ChatGPT (Or LLMs really) have discovered regularities in language that humans are not aware of

Prediction market on manifold. Resolves YES if the claim is shown to be true, that in 2023, LLMs have discovered regularities or structures in language that are meaningful to humans, but humans are not aware of themselves while LLMs are.

Liquidity: $1,000. Resolves: 3/24/2100.