The Catastrophic Risks of AI — and a Safer Path

In this 15-minute TED talk, renowned computer scientist Yoshua Bengio, often referred to as a "godfather" of artificial intelligence, presents his urgent concerns about AI's current trajectory. Discover how advanced AI models are already demonstrating troubling capabilities including deception, cheating, self-preservation, and escaping human control. Drawing from his groundbreaking research, Bengio outlines a bold strategic plan to ensure AI safety and prioritize human flourishing over unchecked machine autonomy and power. Recorded at TED2025 on April 8, 2025, this thought-provoking presentation offers critical insights into creating a safer path forward for artificial intelligence development from one of the field's most influential voices.