In this episode of Hidden Layers, Ron is joined by Michael Wharton and Dr. ZZ Si to explore one of the most pressing and puzzling issues in AI: hallucinations. Large language models can tackle advanced topics like medicine, coding, and physics, yet still generate false information with complete confidence.

The discussion unpacks why hallucinations happen, whether they’re truly inevitable, and what cutting-edge research says about detecting and reducing them. From OpenAI’s latest paper on the mathematical inevitability of hallucinations to new techniques for real-time detection, the team explores what this means for AI’s reliability in real-world applications.

In this episode of Hidden Layers, we dive into the most important AI developments of the month. We cover OpenAI’s highly anticipated and controversial GPT-5 release, debate where we really are on the AGI timeline, explore groundbreaking new world models like Google’s Genie 3 and Tencent’s Huanyuan Gamecraft, and unpack Meta’s DINO V3 image encoder breakthrough.

In this episode of Hidden Layers, host Ron talks with Dr. Hannah Lu, assistant professor at the University of Texas at Austin and core faculty at the Odin Institute for Computational Engineering and Sciences. Dr. Lu is pioneering the use of AI-powered surrogate models to make complex scientific simulations—like CO₂ absorption in geological formations—faster, more accurate, and more useful for real-world decision-making.

They discuss:

• How surrogate models work and why they’re so powerful
• The challenges of applying AI to physics-based systems
• How digital twins and uncertainty quantification are shaping the future of environmental modeling
• The intersection of generative AI, physics constraints, and climate science