Scott Frothingham posted a collection of 15 sayings from Yogi Berra on LinkedIn. They’re funny, a little confusing, and insightful.
That made me wonder, “Can AI do this?” So I asked ChatGPT to give me ten Yogi Berra-ish sayings.
The results were decent. Some were clever. But they weren’t great. Not like the originals.
Why not?
AI can certainly mimic the syntax and style. It can learn the patterns that make a quote sound like Yogi. But what it can’t do — at least not well yet — is figure out which lines hit and which ones flop. It doesn’t know which phrases people will actually laugh at or remember.
That’s because the best Yogi Berra quotes weren’t just clever. They were popular. They resonated with people. Yogi probably said all kinds of odd things — but only a few became legendary. The crowd decided what stuck, and those are the sayings that made Scott’s list.
This is a key insight: AI can replicate style, but it struggles to replicate impact without human feedback.
That’s no slander on AI. Even genius comics have to try out their jokes on an audience to see what works, how to time the lines, etc.
The next frontier in AI training is to give these systems similar feedback. They’ll need more and better human reactions to learn what works — that is, what’s funny, what’s wise, and what’s emotionally effective. AI gets some of that feedback right now. You can click a thumbs-up or thumbs-down, and sometimes it asks you to choose between two responses. But that’s primitive compared to what’s coming.
I suspect the future involves more passive, real-time feedback — like emotional responses captured by wearables. Smartwatches and similar devices can already detect heart rate, skin conductivity, even micro-expressions. Imagine AI getting subtle, immediate signals about how you actually feel about what it just said.
Yes, I know. That’s more than a little creepy. But it’s also the logical next step if we want AI that can truly understand and influence human emotion. The next generation of AI won’t just be about smarter algorithms. It’ll be about tuning into us more effectively.
That’s something to watch. Or, if you can, to experiment with.
