Bloomberg

Artificial intelligence experts and industry leaders, including Elon Musk, University of California Berkeley computer science professor Stuart Russell and Apple co-founder Steve Wozniak, are calling on developers to hit the pause button on training powerful AI models.

More than 1,100 people in the industry signed a petition calling for a six-month break from training artificial intelligence systems more powerful than the latest iteration behind OpenAI’s ChatGPT, to allow for the development of shared safety protocols.

“Recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control,” said an open letter published on the Future of Life Institute website. “Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable.”

The call comes after the launch of a series of AI projects in the past several months that convincingly perform human tasks such as writing emails and creating art. Microsoft-backed OpenAI released its GPT-4 this month, a major upgrade of its AI-powered chatbot, capable of telling jokes and passing tests like the bar exam.

OpenAI Chief Executive Officer Sam Altman – whose name appeared on the list of signatories Tuesday night – didn’t sign the petition, said OpenAI spokesperson Hannah Wong. “Also I think it’s important to point out that we spent more than six months – after GPT-4 finished training – on the safety and alignment of the model,” Wong said.

Alphabet’s Google and Microsoft are among the companies using artificial intelligence to enhance their search engines, while Morgan Stanley has been using GPT-4 to create a chatbot for its wealth advisers.

Developers should work with policymakers to create new AI governance systems and oversight bodies, according to the letter. It called on governments to intervene in the development of AI systems if major players don’t imminently agree to a public, verifiable pause.

“AI research and development should be refocused on making today’s powerful, state-of-the-art systems more accurate, safe, interpretable, transparent, robust, aligned, trustworthy and loyal,” it said.

Yoshua Bengio, the founder and scientific director of Canadian AI research institute Mila, signed the petition, according to a statement from the institute. Emad Mostaque, founder and CEO of Stability AI, also said he signed it.

“We have seen the amazing capabilities of GPT-4 and other massive models. Those making these have themselves said they could be an existential threat to society and even humanity, with no plan to totally mitigate these risks,” Mostaque said. “It is time to put commercial priorities to the side and take a pause for the good of everyone to assess rather than race to an uncertain future.”

The Future of Life Institute is a nonprofit that seeks to mitigate risks associated with powerful technologies and counts the Musk Foundation as its biggest contributor.

“All of the top signatories on the list have been independently verified,” said Anthony Aguirre, a spokesperson for the institute. “Doing so for the whole list exceeds our capacity.”

A spokesperson for the Center for Humane Technology, whose executive director, Tristan Harris, signed the letter, said steps are being taken to prevent fake signatures. New signatories are now requiring human review before going up on the site, and all high-profile signatories listed have been vetted by direct communication, the spokesperson said in an email.