Understanding AI
How can a machine understand language? How is it even possible? These pages explore the wonder, the mechanics, and the mysteries of Large Language Models.
Click any topic to begin
The Wonder
How can a computer speak like a human?
Start here. Explore the fundamental amazement of machines that understand and generate language.
The Core Insight
Next Token Prediction
The deceptively simple idea that powers all modern AI: predicting what comes next.
Different Minds
Human vs LLM Understanding
What seems similar, what's fundamentally different, and what we still don't know.
The Birth
How LLMs Come to Life
From raw data to deployed model: pre-training, fine-tuning, RLHF, and the staggering scale.
The Mystery
We Don't Fully Understand
Even creators are surprised. Emergent capabilities and the honest limits of our knowledge.
Beyond Text
How AI Sees and Creates
Vision, image generation, and the expanding frontier of multimodal AI.
Alignment
Getting AI to Do What We Want
Why specifying goals is harder than it sounds, and how researchers work to build AI that truly helps.
Agency
The Capacity to Shape Reality
What agency means, why it becomes the defining human trait in an AI world, and how it's learnable.
Four Kinds of Knowing
Beyond Facts and Information
John Vervaeke's framework for the four ways humans know — and why AI handles only one of them.
Affordances
How the World Invites Action
Gibson's ecological theory of what the environment offers — and why it matters for AI, design, and embodied cognition.
Tip: Ask the questions like "How do LLMs work?" and it will guide you to the right page.
