Musk, Wozniak Support Pausing the Training of ‘Powerful’ AI

Billionaires Elon Musk and Steve Wozniak, Apple’s cofounder, joined more than 1,000 people to sign an open letter calling for a 6-month pause on development of advanced AI systems as they “pose profound risks to society and humanity.”

Penned by nonprofit Future of Life Institute, the letter said that in recent months AI labs have been “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.”

OpenAI creator ChatGPT, released to the public in November 2022, has become the fastest growing app of all time. The AI chatbot’s human-like responses to queries, or prompts, has led to an increase in investment from and much closer collaboration with Microsoft. ChatGPT capability is now being incorporated across the software giant’s products. This sparked a competitive response from Google, which rushed to match these announcements.

But the Institute said such advanced tech needs to be “planned for and managed with commensurate care and resources. Unfortunately, this level of planning and management is not happening.”

The letter has garnered nearly 1,200 signatures with many more names on hold for verification “due to high demand," the nonprofit said in a blog.

Besides Musk and Wozniak, other notable signers include Turing award winner Yoshua Bengio, AI luminary Stuart Russell, former presidential candidate Andrew Yang, and the co-founders of Skype, Pinterest, Ripple among many others.

Even Emad Mostaque, CEO of Stability AI, the maker of popular text-to-image generator Stable Diffusion, signed. There were a few signers from Google, DeepMind and Microsoft, but none from OpenAI.

One notable dissenter is Turing award winner Yann LeCun, who is also Meta’s chief AI scientist. “Nope. I did not sign this letter. I disagree with its premise,” he tweeted. LeCun later deleted the tweet.

Fearing generative AI

The Institute – headed by MIT physics professor Max Tegmark − said current AI systems are now becoming as good as humans at doing general tasks.

“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.”

The nonprofit even pointed to OpenAI’s own statement regarding artificial general intelligence: “At some point, it may be important to get independent review before starting to train future systems, and for the most advanced efforts to agree to limit the rate of growth of compute used for creating new models.”

The Institute said it agrees with OpenAI and thus is calling for all AI labs to pause for six months the training of AI systems more powerful than GPT-4, which is OpenAI’s latest large language model (LLM). ChatGPT was trained on an older version, GPT-3.5.

During this respite, AI labs and independent experts should come together to develop safety protocols for advanced AI design that will be audited by third parties.

However, the Institute was quick to clarify that this does not mean all work on AI must stop, but “merely a stepping back from the dangerous race to ever-larger unpredictable black-box models with emergent capabilities.”

Gary Marcus, professor emeritus at New York University, said he is not exactly sure “what counts as more powerful than GPT-4” but signed the letter nonetheless because he agrees with the spirit of it.