In short

Cambridge's Seán Ó hÉigeartaigh argues the Western narrative of a US–China 'race' for AGI is significantly exaggerated. China frames AI as a practical economic 'tool', its 'AI Plus' strategy emphasises broad integration, not superintelligence, with no evidence of a centralised 'Manhattan Project'. He contends the race narrative serves tech lobbyists, investors and hawks seeking deregulation, and urges track-two dialogues, verification mechanisms and hardware-level governance to defuse genuine race dynamics.

“China is racing us to AGI” has become one of the most powerful sentences in AI politics, used to justify speed, secrecy, and deregulation. This episode asks whether it’s actually true.

Seán Ó hÉigeartaigh has spent a decade watching China’s AI ecosystem up close, and his answer is: much less than you’d think. China’s official strategy (“AI Plus”) treats AI as a practical tool for broad economic integration, not a Manhattan Project for superintelligence. The race framing, he argues, tells us more about who benefits from it in the West than about Beijing’s intentions.

Crucially, this isn’t complacency. If a genuine race did emerge, it would be dangerous, which is why he spends the back half of the conversation on the concrete machinery that could defuse it: track-two dialogues, verification, and governance built into the hardware itself.

Key takeaways

  • Little evidence supports a top-down Chinese national push toward AGI; China frames AI as a practical 'tool' via its 'AI Plus' strategy.
  • The winner-takes-all AGI race narrative serves vested interests, tech lobbyists, VCs and hawks seeking deregulation.
  • US–China decoupling since 2019 widens information gaps and fuels misconceptions on both sides.
  • China is not monolithic; internal departmental rivalry resembles Western bureaucratic dynamics.
  • Track-two dialogues, verification mechanisms and hardware-level governance (e.g. FlexHEG) could reduce real race risks.

Listen to the full episode

This is a summary. Listen to the full conversation on your platform of choice, or read the companion essay on Substack.

Read on Substack → Watch on YouTube Listen on Spotify

Frequently asked questions

Is China really racing the US for AGI?
According to Seán Ó hÉigeartaigh, the race narrative is significantly exaggerated, China frames AI as a practical economic tool, with no evidence of a centralised national AGI push.
Who benefits from the AGI race narrative?
Primarily Western tech companies, venture capital and geopolitical hawks, who use competition framing to justify deregulation, lighter oversight and increased funding.
How could a dangerous AGI race be prevented?
Through track-two diplomatic dialogues, international verification mechanisms, and hardware-level governance such as FlexHEG that embeds agreements directly in chips.

People & ideas in this piece

Seán Ó hÉigeartaighCentre for the Future of IntelligenceDario AmodeiLeopold AschenbrennerArvind NarayananAI PlusTrack-two dialoguesFlexHEGAI as normal technology

Topics: AI & Geopolitics