AIhub monthly digest: May 2026 – AI for science, the lottery ticket hypothesis, and world models

LeCun's findings could redefine AI's approach to understanding complex systems, but real-world application remains challenging due to environmental variability. The post Yann LeCun’s paper reveals conditions for LeJEPA to learn world models appeared first on Crypto Briefing.

DeepMind's shift to 'world models' could redefine AI's role in robotics and scientific discovery, emphasizing causality over language processing. The post Google DeepMind CEO Demis Hassabis says language models can’t understand reality, pushes for ‘world models’ appeared first on Crypto Briefing.

The AIhub coffee corner captures the musings of AI experts over a short conversation. This month we delve into world models. What are they, and what potential do they have? Joining the conversation this time are: Sanmay Das (Virginia Tech), Rina Dechter (University of California, Irvine), Tom Dietterich (Oregon State University), Sabine Hauert (University of […]

Listen to the session or watch below AI companies want to build systems that understand the external world and overcome the limitations of LLMs. Recent developments have brought world models to the forefront of the AI discussion. Watch a conversation with editor in chief Mat Honan, senior AI editor Will Douglas Heaven, and AI reporter…

In this crosspost from AI Matters – a publication of the ACM SIGAI – Ella Scallan sat down with Jonathan Frankle to discuss the lottery ticket hypothesis, for which he was awarded the 2023 AAAI/ACM Doctoral Dissertation Award. In this wide-ranging conversation, Jonathan delves into empiricism vs theoretical proofs, how the approach to computer science […]

AI video generation startup Runway is betting that video generation is the path to world models. And that being an AI outsider is an advantage, not a liability.

Insider Brief PRESS RELEASE — Origin Lab, the technology platform turning licensed game worlds into structured training data for world models and multimodal AI, announced an $8M seed round led by Lightspeed Venture Partners. The financing will accelerate Origin Lab’s software, capture, enrichment, QA, search, and delivery systems, while expanding its applied research work in […]

GRASP is a new gradient-based planner for learned dynamics (a “world model”) that makes long-horizon planning practical by (1) lifting the trajectory into virtual states so optimization is parallel across time, (2) adding stochasticity directly to the state iterates for exploration, and (3) reshaping gradients so actions get clean signals while we avoid brittle “state-input” gradients through high-dimensional vision models. Large, learned world models are becoming increasingly capable. They can predict long sequences of future observations in high-dimensional visual spaces and generalize across tasks in ways that were difficult to imagine a few years ago. As these models scale, they start to look less like task-specific predictors and more like general-purpose simulators. But having a powerful predictive model is not the same as being able to use it effectively for control/learning/planning. In practice, long-horizon planning with modern world models remains fragile: optimi