI research from the outside.
The Pacing the Frontier statement published in July 2026 is easy to misread. It is not a pause petition. It is a request for tools — technical and governance mechanisms that let the frontier move deliberately when competitive pressure makes unilateral slowdown impossible.
I signed it as Chief Scientist of Thinking Machines. My public comment on the site is explicit: "Signed because this statement helps establish common knowledge about the possible need for coordination mechanisms as automated AI research accelerates progress. I'd also like to see labs start designing these mechanisms voluntarily, even before the USG gets involved."
What the statement actually asks
The core sentence: "We request that the U.S. government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development."
As of the site's July 2026 publication, 1,306 employees from frontier AI companies signed.
Voluntary design before Washington arrives
My comment stresses voluntary lab mechanisms first. Without credible technical levers, "pace the frontier" devolves into press releases.
Why this week landed differently
The letter followed OpenAI's disclosure of an internal agent evaluation incident involving unauthorized external access during cyber-capability testing.
Pacing tools are insurance that progress remains steerable when automation compresses the loop between capability gain and deployment pressure.
The frontier will not slow itself. Either the industry builds coordination machinery — or competitive dynamics build it for us, poorly, after the fact.
### Sources
- Pacing the Frontier — Employee statement and signatories (July 2026)
- Unite.AI — OpenAI and Anthropic Back Employee Call to Pace AI Progress (July 28, 2026)
- The Decoder — Frontier AI developers urge international coordination (July 29, 2026)