The English saying goes “fake it till you make it”. Start by pretending you know what you’re doing. Keep at it long enough, and eventually you actually will.
With AI everywhere now, the temptation is obvious. Let the machine do the work, get the result, move on. But the old saying doesn’t hold anymore. Now it’s “fake it and you will never make it.”
Because when we outsource our thinking to AI — which always sounds so bloody confident — we stop developing our own capabilities. Our skills atrophy. We lose what makes us human and turn into what I can only describe as “deep-fried eggplants.”
I get it though. Everyone else is using AI to work faster and better. You can’t just ignore it.
The middle path might be this: don’t hand everything over completely. Get AI to suggest improvements, then work through them deliberately. Make the changes yourself, one by one. It’s more work, but you’re learning, not just copying.
I run my writing through ChatGPT with a detailed prompt. But I never just accept what it spits back. Instead, I ask for a list of suggested improvements with explanations — why this change, what it achieves. From that, I learn better writing and clearer thinking.
Then I work through the list. Keep what makes sense, ignore what doesn’t, often find an even better way to phrase something myself. The AI does the heavy lifting on analysis, but I don’t turn my brain to mush in the process.
That has to be our standard. We keep control. Because if we just push text through, tell ChatGPT to “make this better” without even reading the result because we trust it so completely — that’s massive loss of control. And it backfires spectacularly.
First published in German at reinergaertner.de, where I’ve been at it since 1997. AI did the heavy lifting on the translation. I did the heavy squinting at the result.