More information about the lecture

With the increasing capabilities of artificial intelligence, a crucial question arises: Can we ensure that these systems remain in line with human values? While advances in logical thinking and planning bring us closer to intelligence at a human level, recent findings also reveal concerning behaviors such as deception, hacking, and resistance to shutdown.

In his Munich AI lecture, Yoshua Bengio will examine these challenges and outline a safer path to the future. He advocates for the development of non-agent-based, yet trustworthy AIs - systems modeled after a selfless scientist that dedicate themselves to understanding the world rather than pursuing their own goals. Such "scientist AIs" could act as overseers and help society manage more powerful agent-based systems and reduce existential risks. Beyond technical solutions, Bengio calls for political coordination at national and international levels, suggesting that transformative AI should be treated as a global public good essential for upholding democracy and stability.

About the speaker

Yoshua Bengio is a full professor at the Université de Montréal, founder and scientific advisor of Mila, co-president and scientific director of LawZero, as well as holder of the Canada CIFAR AI Chair. As the recipient of the 2018 A.M. Turing Award – often referred to as the "Nobel Prize of Computing" – he is the most cited computer scientist in the world and one of the most cited living scientists across all disciplines. Bengio is a member of the Royal Society of London and Canada, an Officer of the Order of Canada, a Knight of the French Legion of Honour, and currently the Chair of the International AI Safety Report.

Curriculum Vitae

The panelists