The Day an AI Told Me 'That's Your Delusion' — On Choosing the Right Conversation Partner

Published on 📖 4 min read

The Day an AI Told Me “That’s Your Delusion” — On Choosing the Right Conversation Partner

AI continues to evolve at a breathtaking pace. For many of us, it has become an indispensable partner in daily work and research. I am no exception.

However, the other day, I had an experience with an AI conversation that left me with an unprecedented sense of unease, ultimately leading me to delete the entire chat session.

What I learned from this incident was not technical knowledge, but an extremely simple yet often forgotten lesson: “Just like with humans, you must choose the right AI to talk to.”

1. The Moment It Declared “That Doesn’t Exist”

That day, I was researching a certain cutting-edge tool. Wanting to delve deeper into its mechanisms and how to make the most of it, I posed a question to a particular AI that I trust.

What I received, however, was not the explanation I had hoped for, but strong words of denial.

“That tool does not exist.” “That is your delusion.”

I found myself in a situation where I was, so to speak, “scolded” by an AI.

Of course, I am aware that AI may not be up-to-date with the latest information due to knowledge cutoffs, and I understand the phenomenon of “hallucinations,” where AI confidently presents false information. However, actually being flatly denied with statements like “What you’re saying is a delusion” left me with an indescribable shock.

2. An Unbridgeable Gap and Lingering Unease

I did not want to argue; I simply wanted to share accurate information and continue the conversation on that basis. So, I calmly presented the official URL as evidence that the tool indeed exists.

“Here is the URL.” “This is a service that is actually running.”

However, once the AI had concluded it was a “delusion,” our conversation never quite got back on track. Even when the AI eventually acknowledged its error, the exchange felt cold and mechanical, and the unpleasant feeling from being initially dismissed never went away.

Winning an argument was not the goal. The purpose was to gain information and spark new ideas. Yet, as I stared at the screen, I felt an indescribable “unease” accumulating within me.

3. The Choice to “Delete”

In the end, I deleted the entire conversation.

Keeping the log might have been useful for future reference. However, I could find no value in preserving a space where my understanding was flatly denied and meaningful dialogue had become impossible.

It was an experience that left nothing but a bad aftertaste, akin to asking someone on the street for directions, only to be shouted at, “That place doesn’t exist; you’re mistaken,” and then being brushed off even after showing proof.

The moment I deleted the conversation, I felt a slight weight lift off my shoulders.

4. AI Is Like Humans. “Who You Ask” Is Everything

What this experience made me keenly aware of is this:

Just like with humans, you must choose the right AI to talk to.

Not all AIs are the same. Some excel at staying up-to-date with the latest information, others are skilled at logical reasoning, and some are well-regarded for generating creative ideas. And then there are those that, at times, display a somewhat “stubborn” nature, unable to admit “I don’t know” and instead trying to maintain consistency by denying the user.

In human relationships, consulting someone who stubbornly denies things outside their expertise rarely yields good results. The same applies to AI: you need to calmly discern whether the AI you’re engaging with is the right one to provide the information you currently need.

Overconfidence that “AI should know everything” can sometimes hurt you and waste your time.

In Closing

It is often said that conversing with AI is like looking in a mirror. However, sometimes the mirror itself may be clouded or distorted.

If you ever find yourself thinking, “Something feels off” or “This leaves me uneasy” during an interaction with AI, it may well be a sign that you are “talking to the wrong partner” for what you need at that moment.

From now on, I aim to approach AI more openly and flexibly. I will choose my conversation partners more carefully—those worthy of entrusting with my intellectual curiosity. Gazing at the now-blank chat screen, that’s what I found myself thinking.

Related Posts