28.07.2025|
What we can learn about communication from using AI
We optimise prompts for AI chatbots right down to the smallest detail, but send contradictory emails and are surprised if the recipient misunderstands them. Many basic rules for using AI also apply to interpersonal communication.
Anyone who regularly works with AI has probably already noticed that the wording of the text you input has a significant impact on the result. Perhaps your company already has clear guides or guidelines on how to “prompt” efficiently. But how efficiently do you communicate with your colleagues, friends and partners? In fact, many of the basic principles for chatting with ChatGPT and other AI engines are also vital for interpersonal communication. However, this is precisely where they are often forgotten.
Clarity and precision
Unclear communication can result in misunderstandings and unusable answers. If you want your counterpart – whether human or machine – to correctly understand what you are trying to communicate, you should formulate your sentence to reflect exactly what you mean. Let’s say you need an AI-generated picture of a llama in a superhero costume for a social media post. You may already have a clear idea in your head of what this llama should look like. But if you just type in: “Create a picture of a llama as a superhero,” the result is unlikely to match this idea exactly. However, the more details you provide, the more satisfying the result will be
The example of the “Super Lama” shows that AI cannot read minds. And neither can your fellow human beings. But with enough helpful details, they don’t have to.
Context is critical
“Can you please check this invoice?” One and the same sentence can be considered a simple instruction for the long-standing employee, but it can be completely overwhelming for the new intern: What do you need to look out for when checking? What does the invoice need to include in order for it to be correct? What errors can creep in? All of this is valuable context that not every recipient knows. Likewise, chatbots benefit from context, i.e. for whom and for what purpose they are creating a text, an image or an analysis. It is the client’s obligation to provide this context. Otherwise, they should not be surprised if the invoice still contains errors after having been checked or if the PR text from ChatGPT is peppered with marketing elements.
Politeness as a life insurance policy
Are you always polite to ChatGPT? If so, then a study by Future reveals that you are one of the 70% of users who pay attention to your tone when using chatbots or smart speakers. According to the study, 58% do so out of pure decency, while the remaining 12% say they consider it to be life insurance in the event of a robot revolution. If you’re nice to AI, you won’t get pulverised is their reason for doing so. But even without the threat of destruction, a certain degree of politeness is worthwhile and demonstrably leads to better results. This also applies to your professional and private lives. Politeness creates a sense of trust and mutual respect, both of which are fundamental to communication and, in turn, to the quality of the results. Or to put it even simpler: If you ask nicely, you’re more likely to get what you want.
Know who you are talking to!
ChatGPT, Grok, Gemini, Claude, Perplexity – the list of AI chatbots is long, and they all have their own quirks. The perfect prompt for one chatbot might not be as successful with another. The same applies to our fellow human beings. If you know the person you are talking to, you can communicate more efficiently than someone who tars everyone with the same brush. We already do this subconsciously: That’s why we kiss the feet of some bosses, but we make fun of our best friend at every opportunity.
Some people practically have it written on their faces how to communicate with them, while others are much more difficult to read. Developing a feel for who prefers which communication style can take a long time, but it’s enormously rewarding to do so in the long run. In excess, it may seem ingenuine to adapt your communication style to the other person. However, it can be very effective in the right amount.
In summary: If you communicate better with chatbots than with your fellow humans, you may survive when AI seizes power, but you won’t optimise your processes, let alone your relationships.
About the Author
Maximilian Schenner is interested in the impacts of modern technologies on human interaction.