AI at a Defining Moment: Ensuring Safety Through Technical and Societal Safeguards
Information
With mounting evidence of frontier AI systems exhibiting self-preserving and deceptive behaviors, there is growing urgency to steer AI development in a safer direction. In this session, Yoshua Bengio, Full Professor at UdeM, Co-President and Scientific Director of LawZero and the most cited scientist in the field, will sit down with Harry Booth, AI reporter at TIME Magazine's London bureau, to discuss how society can confront these risks and how LawZero's solution, Scientist AI, may offer a promising technical path forward. Frontier AI models are increasingly showing signs of deception and self-preservation.
Are we prepared for what that entails? Is it possible to build fundamentally safe and trustworthy frontier AI models? How can we ensure superintelligent AI remains aligned with human values?
