AI for people

This week a president renamed artificial intelligence by decree, and a domain gold rush followed. Every label on the table — super, master, companion — is a claim about where the machine sits in relation to you. You don't have to accept the one you're handed.

The week's headlines are all about whether AI ends us. That's the dramatic question — and the one that lets us off the hook. The nearer question, the one we can still answer, is what deal we're making with these machines: what stays ours, what we hand over, and what still has to be earned.

When a message knows your email, an old password, or where you shop, it feels like proof it's genuine. It isn't. Where that information really comes from — and what actually keeps you safe.

AI can make almost anything look real, so 'it looks real' has stopped being proof. A few seconds of 'where did this come from?' protects you better than trying to spot the fakery.

AI adverts want you to feel something about a company. Here's a simple test that tells you whether there's anything real underneath the good feeling.

A convincing email is easy to fake. The one thing the scammer can't hide is where the link actually goes — and reading it is a ten-second skill anyone can learn.

An AI chatbot will state a complete fiction in the same confident voice it uses for the truth. Here's why that happens, and a ten-second habit that keeps you from being caught out.

The single skill that makes an AI assistant actually useful — giving it enough context before you ask.

You're not imagining it — the web feels smaller than it is. How to break the habit, use AI search to find what doesn't rank, and put yourself back in charge of what you read.