CloudYetiFREE
AI tip

Make the AI show which parts of its answer are guesses

An AI sounds just as sure when it is guessing. Ask it to tag each claim by where it came from, then list the claims you should check. Anthropic's guide on reducing hallucinations recommends letting the model say it does not know and checking each claim against a source.

Why it matters

Asking "are you sure?" is not enough: the AI can fold and change a right answer just because you pushed.

Anthropic's guide says to give Claude explicit permission to admit uncertainty, and to make answers auditable by checking each claim against a quote or source.

Tagging each claim turns one confident answer into a short list of what to check.

Do this

  1. Paste the material the answer should come from (notes, a report, a page).
  2. Ask your question as usual.
  3. Add the line below so each claim is tagged: from what you gave it, or from its memory.
  4. Check the listed claims yourself before you use the answer.

A line to add at the end of your request

For each claim, tell me if it comes from what I gave you or from your memory. If you are not sure, say so. Then list the claims I should check before I use this.

One thing to keep in mind: Its own tags can be wrong too. Use the list as a to-do list, not as proof.

The guide is written for Claude; the same line works in other AI chats. Source: Anthropic: Reduce hallucinations ("Allow Claude to say I don't know", "Verify with citations").

Free · by emailGet the next AI tip by emailOne short tip like this one when a new one is ready. Free. Unsubscribe any time.

Ask me anything about AI and cloudA paid call on your own setup: your agents, your bill, your plan.Book a call →

All tips in: Prompt lines to copy

More free AI resources

Send me the tips: free →