The world is ill-prepared for breakthroughs in artificial intelligence, according to a group of veteran experts including two of the “godfathers” of AI, who warn that governments have not made enough progress in regulating the technology.
Tech companies’ move toward autonomous systems could “massively amplify” the impact of AI, and governments need safety regimes that trigger regulatory action if products reach a certain level of capability, the group said.
The recommendations are issued by 25 experts, including Geoffrey Hinton and Yoshua Bengio, two of the three “godfathers of AI” who won the ACM Turing Award for their work, the computer equivalent of the Nobel Prize.
The crackdown comes as politicians, experts and technology executives prepare to meet at a two-day summit in Seoul on Tuesday.
The academic paper, titled “Managing the Extreme Risks of Artificial Intelligence as it Advances Rapidly”, recommends government security frameworks that introduce stricter requirements if the technology develops rapidly.
It also calls for increased funding for newly established bodies such as the UK and US AI Safety Institutes; forcing technology firms to perform more rigorous risk checks; and limiting the use of autonomous artificial intelligence systems in key societal roles.
“Society’s response, despite promising first steps, is not commensurate with the potential for rapid transformative progress that many experts expect,” said the article, published Monday in the journal Science. “AI security research lags behind. Current governance initiatives lack mechanisms and institutions to prevent abuse and recklessness and barely address autonomous systems.
A global summit on AI security at Bletchley Park in the UK last year brokered a deal on voluntary testing with tech firms including Google, Microsoft and Mark Zuckerberg, while the EU introduced AI legislation and the White House issued an executive order in the US. new AI security requirements.
The document says that advanced artificial intelligence systems — technology that performs tasks typically associated with intelligent beings — could help cure disease and raise living standards, but also carry the threat of disrupting social stability and enabling automated warfare. However, he warns that the technology industry’s shift towards the development of autonomous systems poses an even greater threat.
“Companies are focused on developing general artificial intelligence systems that can act autonomously and track targets. The rise of capabilities and autonomy may soon massively amplify the impact of AI, with risks that include widespread social harm, harmful use and the irreversible loss of human control over autonomous AI systems,” the experts said, adding that unchecked AI development could lead to the “marginalization or extinction of humanity.” .
after promoting the newsletter
The next stage in the development of commercial AI is “agent” AI, a term for systems that can act autonomously and theoretically perform and complete tasks such as booking a holiday.
Last week, two tech firms got a glimpse of that future with the OpenAI GPT-4o, which can conduct real-time voice conversations, and Google’s Project Astra, which could use a smartphone camera to identify locations, read and explain computer code. and create alliterative sentences.
Other co-authors of the proposals include Sapiens bestselling author Yuval Noah Harari, the late Daniel Kahneman, Nobel laureate in economics, Sheila McIlraith, a professor of artificial intelligence at the University of Toronto, and Dawn Song, a professor at the University of California, Berkeley. The document published on Monday is a peer-reviewed update of original proposals made before the Bletchley meeting.