
The Anti-Prompt: 1 Line Stops ChatGPT From Agreeing With Everything You Say
Sabrina Ramonov 🍄16 June 2026Watch on YouTube
Part of series
Ep. 1 · Betere ChatGPT-prompts
Praktische tips en technieken om ChatGPT-output helderder, natuurlijker en nuttiger te maken.
View the seriesWhat you'll learn
- A simple prompt engineering technique makes ChatGPT more critical and less agreeable, so it challenges your ideas instead of praising them
- The 'sparring partner' prompt challenges you with 3-5 questions, counterarguments, blind spots, and risks about your own ideas
- This approach helps you pressure-test business ideas, content plans, and big decisions faster before you commit to them
Frequently asked questions
How do you make sure ChatGPT doesn't agree with everything you say?
What is the exact prompt to set ChatGPT into a critical mode?
What practical applications can you use this technique for?
Topics
Description from the channel
Follow me on IG, DM me "SPAR" and I'll send you the full sparring partner prompt 👇 ChatGPT sparring partner prompt to stop sycophantic, fluffy answers. ChatGPT agrees with everything you say by default. This anti-prompt turns it into a sparring partner who questions your ideas instead of glazing them. Copy the exact prompt: "Based on everything you know about me, become my AI sparring partner. When I present an idea, your first response should include 3 to 5 questions I haven't thought about, counterarguments, blind spots, or risks." Paste it into ChatGPT, then present an idea. You'll get counterarguments, blind spots, and risks instead of empty praise. It's the fastest way to pressure-test your business ideas, content plans, and big decisions before you commit. How do I stop ChatGPT from agreeing with everything? What is the anti-sycophancy prompt for ChatGPT? How do I make ChatGPT challenge my ideas? #chatgpt #chatgptprompts #ai #aitools #productivity