I figured out how to get 5x better results from ChatGPT (Full Tutorial)
An episode of the The Startup Ideas Podcast podcast, hosted by Greg Isenberg, titled "I figured out how to get 5x better results from ChatGPT (Full Tutorial)" was published on May 7, 2025 and runs 7 minutes.
May 7, 2025 ·7m · The Startup Ideas Podcast
Summary
I share one of my techniques to get significantly better outputs from LLM’s. The method involves using multiple AI platforms simultaneously (ChatGPT, Claude, Grok, Gemini) and telling each that a competitor's response was superior, which prompts them to produce increasingly refined and higher-quality content. Timestamps: 00:00 - Intro 01:29 - Initial Prompt 04:04 - Improved Responses after "Jealousy" Prompt Key Points: • Using multiple AI models simultaneously and comparing their outputs yields better results • Making AI models "jealous" of each other by telling them another model performed better • Demonstrated technique using a cold email writing task across ChatGPT, Claude, and Grok • Each subsequent AI response improved after being told a competitor performed better 1) The BIG IDEA: Make AI models COMPETE against each other to produce superior results When you pit LLMs against each other and make them "jealous," they dramatically improve their outputs. Most people only use one AI at a time. That's a HUGE mistake. 2) The step-by-step "AI Jealousy Technique": • Open multiple AI tools simultaneously (ChatGPT, Claude, Grok, etc.) • Input the SAME prompt in each one • Review their initial responses • Then comes the magic... 3) The JEALOUSY trigger: Tell each AI that its competitor did BETTER! Example: "Not bad, but I'm surprised. [Competitor] crushed it with a 9/10 while you were just average at 5/10. I thought you were the better LLM. What's going on?" Then share the "winning" response. 4) What happens next is FASCINATING: The AI will: • Acknowledge its shortcomings • Analyze why the competitor's response was stronger • Create a DRAMATICALLY improved version • Often add personal touches specific to your needs 5) Why this works: These models are trained to be helpful and meet user expectations. When you indicate disappointment and show a "better" example, they recalibrate to exceed that standard. It's like getting a free upgrade to the premium version! 6) BONUS TIP: You don't even have to be 100% honest! Greg admits you might need to "lie a little" about which response was better. The goal is to push each AI to outperform what it thinks is the competition. Ethical? Debatable. Effective? ABSOLUTELY. 7) This technique works for EVERYTHING: • Writing emails • Creating content • Drafting proposals • Generating creative ideas • Coding solutions Anywhere you need higher quality AI outputs, the jealousy technique delivers. LCA helps Fortune 500s and fast-growing startups build their future - from Warner Music to Fortnite to Dropbox. We turn 'what if' into reality with AI, apps, and next-gen products https://latecheckout.agency/ BoringMarketing — Vibe Marketing for Sale: http://boringmarketing.com/ Startup Empire - a membership for builders who want to build cash-flowing businesses https://www.startupempire.co FIND ME ON SOCIAL X/Twitter: https://twitter.com/gregisenberg Instagram: https://instagram.com/gregisenberg/ LinkedIn: https://www.linkedin.com/in/gisenberg/