A group of over 1,100 tech executives and researchers, including Elon Musk, have warned of the dangers of artificial intelligence (AI) and called for a six-month pause on the developments of the likes of ChatGPT and Google's Bard.
Citing concerns over a “dangerous” arms race, a letter was published by the non-profit campaign group Future of Life Institute, with other signatories including Apple co-founder Steve Wozniak, top AI professors Stuart Russell, Yoshua Bengio and Gary Marcus, and founders of AI startups Stability AI, Character.ai, Unanimous AI and Attain.ai.
The letter urged all labs developing AI to halt the training of AI systems more powerful than GPT-4, or suggested there should be government intervention.
It also called for the creation of shared safety protocols audited by independent experts to ensure that AI systems with human-competitive intelligence are safe beyond a reasonable doubt.
The intervention comes as governments around the world formulate policy responses to the rapidly evolving field of AI.
The letter said people should ask themselves: "Should we let machines flood our information channels with propaganda and untruth? Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization?
"Such decisions must not be delegated to unelected tech leaders. Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable. This confidence must be well justified and increase with the magnitude of a system's potential effects."
Last month, OpenAI, the developers of ChatGPT, said "At some point, it may be important to get independent review before starting to train future systems, and for the most dvanced efforts to agree to limit the rate of growth of compute used for creating new models."
The Future of Life letter said, "we agree. That point is now."
Geoffrey Hinton, the former University of Sussex professor known as the 'godfather of artificial intelligence', who works with Google and mentors many AI startups, said it was "not inconceivable" that AI could try to wipe out humanity.
As for when general-purpose AI might be ready, he told CBS (NYSE:CBS): "Until quite recently, I thought it was going to be like 20 to 50 years before we have general purpose AI. And now I think it may be 20 years or less."
With computers close being able to come up with ideas to improve themselves, he said this was "an issue" and "we have to think hard about how you control that."
More importantly Hinton suggested regulation was important, as the technology could give a handful of companies or governments tremendous power: "I think it's very reasonable for people to be worrying about these issues now, even though it's not going to happen in the next year or two. People should be thinking about those issues."
The UK government said today it proposes to "avoid heavy-handed legislation" on AI so as to not stifle investment.