If your goal is to understand what AI is and is not
Take these in order. Fifteen minutes gets you step one; forty-five gets you the first two; two hours gets all three.
AI literacy depends on who is using the system, for what purpose, at what level of risk, and who may be affected. Build a route below or use the curated starting points. No email, account, or server-side profile.
Choose your context, goal, stakes, and time. The result is deterministic, stays in your browser, and routes you through the existing curriculum with a human-review rule suited to the work.
The builder above is a convenience. Nothing is hidden behind it — here is every route it can produce, and every human-review rule it can apply, so the page is complete whether or not scripting runs.
Take these in order. Fifteen minutes gets you step one; forty-five gets you the first two; two hours gets all three.
Take these in order. Fifteen minutes gets you step one; forty-five gets you the first two; two hours gets all three.
Take these in order. Fifteen minutes gets you step one; forty-five gets you the first two; two hours gets all three.
Take these in order. Fifteen minutes gets you step one; forty-five gets you the first two; two hours gets all three.
Take these in order. Fifteen minutes gets you step one; forty-five gets you the first two; two hours gets all three.
Pick the one that matches the consequence of being wrong, not the one that matches how confident the output sounds. This is the single judgement the rest of the site keeps returning to.
Low stakes still require a quick factual check. Do not paste confidential, personal, privileged, or licensed material into a tool without authorization.
Use a named human reviewer, preserve important sources, and test the output against a known example before relying on it.
Do not treat model output as the decision. Require qualified human review, source verification, documented approval, and a non-AI fallback or escalation path.
Human-centred mindset, ethics, foundations and applications, and responsible design—not tool tricks alone.
Govern, map, measure, and manage risk in context, with testing and documentation proportional to impact.
Learning should reflect technical knowledge, experience, context of use, and the people affected by the system.
Framework references: UNESCO AI Competency Framework · NIST Generative AI Profile · European Commission AI Literacy FAQ.
You want to understand what generative AI actually is and where it came from, without hype or jargon.
You already use ChatGPT or Claude sometimes and want to get noticeably better at prompting and applying it.
You're choosing between ChatGPT, Claude, Gemini, and Perplexity, or trying to build a lean toolkit.
You want the real 2026 numbers and the honest, unresolved debate over artificial general intelligence.
You want to verify a claim, or you're looking for a specific free resource — official docs, papers, or tutorials.
Skim the Frontier & AGI stat grid, or click one entry on the History timeline.
Read one full tab on Techniques, or one category on Use Cases, start to finish.
Read front to back: History → Concepts → Techniques → Use Cases → Tools → Templates → Frontier & AGI.
Turing, Dartmouth, two winters, and the transformer — the sourced history, in full.