Season 2 · Episode 336
The World Model Revolution: Beyond LLM Token Prediction
Herman and Corn explore why LLMs struggle with logic and how the shift to world models is giving AI a sense of physics and spatial reality.
My Weird Prompts · Daniel Rosehill
January 28, 202628m 45s
Audio is streamed directly from the publisher (dts.podtrac.com) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.
Show Notes
In this episode of My Weird Prompts, Herman and Corn tackle a growing frustration in the AI community: the "reasoning wall" hit by traditional large language models. As users notice coding assistants collapsing under the weight of complex architectural changes, the brothers discuss why statistical token prediction is no longer enough. They explore the emergence of world models—AI systems designed to internalize the laws of physics, causality, and 3D space. From Meta’s JEPA architecture to the spatial intelligence breakthroughs at World Labs, this conversation maps out the transition from AI that merely "speaks" to AI that truly "understands" the environment it operates in. By examining the synergy between intuitive "System 1" language models and logical "System 2" world simulators, Herman and Corn provide a roadmap for the next stage of artificial general intelligence and what it means for the future of robotics, autonomous systems, and software development.