Some of the world’s leading experts in artificial intelligence are sounding the alarm about the “societal-scale risks from AI,” including the possibility of human extinction.
Hundreds of AI scientists and other notable figures have signed a statement on AI risk from the Centre for AI Safety (CAIS), a San Francisco-based non-profit organization that’s supported by private contributions.
Noting the difficulty in parsing all of the risks presented by the advancement of AI, the single-sentence statement is intended to “open up discussion” and create “common knowledge of the growing number of experts and public figures who also take some of advanced AI’s most severe risks seriously.”
“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war,” the statement reads in its entirety.
Signatories of the statement include Sam Altman, the CEO of OpenAI; Demis Hassabis, CEO of Google DeepMind; and Geoffrey Hinton, a former University of Toronto computer science professor who has been called the Godfather of AI.
CAIS says its mission is to “reduce societal-scale risks from AI” with a focus on serving the public interest rather than those of donors.
The organization details the heightened risks that accompany AI in a 36-page existential risk analysis, including AI’s applications in warfare, the spread of misinformation and the enabling of power-seeking behaviour and related imbalances.
“Currently, society is ill-prepared to manage the risks from AI. CAIS exists to equip policymakers, business leaders, and the broader world with the understanding and tools necessary to manage AI risk,” notes the CAIS website.
Hinton, 75, spent the last decade working for both a deep learning AI research team under the umbrella of Google AI and the University of Toronto before publicly announcing his departure from Google earlier this month.
Speaking with The New York Times from his Toronto home, Hinton said he left the company so he could talk about the dangers of AI without considering the effect on Google. He has defended the company’s handling of AI, saying Google has acted “very responsibly,” but he still has concerns about the future, along with some regrets about his previous work.
“Look at how it was five years ago and how it is now,” Hinton told the Times. “Take the difference and propagate it forwards. That’s scary.”
The risks Hinton warns about include AI’s ability to generate false photos, videos and audio, blurring the lines between truth and fiction, the effect on the job sector, and a “nightmare scenario” where AI machines learn to write their own computer code.
“Right now, they’re not more intelligent than us, as far as I can tell. But I think they soon may be,” Hinton told BBC in another interview.
Earlier this year, the Future of Life Institute, a non-profit focused on steering “transformative technologies away from extreme, large-scale risks and towards benefiting life,” issued an open letter urging all AI labs to pause training systems more powerful than GPT-4, the newest version of OpenAI’s language model systems, for a period of six months.
The letter, which has nearly 32,000 verified signatories, including Elon Musk and Apple co-founder Steve Wozniak, notes that advanced AI could lead to “profound change in the history of life on Earth and should be planned for and managed with commensurate care and resources.”
“Unfortunately, this level of planning and management is not happening, even though 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,” it adds.
The letter called for the pause to allow labs to develop safety protocols and risk management strategies “that are rigorously audited and overseen by independent outside experts.”
The letter also calls on policymakers to develop “robust AI governance systems,” including establishing new regulatory authorities, oversight and tracking tools, provenance and watermarking systems to help distinguish real from synthetic, as well as liability for AI-caused harm and robust public funding for technical AI safety research.
More resources are needed to cope with “the dramatic economic and political disruptions (especially to democracy) that AI will cause,” the letter notes.
Earlier this month, Altman issued a similar warning, telling U.S. Congress during a Senate hearing that government intervention will be critical in mitigating the risks of increasingly powerful AI systems.
“As this technology advances, we understand that people are anxious about how it could change the way we live. We are too,” he said.
Post a Comment