The Physea Wiki
How AI actually works.
A plain-language reference on modern AI, written for people who want to understand the machinery instead of the hype. Built bottom-up: each subject rests on the one before it, every claim is cited to a primary source, and the ideas come with diagrams, not just words.
- 01 Foundations 22 pages What AI, machine learning, and language models actually are. Tokens, training versus inference, and the vocabulary every later subject assumes. Start here — no prerequisites
- 02 Language Model Architecture 26 pages How a language model is built: the transformer, attention, embeddings, context windows, and how text is generated one token at a time. Builds on Foundations
- 03 Models 22 pages Frontier and local models in plain language. Families, sizes, licensing, context, pricing, and how to choose one. Builds on Architecture
- 04 Prompting & Directing Models 21 pages How to steer a model you have chosen: prompts, system prompts, structured output, in-context examples, reasoning, and the common failure modes. Builds on Models
- 05 Agents & Tooling 28 pages An agent is a prompted model given tools, retrieval, and a plan. Agents, tool use and MCP, retrieval, skills, workflows, rules, and the harnesses they run in. Builds on everything above
- 06 Safety & Security 23 pages Prompt injection, jailbreaks, alignment, and data privacy. Where tool use turns a wrong answer into a wrong action. Cross-cutting; matters most once agents can act
- 07 Running AI Yourself 22 pages The practical track: local inference, hardware and VRAM, quantization, serving runtimes, and the local-versus-cloud trade-off. Builds on Architecture + Models
- 08 History & Context 22 pages How the field got here: symbolic AI, the deep-learning revolution, the transformer moment, and the questions still open. Read last — it contextualizes the rest