- Published on
Thousands of employees call on US government to control AI development, OpenAI CEO Altman responds
- Authors
- Name
- DeepHorus
On July 29, Beijing time, over 1,100 AI employees from companies such as OpenAI, Anthropic, Google, and Meta jointly signed an open letter calling on the U.S. government to control the pace of artificial intelligence development. This initiative, called "Pacing the Frontier," aims to promote international cooperation and develop necessary technical and governance tools to address the risks posed by potential automated AI development.
The open letter specifically focuses on "automated AI development," which refers to the future ability of AI to develop and improve its own systems, known in the industry as "recursive self-improvement." Anthropic previously stated that its Claude series models are rapidly approaching this threshold. The letter warns that the speed of AI capability development could accelerate quickly, surpassing the ability of humans to understand or control.
Signatories include Anthropic CEO Dario Amodei, co-founders Jack Clark and Jared Kaplan, OpenAI Chief Scientist Jakub Pachocki, Meta Chief Scientist Zhao Shengjia, and Google DeepMind Security Chief Anca Dragan. In the open letter, these executives emphasized that AI companies are currently competing to develop more powerful systems, but safety guarantees and regulatory mechanisms may lag behind.
OpenAI posted on X platform that AI acceleration in future frontier model development could be so fast that the world needs to pace the progress. Anthropic also pointed out that its research on recursive self-improvement shows that tools are needed to deliberately regulate the speed of AI development, allowing society time to prepare.
OpenAI CEO Sam Altman's stance has shifted. In an interview with the "Invest Like the Best" podcast, he said that it might be necessary to "pace" the rhythm of AI development to allow society to adapt to new capabilities and establish protective mechanisms. He also emphasized that this approach should not be seen as regulatory capture or collusion between labs. Previously, Altman had avoided signing similar initiatives calling for a slowdown in AI development and considered a related open letter in 2023 lacking key technical details.
This collective appeal reflects the industry's growing concern about the potential risk of AI getting out of control. As recursive self-improvement approaches, various sectors of society are seeking cooperative mechanisms to balance innovation and safety.
