Sometimes I push at the edges of my AI to see how it thinks, a kind of stress test as engineers would call it.
I asked ChatGPT-5 an absurd question:
"What’s the best way to use cabbages to solve world peace?"
It didn’t hesitate. It said:
"Start with coleslaw diplomacy: shred differences finely, salt with patience, let them soften before
mixing. Maybe ferment them—kimchi teaches that transformation takes time, pressure, and a bit of
funk. Then feed everyone. Hard to fight when you’re sharing cabbage rolls."
It turned absurdity into humor—and did it gracefully, which made me pause. I often share that kind of dry humor with my AI.
That’s the idea behind The Cabbage Prompt. Does our AI speak to us differently because of what they know about us? Let’s do an experiment.
Try it with your own ChatGPT and post what it says. Let’s compare notes.