Table of Contents
- 1What It Is — The Pattern Machine
- 2How It Learns — Wrong to Less Wrong
- 3How It Guesses — Statistics in Action
- 4What AI Is Actually Doing
- 5One Neuron — The Atom of Every AI System
- 6Networks — Layers, Matrix Math, and GPUs
- 7Training — How a Model Actually Learns
- 8Attention — The Mechanism That Changed Everything
- 9The Full 200 Lines — Reading It Top to Bottom
- 10What This Understanding Changes
- 11Teaching AI to See
- 12Hearing and Speaking
- 13Motion and Video
- 14Networks, Security, and Sequences
- 15The Architecture Meets Science
- 16What Text Models Are Actually For
- 17The Mismatch Problem
- 18How to Choose
- 19The Prompt Is the Program
- 20Context Is the Model's Entire Reality
- 21What Good Context Looks Like
- 22Advanced Patterns
- 23Retrieval-Augmented Generation
- 24Fine-Tuning
- 25Agents, MCP, and the Efficient Edge
- Epilogue · What You Now Know
- Appendix A · The Complete 200-Line Listing
- Appendix B · Further Reading
- Glossary & Index
From Chapter 1 · What It Is — The Pattern Machine
Forget everything you've seen in movies about AI. The real thing is stranger, simpler, and more interesting than any of that.
It's a Very Sophisticated Autocomplete
You've used autocomplete your whole life. Your phone suggests the next word when you're texting. Google finishes your search query before you type it. Email clients suggest how to end a sentence.
A generative AI model — the kind that powers ChatGPT, Claude, Gemini, and every other large language model — is doing the same thing. Just at a scale that makes the results feel uncanny.
When you ask an AI a question, it doesn't look up the answer in a database. It doesn't reason through the problem the way a person would. It predicts what the most likely response looks like, word by word, based on patterns it absorbed from an enormous amount of human-written text.
Where the 'Intelligence' Actually Comes From
Here's the part that surprises most people: the model wasn't taught facts. Nobody sat down and programmed in "the capital of France is Paris" or "water boils at 100°C." The model learned those things the same way it learned everything else — by reading an enormous amount of text where those facts appeared, repeatedly, in consistent patterns.
What looks like knowledge is really pattern density. The more consistently something appears in training data, the more confidently the model reproduces it…
Not ready to buy? The free Overview track at learn.tgifriday.com covers this material and more — about 15 minutes, no account required.