About this explorable

The Confident Machine is an interactive, sequential explainer about hallucinations in large language models: what they are, why they happen, how they are measured, and how to build and use AI systems despite them.

It is inspired by the tradition of explorable explanations: essays where diagrams are instruments you can play with, not just pictures you look at.