Geoffrey Hinton Claims Congress Has One Year to Act on AI
Must Congress halt advanced AI before it escapes human control or reject a freeze that could hobble U.S. innovation?
Above: Geoffrey Hinton in Washington, D.C. on Sept. 16. Image credit: Roberto Schmidt/Getty Images
The Spin
Establishment-critical narrative
Congress must slam the brakes on advanced AI before systems escape human control. Corporate promises cannot substitute for federal safety rules when less-monitorable models are released despite cyber risks. A pause on dangerous development and firm limits on superintelligence are the only responsible course.
Pro-establishment narrative
Congress should reject an AI freeze that would hobble American innovation while foreign rivals race ahead. Flexible industry standards can address concrete risks faster than rigid rules and avoid handing entrenched firms a regulatory moat. Target frontier systems with company-led safety tests, not a sweeping ban imposed by politicians who don't understand the technology they legislate.
