Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore the reality behind AI agent hype in this conference talk that critically examines whether 2025 will truly be the breakthrough year for AI agents and reasoning models. Discover why current AI agents, from software engineering tools to web automation systems, fall significantly short of their marketed capabilities, using examples like the Rabbit R1 and Humane Pin failures. Learn about the substantial gap between claimed performance and real-world utility, while gaining insights into best practices for proper agent evaluation. Understand common pitfalls in agent development and evaluation methodologies that lead to inflated performance claims. Master practical approaches for building AI agents that deliver genuine value rather than empty promises, with guidance on avoiding known shortcomings that plague current implementations. Gain a critical perspective on the AI agent landscape from an expert who co-authored "AI Snake Oil" and has been recognized as one of TIME's 100 most influential people in AI, helping you separate legitimate advances from marketing hype in the rapidly evolving field of AI agents.
