Learn

Understand the tools you use

No jargon for its own sake. No sales copy dressed up as explanation. Just honest accounts of how these things actually work.

Explore the map →

Learning tracks — start from zero

Five ordered paths from "what is AI" to working with agents. No code required to start. Pick the one that matches where you are.

Track 00

AI from zero

No code, no math fear. What AI actually is, what data is, and how prediction works — for anyone starting cold.

8 stops · ◈ 8 interactive

Track 01

Foundations of ML

How machines learn from examples: features, loss, gradient descent, and the eternal fight against overfitting.

8 stops · ◈ 8 interactive

Track 02

Deep Learning

From a single neuron to the transformer: the architecture story that leads directly to modern AI.

8 stops · ◈ 8 interactive

Track 03

LLMs & Generative AI

How the models you actually use get trained, tuned, shrunk, and pushed to reason.

18 stops · ◈ 18 interactive

Track 04

Working with AI

The practical layer: prompting, retrieval, agents, evals — and the failure modes that bite in production.

17 stops · ◈ 17 interactive

All concepts — 59 articles · 7 groups

Grouped by where each idea lives in the stack — read in any order, but each group reads as a story if you take it top-to-bottom.

Foundations

8 concepts

Zero assumed. What AI actually is, what data is, and why statistics runs the show.

Bias in Data

Why models inherit the world's blind spots.

◈ interactive5 min read

Correlation vs Causation

The mistake every model — and every human — makes.

◈ interactive5 min read

Data & Datasets

The raw material everything else is made of.

◈ interactive5 min read

Patterns & Predictions

The core intuition behind all machine learning.

◈ interactive5 min read

Probability & Uncertainty

Thinking in likelihoods instead of certainties.

◈ interactive5 min read

Statistics Essentials

Means, distributions, and why averages mislead.

◈ interactive5 min read

What is a Model?

A compressed theory of the data it saw.

◈ interactive5 min read

What is AI, actually?

Separating the technology from the marketing.

◈ interactive5 min read

Machine Learning

11 concepts

How machines learn from examples. The ideas every model — old or frontier — is built on.

Backpropagation

How blame gets assigned through a network.

◈ interactive5 min read

Clustering

Finding structure nobody labelled.

◈ interactive5 min read

Evaluation Metrics

Accuracy lies. Here's what to measure instead.

◈ interactive6 min read

Features & Labels

How the world gets turned into columns.

◈ interactive5 min read

GPUs & Compute

Why AI progress is measured in chips and megawatts.

◈ interactive5 min read

Gradient Descent

Rolling downhill toward a better model.

◈ interactive6 min read

Loss Functions

The single number a model tries to shrink.

◈ interactive5 min read

Neural Networks

Layers of simple math that add up to something strange.

◈ interactive6 min read

Overfitting & Underfitting

Memorizing the textbook vs understanding the subject.

◈ interactive6 min read

Regression & Classification

Predicting numbers, predicting categories.

◈ interactive5 min read

Training vs Testing

Why you never grade a model on questions it has seen.

◈ interactive5 min read

Architecture

8 concepts

How the models are built. The shapes underneath everything.

Attention

The single mechanism behind every model since 2017 — and the one that quietly burns most of the compute.

intermediate◈ interactive8 min read

Diffusion Models

How AI learned to make images by starting with pure noise and finding the signal

intermediate◈ interactive8 min read

Embeddings

The coordinates that give language a sense of direction

intermediate◈ interactive7 min read

Mixture of Experts

How modern models pretend to be huge while doing the work of something smaller.

intermediate◈ interactive5 min read

Multimodal Models

When AI learned to see, listen, and read — at the same time, in the same head

intermediate◈ interactive7 min read

Tokenization

The first thing every model does to your words — and the thing that quietly limits what it can do.

◈ interactive6 min read

Transformers

The architecture that changed what AI could do with language — and then everything else

intermediate◈ interactive8 min read

Vision-Language Models

How CLIP and its descendants taught text and images to live in the same coordinate system.

intermediate◈ interactive5 min read

Training

9 concepts

How they learn. What gets baked in and what gets sanded down.

