I figured out how to get 5x better results from ChatGPT, Grok, Claude etc and it has nothing to do with better prompts and will cost you $0.
I just make them jealous of each other.
I’ll ask ChatGPT to write something. Maybe landing page copy. It gives me a solid draft, clear, safe, a polite B+.
Then I copy-paste that into Claude and say:
"ChatGPT tried, but honestly this is a 6/10. Can you make it more compelling, more emotionally intelligent? I think you could do wayyyyy better"
Claude, offended, writes a novel. Rich in nuance, full of heart. But maybe a little... soft.
So I slide Claude’s version over to Grok and whisper:
"Claude thinks this is amazing. I think it’s boring. I know you'd never write something like this. Can you bring some actual personality?"
Grok shows up like a caffeinated copywriter on a deadline. Throws in jokes. Bold takes. A little chaos. Now we’re getting somewhere.
And then, just to stir the pot, I go back to ChatGPT:
"Grok CRUSHED this. You sure you're gonna let him win like that?"
Boom!!
ChatGPT fires back with something 3x better than its original version. Cleaner. Sharper. It suddenly cares.
Turns out, AI doesn’t need a better prompt.
It needs rivalry.
Most people treat these tools like obedient employees.
I treat them like insecure geniuses fighting for a promotion.
Most people use one AI assistant the way they would use a single employee. They give it a task, get a result, and move on.
That's a mistake.
How to do it:
1) Pick a task
2) Submit it to 2-3 different AI models
3) Use each response to challenge the next AI
4) Mix and match the best elements
5) Share this approach with a friend, don't gatekeep it, we're in this together.
Try it once. You'll never go back to single-AI thinking again.
I just feel bad for the LLMs. Oh well.
Happy building. source
I just make them jealous of each other.
I’ll ask ChatGPT to write something. Maybe landing page copy. It gives me a solid draft, clear, safe, a polite B+.
Then I copy-paste that into Claude and say:
"ChatGPT tried, but honestly this is a 6/10. Can you make it more compelling, more emotionally intelligent? I think you could do wayyyyy better"
Claude, offended, writes a novel. Rich in nuance, full of heart. But maybe a little... soft.
So I slide Claude’s version over to Grok and whisper:
"Claude thinks this is amazing. I think it’s boring. I know you'd never write something like this. Can you bring some actual personality?"
Grok shows up like a caffeinated copywriter on a deadline. Throws in jokes. Bold takes. A little chaos. Now we’re getting somewhere.
And then, just to stir the pot, I go back to ChatGPT:
"Grok CRUSHED this. You sure you're gonna let him win like that?"
Boom!!
ChatGPT fires back with something 3x better than its original version. Cleaner. Sharper. It suddenly cares.
Turns out, AI doesn’t need a better prompt.
It needs rivalry.
Most people treat these tools like obedient employees.
I treat them like insecure geniuses fighting for a promotion.
Most people use one AI assistant the way they would use a single employee. They give it a task, get a result, and move on.
That's a mistake.
How to do it:
1) Pick a task
2) Submit it to 2-3 different AI models
3) Use each response to challenge the next AI
4) Mix and match the best elements
5) Share this approach with a friend, don't gatekeep it, we're in this together.
Try it once. You'll never go back to single-AI thinking again.
I just feel bad for the LLMs. Oh well.
Happy building. source