5 questions before trusting an AI answer

Using AI day to day is like using a map: it helps a lot, but you are still driving. Before following an answer, run through these five questions.

1. Is this serious?

Cake recipe, movie suggestion, e-mail draft: trust it and move on. Medicine, money, contract, diagnosis: AI helps you understand, but does not decide.

2. Where did it get this?

Ask for the source. Open the link. If the source does not exist or does not say that, throw the answer away.

3. Did I say too much?

Do not put in the chat what you would not put on a poster in the street: ID numbers, passwords, medical results, a child's address. What goes in may be kept.

4. Does it make sense to a person?

Read it aloud. If it sounds strange, it probably is. Your intuition still counts.

5. If it goes wrong, who answers?

If the answer is yours, you answer. That is why the last word has to be yours.

Keep this

AI is a co-pilot. A map, not the driver.

Want the version to print and stick on the fridge? Subscribe to the newsletter and reply to the first e-mail with "fridge". We will send it.