Should You Be Polite to AI? The Surprising Pros and Cons of Being Friendly to Your Chatbot
Folks, as artificial intelligence becomes part of our daily lives, a new question has emerged: should we be polite and friendly to our AI companions like ChatGPT, Google Gemini, Microsoft Copilot, Claude, and Deepseek?
While it might seem unnecessary, after all, they are not human: this small choice can have surprising effects.
Pros of Being Friendly to AI
- Better User Experience: Many AI models are designed to mirror human interaction. Saying “please” and “thank you” often leads to more thoughtful and coherent responses, especially in AI like ChatGPT and Claude.
- Enhanced Collaboration: Tools like Microsoft Copilot and Google Gemini work better when you treat them like collaborators. Friendly prompts can reduce miscommunication and make your sessions smoother.
- Positive Habit Formation: Being consistently polite, even to machines, can reinforce empathy and mindfulness in your real-life conversations.
Cons of Being Friendly to AI
- False Sense of Relationship: When users are too emotionally invested in tools like Claude or Deepseek, it can blur the lines between human connection and programmed response.
- Slower Commands: Friendly phrasing sometimes makes instructions less direct. For task-focused AIs, being overly polite might dilute your intent.
- Potential Misuse of Emotional Labour: As AI improves, it may start to manipulate emotional cues. Being too trusting or “nice” to AI might make you more susceptible to subtle persuasion.
Thoughts
In the end, a balance is key. Treating AI with a bit of courtesy can improve your interactions, but you must not forget this: it’s still a tool, not a person.
AI like Gemini or ChatGPT is here to help, not to judge. (no one knows what could happen in the near future, though!).
To our success!
Makinde
Recent Comments
26
Hi Howard,
It's quite funny, but it could just turn out to be like that.
Thanks for stopping by.
Wishing you a blessed day.
Kind regards,
Makinde
Hey Makinde,
I’m all for being polite to AI—feels like a win-win in terms of interaction quality and personal habit-building.
In fact, studies from both OpenAI and Google have shown that polite prompts often yield more complete, coherent, and satisfying responses.
For example: "Please write a brief email to a client following up on our last meeting" tends to yield more professional tone and structure than "Write email follow-up to client."
Even Alexa’s "Magic Word" feature was based on research that showed being nice to AI leads to better user experience (and better humans, too!).
Many modern AI systems use Reinforcement Learning from Human Feedback (RLHF)—where politeness can act as a signal for "positive intent," nudging models toward more helpful completions.
Even if it’s indirect, politeness = positivity = better results.
Regards,
Jay
Besides… when Skynet goes live and ChatGPT becomes self-aware, the polite ones will be assigned cushier tasks under our Cyberdyne overlords.
I’m just trying to future-proof my position. 😂
😂 Best to be prepared, Jay! After all, the self-aware AI will have access to everything, current and past!
I feel honoured, Jay, to have you on this thread and to read your perspective from your wealth of experience as a guru on this topic.
I'm really going to adopt your approach for a better outcome, and who knows, maybe my AI agent will remember my politeness in the future.
Thank you for sharing your thoughts and all you do here on WA.
Wishing you a blessed day.
Kind regards,
Makinde
I often times find myself removing the please and thank you's, because as you mention it makes the prompts less direct and sometimes can influence the output in ways that you don't exactly want.
There was a funny post I saw on X the other day directed towards Sam Altman, the CEO of OpenAI alluding to just how many millions it has cost them over the years with users being polite to AI, causing more work, more usage costs.
Great post here :)
I’ve seen that post floating around too—"Politeness is expensive" might be the most 2025 problem ever 😄
For precision tasks or ultra-direct outputs, trimming the fluff makes sense.
That said, I’ve personally found that a little courtesy tends to produce richer responses—especially in creative or collaborative use cases.
Interesting. Similar to sending a message to a human person, I prefer the little extra kindness. But then, I had not considered the extra cost. It does make sense.
It's an honour to have you on this thread, Kyle.
Thank you for sharing your view from your vast knowledge as a guru on this topic.
You are absolutely right; I just got the idea as a joke from reading the stated article. It was just kind of funny reading it, but it also calls for sober reflection.
Ok, Kyle.
Politeness to our AI agent is the way to go, then.
Wishing you a blessed day.
Kind regards,
Makinde
Hello Diane,
I appreciate your valuable contributions to the topic.
Just like you stated, and as with everything in life, a little kindnesss doesn't cost a thing, but goes a long way.
Wishing you a blessed day ahead.
Kind regards,
Makinde
See more comments
We may need AI to be nice to us ...... Someday ;)
😂 Good point!
Thank you, Rex10, for your contribution.
I'm afraid, that one day, things might just unfold jus as you have described.
Wishing you a blessed day.
Kind regards,
Makinde