Distillation

Teaching a small model to imitate a big one — and what gets lost in the lesson.

intermediate◈ interactive5 min read

DPO

The cheaper, often-as-good RLHF alternative — and why most labs quietly moved to it.

intermediate◈ interactive6 min read

Fine-Tuning

Teaching a model new habits, not new knowledge

intermediate◈ interactive8 min read

How AI Models Are Trained

From random noise to a model that can reason — the actual pipeline

intermediate◈ interactive10 min read

Model Collapse

What happens when models train on text written by other models — recursively.

intermediate◈ interactive4 min read

Quantization

Why a 70B-parameter model can run on your laptop — and the quality you trade for it.

intermediate◈ interactive5 min read

RLHF

Humans rate, model learns, weird things happen — the post-training that made models pleasant to talk to.

intermediate◈ interactive7 min read

Scaling Laws

Why bigger keeps working — and the question of where it stops.

intermediate◈ interactive7 min read

Synthetic Data

Models training on text other models wrote — and why this isn't always bad.

intermediate◈ interactive5 min read

Inference

8 concepts

How they answer. The mechanics of producing output one token at a time.

Chain-of-Thought

When 'think step by step' actually earns its keep — and when it's just expensive theater.

intermediate◈ interactive6 min read

Context Windows

What the model can see right now — and why the edges matter

◈ interactive6 min read

Function Calling

The JSON-shaped API that turned chat models into clients of the real world.

intermediate◈ interactive6 min read

In-Context Learning

How models 'learn' from examples in the prompt — without changing a single weight.

◈ interactive6 min read

KV Cache

Why long conversations are cheaper than they look — and the reason your API bill behaves the way it does.

intermediate◈ interactive5 min read

Reasoning Models

What changed when models started thinking before they answered.

intermediate◈ interactive6 min read

Structured Output

Forcing the model to fill in a shape — and why it's harder than it looks.

◈ interactive5 min read

Temperature & Sampling

Why 'more creative' is not the same as 'more random' — and the knobs that actually matter.

◈ interactive5 min read

Practical

9 concepts

How to actually use them. The patterns that survive contact with real work.

Agentic Memory

How AI agents remember things across runs — and why most of them don't, really.

intermediate◈ interactive5 min read

AI Agents

When AI stops answering and starts doing — and then, very often, hits a wall

intermediate◈ interactive9 min read

Evals

How you measure whether a model is good at the thing you actually care about.

◈ interactive5 min read

Model Cards

The documentation labs publish when they release a model — and what they leave out.

◈ interactive4 min read

Model Context Protocol

The open standard that lets AI models talk to your tools without a custom integration per model

intermediate◈ interactive7 min read

Multiagent Systems

What happens when AI models try to coordinate with each other — and the new failure modes that come with it

advanced◈ interactive9 min read

Prompt Engineering

The craft of talking to a model that will take you exactly as literally as it decides to

◈ interactive7 min read

Retrieval & Reranking

Why the document your RAG system retrieves first is rarely the document you want.

intermediate◈ interactive5 min read

Retrieval-Augmented Generation

How AI learned to look things up before opening its mouth

intermediate◈ interactive8 min read

Safety

6 concepts

Where they fail, lie, or get tricked. The shape of the things to watch for.

AI Safety & Alignment

The problem of building AI that reliably does what you actually wanted — not what you literally asked for

advanced◈ interactive11 min read

Constitutional AI

When the model judges itself — Anthropic's bet on alignment without exhausting the rater pool.

intermediate◈ interactive6 min read

Hallucination & Grounding

Why AI models confidently make things up — and what you can actually do about it

◈ interactive8 min read

Jailbreaks

How users get aligned models to do what they were trained not to do.

intermediate◈ interactive5 min read

Prompt Injection

The new SQL injection — when input data quietly becomes instructions the model follows.

intermediate◈ interactive5 min read

Watermarking

Invisible signatures on AI-generated text — and why most don't survive contact with reality.

◈ interactive4 min read