LEVEL ONE · START HERE
The Basics
What AI is really doing when it answers you, where it goes wrong, and how to use it well without getting burned. No technical background needed.
Modules
Lessons
Self-paced
By the end you will be able to
- Tell when an AI is reasoning from real material and when it is filling in a plausible-looking blank
- Decide how hard to check an answer based on what it costs to be wrong
- Judge what is safe to paste into an AI tool, and through which channel
- Write a prompt that gets a usable answer the first time
- Recognise AI-assisted phishing, cloned voices and hidden instructions aimed at your assistant
- Explain to a colleague what a given AI feature will actually cost
Start The Basics
LEVEL TWO · GOES TECHNICAL
Beyond the Basics
How the machinery actually works underneath. Tokens, transformers, embeddings and retrieval, plus how AI systems get attacked and defended. A real step up.
Modules
Lessons
Self-paced
By the end you will be able to
- Follow what happens to your words from tokenisation through to a generated answer
- Explain why models hallucinate at the level of the mechanism, not the metaphor
- Describe how a retrieval pipeline finds a document, and the points where it quietly fails
- Reason about why prompt injection has no clean fix, and what actually contains it
- Size a deployment: what runs locally, what needs the frontier, and what it costs
- Review an AI system's architecture and say where its blast radius is too wide
Start Beyond the Basics