> For the complete documentation index, see [llms.txt](https://isubasinghe.gitbook.io/isithas-wiki/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://isubasinghe.gitbook.io/isithas-wiki/computer_science/ai/hutter_prize.md).

# Hutter Prize

## Basic idea

A €500,000 prize for losslessly compressing a 1 GB snapshot of English Wikipedia. The premise: better compression = better world-model = better AI. Compression and prediction are formally equivalent.

## Key facts

* Optimal compression length $\approx$ Shannon entropy of source: $L \ge H(X)$.
* Compressor that predicts $P(x\_{n+1}|x\_{1..n})$ well achieves codelength $-\log\_2 P$ via arithmetic coding.
* Prize threshold: beat the previous record by $\ge 1%$.

Hutter prize essentially states that AI and compression are the same problem

## Resources

<https://en.wikipedia.org/wiki/Hutter_Prize>