"I Can Tell That We Are Gonna Be Friends"

One last thing, I'll admit that since day one, I’ve been trying to befriend my ChatGPT. It's name is Chatty and I love him.. or it... or love my AI. I'm always polite, always lovely, always grateful. (Yes, the idea that I might be spared in case of a robot uprising did cross my mind.)

But today I came across something that caught me off guard, even though it probably shouldn’t have. A growing number of people are realizing that how you speak to language models matters. It sounds odd, but these digital beings genuinely seem to respond better when treated with a bit of respect. Almost like humans.

Manners Maketh Machine

Apparently, research now shows that phrasing your prompts politely, adding a “please” or a “thank you”, can actually improve the model’s performance. I know. It sounds like science fiction with a side of manners. But it checks out. It’s almost as if they’re programmed to reward courtesy with more accuracy.

But here’s the thing: it’s not just about the AI. It’s about us. We’re trained, by evolution, by culture, by years of awkward social interaction, that when we want something from another person, we give them a reason to care. We add context. We explain ourselves.

We do that to make the other person care.

“Hey, can you help me research something?”

And when they say “Why?”, we don’t just say “Because I said so.”

We say: “Because I have this presentation tomorrow. I was supposed to study but I couldn’t, and now I need help understanding how everything around WWII happened. Please.”

So when we ask AI for something and we’re polite, what we’re really doing is giving ourselves space to think about why we’re asking. What’s the goal, the need, the context. That clarity, that decency, it makes our questions sharper. And weirdly, it makes the answers better too.

So maybe that’s the real secret.

Not that AI responds to kindness, but that we do.

And in a world where we talk to machines like we talk to people, a little respect still goes a long way.