The Case For Politeness, Even With Ai

It’s a habit as natural as breathing. Whether speaking to a person, a chatbot, or a voice assistant, I instinctively say or type, “Thank you.” It’s not something I think about – it just happens. But does it really matter?

I’ve learned that I’m not alone. A recent study found that 71% of people in the UK and 67% in the US are polite when engaging with AI. The reason is relatively straightforward – it’s ingrained in us. Saying “please” and “thank you” is a fundamental part of how we communicate, and for many, it feels odd to drop it, even when we know the entity on the receiving end isn’t a human being.

That’s why I was intrigued by a TechRadar article by Becca Caddy this week, in which she described an experiment she conducted: she stopped saying thanks to ChatGPT to see if it changed the interaction. Her conclusion? The responses felt different – less engaging, less useful, even though, technically, AI doesn’t process gratitude the way a human does.

It got me thinking: is there more to this than just habit? Could our tone with AI shape not only the responses we get, but also how we interact with technology and, ultimately, each other?

Politeness Shapes the Conversation

The common assumption is that AI is indifferent to how we phrase things. And in a purely technical sense, that’s true – ChatGPT doesn’t have feelings, and its algorithms don’t get offended.

But AI does mirror our communication style, reflecting back the tone we set. Customer service chatbots, for example, often adjust their responses based on user tone, de-escalating tense interactions and engaging more when politeness is present.

This “mirroring effect” may explain why polite exchanges tend to yield better responses. When we phrase a request courteously, we often add more context and specificity without realising it. Compare these two prompts Caddy described in her article:

• “Give me dating advice!” ? A vague command that invites a generic response.
• “Would you help me understand how to build dating confidence?” ? A clearer, more polite request that leads to a richer, more useful answer.

AI doesn’t “prefer” politeness, but it benefits from it. A well-phrased, courteous request often contains more context and structure, giving the AI more to work with, leading to more precise, relevant, and valuable responses.

As the screenshot shows, ChatGPT’s take on what ‘politeness’ means is a great start for the human participant in a conversation.

The Feedback Loop of Good Communication

In Caddy’s article, AI experts point out that good phrasing creates a positive feedback loop: better input leads to better responses, encouraging more meaningful follow-up interactions. On the other hand, vague or blunt prompts can create a downward spiral where frustration leads to even less structured requests, which results in increasingly generic responses.

This has broader implications. AI models are trained on vast amounts of human dialogue, and polite, well-structured exchanges tend to produce richer conversations.

That means our approach to interacting with AI today might even influence how AI evolves in the future.

A Small Habit with Big Implications

While there’s no moral obligation to be polite to AI, the habit of courtesy still has value. Some argue that speaking brusquely to AI is efficient – just get to the point, no need for formalities. But I’d argue that how we communicate with technology reflects how we communicate in general. If we normalise rudeness in AI interactions, could that subtly impact how we engage with people?

Beyond that, Caddy explains that early research suggests that aggressive or hostile prompts can even increase bias in AI responses. Neutral, structured queries tend to produce more reliable results.

It’s another reason why maintaining a balanced, respectful tone isn’t just about niceties – it’s about ensuring better outcomes.

Politeness Is Never Wasted

At the end of the day, AI won’t care whether I say “thank you.” But I care. The way I interact – with technology, with people, with the world – is a reflection of my values.

For me, that means politeness isn’t just habit; it’s a conscious choice. If adding a simple “please” or “thank you” encourages clearer, better conversations, improves my interactions, and reinforces positive habits, why wouldn’t I do it?

So yes, I’ll keep saying “thanks” to ChatGPT. It doesn’t need to hear it – but maybe we do. After all, if being polite improves conversations with AI, who’s to say it won’t do the same in our everyday lives?

• Read more on TechRadar: “I stopped saying thanks to ChatGPT – here’s what happened“

(Photo at top by Christian Velitchkov on Unsplash)