About

This site explains how AI systems work: protocols like MCP, agent architecture, model internals, and the training techniques underneath them. The goal is a correct, sourced explanation of the mechanism, not a tutorial for using any specific product.

[QUESTION: this page needs a real author bio and a real answer to 'why should you trust this explanation' — see the open question in .claude/rules/content-drafting.md about this site's equivalent of first-hand operator experience. Placeholder copy until that's answered.]