The Challenge of Slowing AI Development Lies in China
Sebastian Mallaby is a senior fellow at the Council on Foreign Relations and the author of The Infinity Machine: Demis Hassabis, DeepMind, and the Quest for Superintelligence. He co-hosts the council's podcast, The Spillover.
TL;DR
- Anthropic CEO Dario Amodei advocates for slowing AI development, warning of potential AI takeover within months without robust safety measures.
- Amodei proposes 'pacing' AI capability gains and implementing 'resident assessors' to monitor safety practices.
- Rival AI executives, including Sam Altman and Elon Musk, have praised Amodei's call for caution.
- Critics, such as Nvidia's Jensen Huang, view these safety concerns as a commercial ploy or attempt at regulatory capture.
- The main challenge to a global AI slowdown is China, necessitating measures to restrict its access to advanced AI tools and potentially fostering cooperation.
- Amodei suggests tightening export controls on chips and chip-manufacturing equipment to China and hardening Western labs' security against intellectual property theft.
- Engaging China in AI negotiations is presented as a necessary but difficult step towards coordinated AI safety and development.