“The initiative calls for a prohibition on the development of superintelligence until the technology is reliably safe and controllable, and has public buy-in,” according to an open letter published by the Future of Life Institute, a US-based NGO that campaigns against the dangers of AI.
Signatories include the “Godfather of AI” and 2024 winner of the Nobel Prize in Physics Geoffrey Hinton; Computer Sciences Professor at the University of California in Berkeley Stuart Russell; and the world’s most-cited AI scientist Yoshua Bengio of the University of Montreal.
A raft of other public figures have signed: Virgin Group founder Richard Branson, Apple co-founder Steve Wozniak, President Donald Trump‘s former adviser Steve Bannon, and former president Barack Obama‘s national security adviser Susan Rice.
The initiative is also endorsed by the Vatican’s AI expert Paolo Benanti and celebrities such as Prince Harry, his wife Meghan, and the singer will.i.am.
Most major AI developers are striving for artificial general intelligence (AGI), a stage where AI would match all human intellectual capabilities, and even superintelligence, which would exceed them.
Speaking at an event organised by the media group Axel Springer in September, OpenAI CEO Sam Altman, whose company created ChatGPT, said that superintelligence could be achieved within the next five years.
Future of Life Institute President Max Tegmark told AFP that companies should not aim for such an objective without any regulatory framework.
“Many people want powerful AI tools for science, medicine, productivity, and other benefits,” the co-founder of the institute Anthony Aguirre added on Wednesday.
“But the path AI corporations are taking, of racing toward smarter-than-human AI that is designed to replace people, is wildly out of step with what the public wants, scientists think is safe, or religious leaders feel is right.”
The open letter echoes another, published one month ago by AI researchers and sector workers during the United Nations General Assembly, which called for governments “to reach an international agreement on red lines for AI” by the end of 2026.
