Design and Artificial Intelligence: The One Prompting Pro-Tip That Matters

UsefulMBA by John Maeda on YouTubeWatch on YouTube

SaveMark complete

Sign in to save videos and track what you’ve finished.

IntermediateAI2 min

About this video

A short video offering a prompting tip for getting a large language model to hallucinate less by telling it that it is okay to be wrong. The description says the tip works especially well with GPT-4.

What you’ll learn

  • Apply an 'okay to be wrong' instruction in a prompt to reduce hallucinated answers
  • Compare an LLM's answers with and without the tip