Over 1,300 scientists and AI experts seek a six-month stop in the development of powerful AI systems, asking labs engaged in such work not to proceed until risks have become more manageable.
Several major players in the tech sector have signed an open letter, issued by the Future of Life Institute, which calls on AI labs to pause experiments with AI, effective immediately.
A consultation of the EU’s transparency register reveals that the non-profit is primarily funded by the Musk Foundation; concern about the future of AI is a running theme in past comments made by SpaceX, Tesla, and Twitter CEO Elon Musk.
Together with over 1,300 signatories, Musk is worried about the training of systems more powerful than OpenAI’s newly launched GPT-4, the most advanced language model to date.
According to them, such advanced AI systems could pose great risks to society and humans if not handled with care. AI labs, they argue, are in such a race to “deploy ever more powerful digital minds” that the planning and management of those systems are in danger of becoming only secondary considerations.
They believe that AI labs and experts should therefore use the proposed pause to develop and implement much-needed safety protocols.
If this is not done, creators of such systems—which can already compete with humans on certain tasks—will eventually be unable to “understand, predict, or reliably control” them, they warn.
“Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us?” the letter asks, remarking that “such decisions must not be delegated to unelected tech leaders.”
The risk of propaganda and misinformation being spread by these AI systems is another cited risk. “Should we let machines flood our information channels with propaganda and untruth?” reads the next rhetorical question.
Sam Altman, chief executive at OpenAI, was not among those who signed the letter. Notably, other heavyweights, such as Sundar Pichai and Satya Nadella, CEOs of Alphabet and Microsoft, were not included either.
The letter could, however, boast of having garnered the support of Stability AI CEO Emad Mostaque, researchers at Alphabet-owned DeepMind, and AI luminaries Yoshua Bengio (known as one of the ‘godfathers of AI’), and researcher Stuart Russell, a pioneer in the field.