AI Hub
Everyone is confused. That is the honest starting line.
AI is on the news, at work, in the family group chat — and nobody was shown how to use it. This is a calm place to build the basics — then learn the part almost no one teaches: how to check what AI gives you.
Free. No sign-up, no account, nothing to install. Works on a phone.
Before any theory
The kind of thing people use it for
Real shapes of use, in plain words. None of them need any technical skill.
- A restaurant owner pastes next week's staff availability — and gets a rota to argue with. how
- A retired teacher pastes a pension letter — and asks what it means, in plain words. how
- A nurse asks what to ask the doctor — before the appointment, not after. how
- A manager turns an hour of meeting notes into decisions, owners and dates. how
- Someone about to repost a statistic checks the source actually exists. how
- A parent rehearses a hard conversation — with the AI playing the other person. how
These are shapes of use, not testimonials — the pages below show exactly how to do each one, and how to check what comes back.
You cannot break anything, and reading this site commits you to nothing — no account, no download. Three things worth knowing before a first try: the big tools are free to start. Don't type anything private. And it sounds confident even when it is wrong — that is what the Trust page is for. When you do try, the first thing to type can simply be: "Explain [anything you have always wondered about] in plain words." A good first time is with someone beside you.
Before anything else
The short version
Six true things. If you read nothing else here, read these.
-
No human is typing the reply.
It is a program in a data centre. Answers vary because a little randomness is built in on purpose, not because you got a different person. That does not make it private: conversations can be stored, sometimes read by staff checking for misuse, and on a work account seen by your employer. What that means for you is on the Trust page.
-
It is not a search engine.
Search finds pages someone already wrote. A chatbot composes an answer word by word instead. Nobody checked it before it reached you — which is why it can be brilliant, and why it can be fluently wrong.
-
It predicts language. It does not "know" things.
It learned which words tend to follow which, from an enormous amount of writing. Being right is a side effect of that, never a guarantee.
Afterwards, people rate its answers, and it is tuned toward the answers people like. That is what makes it helpful and polite — and also keen to agree with you.
-
It sounds equally confident whether it is right or wrong.
There is no wobble in its voice when it invents something — it sounds just as sure when it is wrong. Confidence carries no information. That one fact is why the Trust track exists.
-
It is not a companion, even when it feels like one.
The warmth is a property of the text, not of a mind that cares. It has no stake in your life. Useful, yes — just don't mistake the feeling for a relationship.
-
The skill is not "prompting". It is judgement.
The people getting real value are not the ones with magic words, but the ones who can tell a good answer from a plausible one.
The six tracks
Pick where you are. Nothing here has to be done in order.
Each track is one page. Each page is skimmable in five minutes and expandable when you want the detail.
What AI actually is
Start here if the words themselves are the problem. A companion one-page guide by the same author, on a separate site, that peels AI apart in ten layers with a quiz at the end. Unlike this hub it does name specific products.
Track 1 · The MapWhat it's good at — and where it breaks
You cannot ask for help you can't imagine. A map of what AI is genuinely good at, genuinely bad at, and unpredictable at — plus five habits for finding the gaps yourself.
Track 2 · TrustHow to check the AI's work
The core of the hub. Six verification moves, a triage rule for how hard to check, the failure modes to recognise on sight, and a one-page checklist you can keep.
Track 3 · Get HelpHow to actually get help from it
The mental model that works, the five parts of a good request, ten patterns worth stealing, the beginner mistakes that waste the most time, and the three levels of tool most people never move past.
Track 4 · ToolsWhich model, and what "free" means
Stop trying to memorise the leaderboard. Learn the six categories of tool, a decision flow for picking one, why free tiers end, and how to run your own five-prompt test.
Track 5 · Go FurtherStudy, work, and building something
A four-week plan at 20 minutes a day. How to find AI-shaped work in your job. What this does to careers, honestly. And what it really takes to start an AI company.
The problem, named
You are not behind. The ground is genuinely unclear.
Built from one page of notes on what makes AI confusing right now. If any of these six are yours, you are in the right place — and in very ordinary company.
Noise everywhere, training nowhere
It is on the news, on billboards, in every company meeting — and in nobody's training budget.
Too many teachers, no way to grade them
The tutorials are endless. Some good, some vague, some far too deep — and nothing tells you which before you've spent the hour.
Too many models, no idea which is right
New names every month, each claiming to be the best, several free until a limit nobody explained. Choosing feels like a guess you're not qualified to make.
The mental model is usually wrong
Nobody was handed a working picture of what it is, so people guess: a better search engine, a person typing back, something like a friend. Every guess is sensible, and each produces a different bad habit.
No way to check the answer
The big one. Nobody teaches how to check what comes back — and an answer you can't check is not an answer you can use.
Same starting line, no visible line
Nobody has twenty years of experience with this — young or old, working or retired. The gap is months of practice, not a degree. The problem is finding the start.
Every section below exists to close one of those six gaps. Nothing here needs any technical background, and everything here is meant to be tested against your own experience rather than believed — including the claims on this page.
Ground rules
What this is, and what it isn't
What it is
- Plain language. No jargon without a translation.
- Free, with no account and nothing to install.
- Honest about what nobody knows yet.
- Focused on judgement, not tricks.
What it isn't
- Not a certificate, not a course, not selling you anything.
- Not sponsored by, or ranking, any AI company. Nothing here is paid for or affiliated.
- Not a live leaderboard — model names age in weeks.
- Not a substitute for a doctor, lawyer, or accountant.
This hub was written on 5 August 2026. The ideas here are durable; specific product names, prices and limits are not, and are deliberately kept out of it. Where a fact could go stale, the page says so instead of pretending. If something here contradicts what you can verify yourself today — trust what you verified.