When a support agent says, “I understand your frustration,” it feels natural. When an AI chat bot says the same, it feels grating. A study from McGill University confirms: simulated AI empathy doesn’t calm users down—it triggers psychological reactance.
The problem isn’t the tech; it’s forcing algorithms to pretend they are human.
The Digital Uncanny Valley
You know you are talking to code. When code tries to “empathize,” it creates dissonance. It feels manipulative. A user wants a fix, not a script about “deep concern.”
The result: customers feel their time is being wasted on empty words instead of actions.
The Principle of Useful Politeness
Minimalism in AI communication isn’t rudeness. It is the highest form of user respect.
Replace “fake soul” with results:
-
Instead of: “Oh, I’m so sorry you’re experiencing this error. Let me try to help you right now.”
-
Specifics: “Error logged. Checking your order status. This will take up to 30 seconds.”
-
Instead of: “I’m so happy to be of service! What else can I do to brighten your day?”
-
Specifics: “Task complete. Do you need help with anything else?”
Setting Up an AI Chat Bot for Real Business
To keep your bot from annoying customers, follow three rules:
-
Acknowledgment over sympathy. The bot should acknowledge the fact of an error (“There is a delay”), not the emotion (“I feel your pain”).
-
Verbs over adjectives. Describe the process (“Restoring access,” “Finding file”), not the attitude (“We highly value,” “This is incredibly important”).
-
Nature transparency. Don’t hide that it’s a bot. People are more loyal to an efficient machine than a clumsy human imitator.
Summary
The era of “emotional” chatbots is ending. Today’s value lies in autonomy and speed. Let the bot be a bot. The best politeness your AI can offer is a solved problem in the fewest messages possible.
