Photo by Anastasiya Gepp on Pexels

A potential paradigm shift is brewing in the AI world, with researchers exploring “World Models” as the next frontier beyond Large Language Models (LLMs). Sparked by discussions on Reddit, the shift suggests that LLMs might be approaching their performance limits, prompting major AI labs to prioritize AI systems capable of understanding 3D space and physics. Evidence cited includes Fei-Fei Li’s “Marble” project, the “From Words to Worlds” manifesto, reports of Yann LeCun’s pivot toward World Models, Jeff Bezos’ involvement in “Project Prometheus” aimed at physical AI, and Tencent’s increased focus on physics-aware models. The anticipated release of AR hardware from Google and Samsung further suggests an emerging ecosystem ready to house these spatially intelligent AI systems. This transition signifies a move from predicting the next word to predicting the next frame and physical interaction, enabling AI to truly understand and interact with the world. The original discussion can be found on Reddit: [https://old.reddit.com/r/artificial/comments/1p0cal1/the_shift_from_llms_to_world_models_am_i_thinking/]