AI Godfather Geoffrey Hinton Doubts Good AI Will Win over Bad AI

Geoffrey Hinton, who has been called the ‘Godfather of AI’ and a professor at the University of Toronto, expressed concern about the rapid development of artificial intelligence (AI) at the Collision 2023 technology conference in Toronto. Hinton, who recently left Google to freely criticize a field he has helped pioneer, cautions about the uncontrolled acceleration of AI development, especially in generative AI tools like ChatGPT and Bing Chat. He expressed doubts about whether good AI will triumph over bad AI and stated that applying AI ethically may come at a large cost.

“AI is Only as Good as Its Creator”

Hinton argued that AI is only as good as its creator and bad technology can still win. He expressed concern about the potential for AI to be used in warfare and the possibility of AI exacerbating wealth inequality. He also reaffirmed his view that AI could pose an existential risk to humanity, stating that if AI becomes smarter than humans, there is no guarantee that humans will remain in control.

Bias and Misinformation in AI

Hinton highlighted the problems that exist with AI, including bias and discrimination due to skewed training data, and the creation of echo chambers that amplify misinformation. He also expressed concern about AI spreading misinformation outside of this echo chamber and stressed the importance of flagging such fake content.

“I think it’s important for people to understand this is not just science fiction; it’s not just scaring”– godfather #AI, @geoffreyhintontalk about the risks of AI in the Center Stage at #CollisionConf pic.twitter.com/u3iM0LO4hP

— Collision Conf (@CollisionHQ) June 28, 2023

The Potential of AI and the Need for Regulation

Despite his concerns, Hinton acknowledged the potential benefits of AI but warned that their realization may come at a high price. He suggests that humans may need to do some empirical work to understand how AI can go wrong and prevent it from taking control. He also called for more balance in AI development efforts, with a greater focus on understanding and managing AI risks.

While Hinton expressed concern, other industry figures at the Collision conference were more optimistic. Adam Selipsky, CEO of Amazon Web Services, discusses Amazon’s ongoing work on its own grand language model, which will be released later this year. Google DeepMind head of business Colin Murdoch and Roblox Chief Scientist Morgan McGuire also expressed hope in the potential of AI to address global challenges and enhance creativity.

.@DeepMind chief business officer Colin Murdoch joins us on the Center Stage to talk all about AI and to explore how DeepMind is working to solve global challenges through the use of AI pic.twitter.com/s0WeR43TKb

— Collision Conf (@CollisionHQ) June 28, 2023

Hinton’s comments reflect a broader debate within the tech industry about the future of AI. While AI has the potential to revolutionize many aspects of society, its rapid development also raises significant ethical and practical issues. Balancing these risks and benefits will be a key challenge for policy makers, technology companies and society as a whole in the coming years.

AI Rules: Recent Developments