AI Is Learning at the Wrong Level of Abstraction — Matthieu Wyart

This episode is sponsored by Notion. Learn more about Notion's Developer Platform today at notion.com/mlstWhy can deep networks discover abstractions that shallow models miss? Statistical physicist Matthieu Wyart joins Tim Scarfe to argue that the answer lies in the hidden hierarchy of data. Language and images are built from parts within parts; depth lets a network recover those coarse-grained variables and escape the curse of dimensionality.The conversation moves from jamming transitions and rough loss surfaces to Chomsky, context-free grammars and machine creativity. Wyart explains why next-token prediction can still recover compositional structure, where current systems fall short of genuine scientific invention, and why predicting latent representations rather than raw tokens could make learning far more sample-efficient.They also examine diffusion models, neural scaling laws and the limits of physics-inspired theory. The final question is on a personal note: if mistakes are the price of leaving the beaten path, how much scientific risk is worth taking?---TIMESTAMPS:00:00:00 Can machines learn abstractions from data?00:02:00 Notion agentic workspace00:02:49 From statistical physics to machine learning00:06:40 What physics can explain about learning00:16:37 From Carnot to Chomsky bulldozer00:21:21 How deep networks recover hidden hierarchies00:32:43 Where machine creativity still falls short00:40:48 How deep nets escape the curse of dimensionality00:52:19 Why predict latents instead of tokens01:02:49 The sample-efficiency case for latent prediction01:08:31 Diffusion, scaling laws and text entropy01:16:40 The scientists we learn from and the mistakes we make---REFERENCES:person:[00:00:43] Noam Chomsky00:02:08]" rel="nofollow noopener ugc" target="_blank">linguistics.mit.edu/user/chomsky/tool:[00:02:08] Notion Developer Platform00:04:43]" rel="nofollow noopener ugc" target="_blank">notion.com/en-gb/blog/introducing-developer-pl… Mastering the game of Go with deep neural networks and tree search00:05:52]" rel="nofollow noopener ugc" target="_blank">nature.com/articles/nature16961[00:05:52] Reconciling modern machine-learning practice and the bias-variance trade-off00:25:54]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/1812.11118[00:25:54] How Deep Neural Networks Learn Compositional Data: The Random Hierarchy Model00:42:12]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/2307.02129[00:42:12] Efficient Estimation of Word Representations in Vector Space00:52:46]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/1301.3781[00:52:46] Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture00:52:54]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/2301.08243[00:52:54] Learn from your own latents and not from tokens: A sample-complexity theory01:08:31]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/2605.27734[01:08:31] A Phase Transition in Diffusion Models Reveals the Hierarchical Nature of Data01:11:39]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/2402.16991[01:11:39] Scaling Laws for Neural Language Models01:12:17]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/2001.08361[01:12:17] Deriving Neural Scaling Laws from the statistics of natural language01:13:34]" rel="nofollow noopener ugc" target="_blank">arxiv.org/abs/2602.07488[01:13:34] Prediction and Entropy of Printed Englishieeexplore.ieee.org/document/6773263---LINKS:D… PDF transcript: app.rescript.info/share/f7644cdaa86c5cc1e41e48…