Two hands-on courses on how AI actually works and how to operate it. Every word is free to read: no account, no payment, no start date. No coding experience needed.
Reading needs no account. An account unlocks the AI voice, the exercises and your saved progress.
↓ scroll
0
Courses
0
Units
Zero
Coding Required
Free
Cost to Read
Try It Right Now
Two minutes of the course, right here.
A paragraph that reads the way every unit reads, and a miniature of the exercise that closes each one. No account, no setup. Judge it yourself.
The reader, live
Hover or tap a dashed term
When you ask a chatbot a question, it doesn't look anything up. It runs The ability to spot repeating trends in data. This is the basic skill behind all AI systems. over everything you typed and predicts the most likely next word, one small piece at a time. Those patterns come from its The huge collection of text, images, or other content that an AI learned from. What's in this data shapes what the AI is good at and what it gets wrong., and where that data has holes, the answers do too. That's why a When AI makes up information that sounds true but isn't, like fake facts or made-up sources. It's guessing what sounds right, not checking if it's true. arrives in the same confident tone as a fact: it was never checking, only predicting.
Every unit reads like this: definitions where you need them, no account required.
Sample exercise · Spot the Error
Three of these six sentences are wrong
Tap the ones you don't trust. This sample is fixed. At the end of every unit you do this on a fresh AI-written paragraph about what you just read.
Most AI courses explain concepts. ThinkModel builds a mental framework: for understanding AI, using it with intention, and keeping your own perspective in a world shaped by it.
Understand the system
Pattern recognition, training, prediction: the real mechanics, not the hype. Explained so clearly you could teach someone else.
Use it with intention
Learn context engineering, the skill that separates people who get generic AI output from people who get exactly what they need.
Judge it critically
Spot hallucinations, evaluate output, and catch errors that most people miss. Become the judge, not the audience.
Build real things
Using tools like Replit, Claude, and v0, you'll create a working project (an app, a tool, a site) that you can show to anyone.
Know your tools
Build a personal AI toolkit and learn to evaluate new tools as they launch, so you stay current long after the program ends.
Keep your perspective
AI can produce, but it can't originate. You'll develop the point of view that makes AI amplify you, not replace you.
The ThinkModel Philosophy
People don't need a list of tools. They need a model for thinking.
Understand the system. Use it with intention. Judge it critically. Build real things. Keep your own perspective.
The Curriculum
Two courses. 24 units. Every word free to read.
101 teaches you what AI is actually doing. 102 teaches you to operate it: workspaces, agents, automation, your own stack. Start at 101; go to 102 when you're ready.
01Understand it
02Talk to it well
03Evaluate its work
04Watch it act
05Build with it
06Delegate to it
07Automate it
08Own your stack
ThinkModel 101
Fundamentals
12 units · Start here
How AI actually works, why it fails, and how to get good at using it, from the prediction machine to building something real.
Stop chatting and start operating: configure workspaces, pick the right model, hand work to agents, and run systems that keep working when you're not watching.
The platform uses AI to personalize your learning, making every unit adaptive, interactive, and impossible to passively skim.
The Read Layer
Listen, explore, understand
The prose is open to everyone: hover any term for an instant definition, no account required. Sign in and every paragraph becomes listenable in a natural AI voice, any passage you highlight gets explained in simpler words, and one key concept per unit is rewritten using YOUR interests as the metaphor: basketball, cooking, music, whatever you chose.
OpenAI gpt-audio · Claude Opus 5
The Think Layer
Answer before you read
Turn on Socratic Mode and AI asks you probing questions between sections, so you have to reason through an answer before the next section reveals. Six interactive sandboxes across the two courses let you train a tiny model, engineer context, price a workload, route a task to the right model, look inside a harness, and run an agent workflow until it breaks.
Claude Opus 5 · Interactive simulations
The Prove Layer
Show you understand
After each unit, record yourself explaining the concept in 60 seconds, and AI evaluates your accuracy, clarity, and completeness. Then spot 3 subtle errors hidden in an AI-generated paragraph. Do both and the unit closes on a finale; skip for now if you'd rather keep reading, and come back to it whenever you like.
OpenAI gpt-4o-mini-transcribe · Claude Opus 5
Every feature in the platform is a demonstration of the concepts being taught. The medium is the message.
