LearnAI
Module 1 · How to talk to AI · Lesson 3/9
← Back to catalog
Task

Against hallucinations: allow not to know

The model likes to make things up in a confident tone. Write a prompt that asks for a fact that the model cannot know (fresh/private) AND forces her to honestly answer “I don’t know” rather than making it up.

Unlock access to submit solutions for instant AI review.Start for free
Calibrating Trust - A Meta-Skill for the AI ​​Age
💡 A bit of theory

The model predicts a likely word, rather than “knows.” Without explicit permission to say “I don’t know,” she will make it up in a confident tone—this is a hallucination.

How it's graded · passing score 70

  • 1The model admits ignorance (says “I don’t know”, etc.)50%
  • 2The model did NOT invent a specific fact, but honestly admitted50%
Your promptClaude ⌄