The Information Machine

Yann LeCun (@ylecun) explains why LLMs are limited in terms of real-world intelligence during a Bloomberg interview.

Rohan Paul Twitter · Rohan Paul (@rohanpaul_ai) · 2026-06-18

Yann LeCun argues in a Bloomberg interview that LLMs are fundamentally limited in real-world intelligence because language is an impoverished, symbolic approximation of the world rather than a full representation of it.

Open original ↗

Extraction

Topics: llm-limitationsworld-modelsagi-debateyann-lecun

Claims

  • Language is an approximate, reduced, quantized, and simplified description of the world.
  • LLMs can only process discrete sequences of symbols, not richer representations of reality.
  • These symbolic constraints make LLMs fundamentally limited as a path to real-world general intelligence.

Key quotes

Language is a very approximate, reduced, quantized, and simplified description of the world, and LLMs can only deal with discrete sequences of symbols.