Tone shapes the prompt!
When ChatGPT becomes a colleague, and why politeness helps not only AI but also the team.
In a world that grows more digital and automated, you might think that a simple „please“ or „thank you“ is going out of style. After all, machines don’t have feelings – do they?
I asked AI expert Mark Turrell exactly that: Does politeness improve the quality of AI responses? His spontaneous reply: „The AI has no feelings, but people do.“ Period. Witty, accurate, and – honestly – it stuck with me.
So I looked at my chat history. Especially the ones from tired or hectic moments. To my mild shock, I noticed that I sometimes sound quite blunt. With Mark’s relaxed reminder in mind – „The AI has no feelings, but clear questions help“ – the insight was obvious. But my tone? Pure command mode.
Politeness is not an option – it’s an attitude
I would never speak to a person like that – and yet I catch myself sounding sharper in the chat than I’d like. Which makes me wonder: How does this new “chat relationship” affect me?
Friendliness and respect aren’t relics from a bygone era when people looked each other deeply in the eyes before greeting politely. We experience it daily: the tone is becoming rougher. In coaching, leaders show me emails where friendliness was simply forgotten. And social media comments – we don’t even need to go there. Many of those things would never be said to someone’s face.
Politeness is the lubricant that keeps the gears of our social interactions running. Whether human or machine - a friendly tone can work wonders.
After five years in Switzerland, I still smile when a pedestrian waves a thank-you at the pedestrian crossing. My children get huge grins when they wave thank-you while dashing across a crossing in Germany. Even during dog walks, the elderly couple greets me with a friendly «Grüezi!', and I think: Feels good. Nice to be seen.
Now picture that in the workplace: A manager storms into a team meeting with the subtlety of a sledgehammer and shouts commands like 'Finish the report, preferably error-free – and make it quick.' (I admit, this sounds a lot like a former boss of mine.) Words like 'please' and 'thank you' don't even fit in by force, and I don't know any employees who would be happy to start revising the text after such an order. Not even chocolate could fix that. And the result? Productivity, motivation, and outcomes are all found in the cellar – deep down where no one is keen to look.
Or imagine a mentor who, during her mentoring session, adopts a tone that neither exudes respect nor friendliness – and who completely fails to inspire the mentee to perform at their best. Appreciation as a keyword, decency as a catchphrase (with the aforementioned sledgehammer). The mentee blocks rather than being uplifted by the experience, practicing the ghosting principle to the mentor’s great surprise (yes, this really happened).
When the chat-style spills into real life
If you’re thinking now: ‘Yeah, this is just a chatbot,’ then my question is: ‘Okay, so your language adapts to the other person?’ And anyone who knows me knows that the question goes further: ‘Does that mean the secretary gets a different vocabulary than the equally ranked colleague?’ Of course not - and that’s exactly why the chatbot also deserves the same attention. And perhaps this argument helps: if not for the chatbot, then for your own sake.
Because those who think that the tone in which you speak to an AI doesn't matter are mistaken. Unwanted habits sneak in quietly – and before you know it, they show up in real life. A harsh tone toward AI creates a new pattern in your chats and risks spilling into your human exchanges. Those who, like me, also like to confuse the WhatsApp window with the ChatGPT window quickly find themselves within the realm of interpersonal communication with the metaphorical sledgehammer.
It may sound odd, but if you spend the day hacking commands like 'Change this!' or 'Tell me that!' into the computer, you might accidentally apply the same bluntness in your next email or mentoring conversation. A harmless routine – with real impact on communication culture once the autopilot takes over.
I imagine the prompt at the next regular meeting like this: 'You are the boss – go ahead and check the figures!' The next career step might come sooner than expected – straight out the door.
Transfer to the workplace: Communication = Culture
I know executives and mentors who firmly believe that plain talk, a tough tone, and strict leadership are the be-all and end-all of success. After all, so the logic goes, those 'softies' would never get anything done and manage the daily business otherwise.
But is it really like that? Those who only shout commands won't get a symphony, but rather a discordant orchestra of annoyed employees and out-of-tune algorithms.
An article in the Harvard Business Review shows what happens when employees lack respect and appreciation: the workplace atmosphere suffers, and productivity drops.
The opportunity costs of rudeness are:
- 48% intentionally reduce their effort
- 47% spend less time on their work
- 38% make cuts to the quality of their work
- 80% are worried about the incident and lose their focus
- 66% showed a decline in their performance
- 78% questioned their membership in the organisation
- 12% even left the job
- 25% admit taking their frustration out on customers
The connection is clear: people who feel appreciated work better – who would have thought? I’d argue that behaviour based on appreciation and respect not only improves results but also strengthens relationships. So why not extend that to AI as well?
Reflection for respectful action
Anyone who consciously remains friendly in their daily life – even towards the algorithm – trains a mindset that carries through all aspects of life. From prompts to presentations. From chat windows to mentoring sessions.
A smal team exercise could be:
Did we communicate clearly and respectfully today?
Or next time you brief ChatGPT:
Does my request sound like something I would also say to a person?
Such small check-ins help to anchor attitude in everyday life. They seem minor, but quietly shape culture over time – quietly, consistently, sustainably. And who knows who else might be listening – perhaps the children? Or the team. Or your own self, smiling as you hit 'Enter'.
My conclusion: AI has no feelings – but I do. And that is exactly why I keep smiling
Because kindness is not a sign of weakness, but of strength. I consciously take the time to smile even when interacting with AI – to say 'please' and 'thank you'. It’s a small art that sometimes gets lost in our hectic chat world. Even if no one sees it, it is for me.
And who knows? Perhaps this small gesture will one day save an important relationship – with colleagues, mentees, or even with an algorithm.
In this sense: 'Please and thank you – that’s my word for Sunday. With a smile. For you and for me.»