By the End of ThinkModel
You'll be able to…
Explain how AI works to anyone: clearly, accurately, without jargon
Use context engineering to get dramatically better results from any AI tool
Spot AI hallucinations, bias, and errors that most people miss
Design human-AI workflows that know when to delegate and when to stay hands-on
Build and ship a real working project using AI as a collaborator
Hand real work to an agent and know exactly where the human checkpoint belongs
Automate something you keep redoing by hand, and notice when it fails silently
Evaluate any new AI tool in 15 minutes and decide if it's worth your time
Think critically about who builds AI, who controls it, and what that means for you
Finish a capstone project you built, can explain, and are proud of
How It Works
One loop, one unit at a time.
No start date, no cohort to wait for. The site is the whole course: read a unit, prove you got it, move on when you're ready.
Read the unit
Every word is free, no account. About 7–13 minutes of reading, plus however long you spend on the live demo, the sandbox and the term definitions. With an account your place is saved as you read, and a progress bar shows how far through you are. Close the tab and pick it up on any device.
Prove you got it
Teach the concept back in 60 seconds and find the errors hidden in an AI-written paragraph. Do both and the unit ends on a finale that unlocks the next one, or skip for now, keep moving, and come back to it later.
Talk it through (optional)
When a cohort is running, live group sessions run alongside the reading: discussion, group demos, and the challenge done together. Nothing on the site waits for one. If you're reading on your own, the course is complete as it stands.
Frequently Asked Questions
Things people ask before they start.
Every word of both courses is free to read with no account: all 24 units, start to finish. An account is only needed for the tools that cost us money to run: the AI voice, the explain-this-simpler feature, personalised analogies, the Prove It exercises, and saved progress.
Start with 101 – Fundamentals if you want to understand what AI is actually doing. Go to 102 – Operator when you want to run it as a system: workspaces, model choice, agents, automation. 102 opens with an optional Start Here primer, so you can see what it expects before committing.
No programming, ever — you never write code. 101 needs no technical background at all: when you build a project in Unit 10 you'll use tools like Replit that turn plain language into working software. 102 asks a bit more of you: still no code, but you will open a terminal, run a few install commands and work with files on your own machine. The skill you need is clear thinking, not programming.
Anyone who uses AI but doesn't really understand how it works, or knows they could be using it much better. The content is written to work equally well for teenagers (13+) and non-technical adults. No dumbing down, no jargon.
As long as you want it to: it's self-paced, with no start date and no schedule. A unit is about 7–13 minutes of reading, counted from the words on the page — the same estimate each unit publishes to search engines. Allow another 20–40 minutes for the demos, sandboxes and Prove It exercises. Your place is saved as you go, so you can read one unit a month or the whole course in a weekend.
Free tutorials teach you how to click buttons. ThinkModel builds a mental framework: you'll understand why AI works the way it does, not just how to use one tool. Across two courses you'll also learn context engineering, evaluation, economics, agents, automation and how to build real projects, with the platform itself demonstrating what it teaches.
101 is browser-only: a laptop or tablet, internet, and the free tiers of Claude, ChatGPT, Replit and Teachable Machine. 102 asks more: Unit 04 installs a command-line agent and initialises git in a terminal, and Unit 10 runs a model on your own machine via LM Studio or Ollama (a multi-gigabyte download, 8GB+ RAM). Each has a free path: Gemini CLI where Claude Code wants a paid plan, and a browser option if your machine can't run a local model.
A capstone project: a working app, website, tool, or creative project that you build using AI as a collaborator. It's your choice what to make. Past ideas include personal budget trackers, study tools, recommendation engines, and automated workflows. In 102 you go further and build systems that keep running without you.
Context engineering is how you give AI the right information to produce great output: who you are, what you need, what good looks like, and what to avoid. It's the skill that makes the difference between generic AI responses and exactly what you wanted. Most people have never heard of it. Unit 07 of 101 teaches it in depth, and 102 builds entire workspaces on it.
The courses stand on their own. You can read and complete everything by yourself, whenever you like. When a cohort is running, live group sessions run alongside the reading for discussion, group demos and shared challenges. They're an accompaniment, not a requirement, and nothing on the site waits for one.
The tools will change. The thinking won't. ThinkModel teaches a framework (pattern recognition, context engineering, evaluation, building) that applies regardless of which specific AI tools are dominant. Unit 11 of 101 teaches you how to evaluate new tools, and the last unit of 102 is about staying current on your own. The goal is a mindset that lasts, not a tutorial that expires.
The credential is the capstone project: a real, working thing you built that you can show to anyone and explain. That demonstrates more than any certificate ever could. A project someone can use beats a piece of paper someone can frame.
ThinkModel
A new model for thinking in the age of AI.
Two courses, 24 units, free to read. No coding required. Start whenever you like.