Three highly respected AI researchers—Nobel Prize winner Geoffrey Hinton, World Labs CEO and co-founder Fei-Fei Li, and Coursera co-founder Andrew Ng—discussed the future of AI at the Ai4 conference in Las Vegas last week. The debate focused on open-source access, regulation, and global competition in AI development.
Andrew Ng voiced concerns about a potential scenario where a limited number of major AI companies control the pace of progress, similar to how Apple and Google influence mobile operating systems. He advocated for promoting openness to ensure broad access to AI technology and foster competition among multiple providers.
Geoffrey Hinton differentiated between open-source software and open-weight models, highlighting that open-weight models, which release the parameters of trained AI models, could facilitate their use for malicious purposes like cyber attacks. Despite these reservations, Hinton acknowledged that open-weight models are now a permanent part of the AI landscape, stating, "I think that battle's been lost."
Fei-Fei Li argued against a strict dichotomy between complete openness and complete closedness in AI development, suggesting a more nuanced approach. She drew parallels with nuclear physics, where different layers of an ecosystem operate with varying levels of openness. Li also emphasised the potential for AI to serve as an infrastructure, benefiting from both open and closed systems, and highlighted the importance of global partnerships and education.
All three experts agreed on the necessity of some level of regulation to guide AI development in a beneficial direction. Hinton stated, "You can't leave it to people like Elon Musk and Mark Zuckerberg to decide how AI should be done."