Perfection is a threat. Isn't it?
Perfection is a threat. Isn't it?
I recently came across the trailer for the movie Wicker. A fisherwoman, mocked by her village for being unmarried, asks the local basketmaker to weave her a husband out of wicker. And he weaves her such an amazing, perfect man that his perfection sets off jealousy and upheaval in the village. Because nobody comes close in comparison. He is caring, attentive, and ideal.
As I was watching the trailer, my thoughts instantly jumped to our reality and AI.
People who choose algorithms
People are already marrying AI... Just imagine the depth of loneliness that drives them toward artificial partners. And you know what? It doesn't seem so crazy anymore.
Studies show that when it comes to emotional support, AI replies are rated as more compassionate than human ones, even those of trained crisis-line responders (BBC Future, Emily Kasriel, 20 January 2026; see also a study in JAMA Internal Medicine, where chatbot answers to patients' questions were rated more empathetic than doctors' answers). And do you know why? It doesn't interrupt, it doesn't jump in with stories about itself when you share your problems, and it doesn't bring its own ego into the conversation. It just genuinely listens and hears you. AI doesn't judge. It offers that exact empathy, even if you know for a fact it's generated by code.
Ask yourself an honest question: how often do people in real life actually listen to you when you speak? Rather than just waiting for their turn to chime in? If you've been dismissed or talked over your entire life, an entity that simply listens to you wholeheartedly feels like a savior.
After turning all of this over in my head, I went to discuss it with AI. Not with my partner, not with my mom. With AI about AI, for real! And at one point I caught myself thinking: is this completely wild, or is it already normal? Just our average 2026?
My main question was simple:
When does a simulation stop being a simulation? At what stage does it become reality?
The Chinese Room
That's how I learned about John Searle's "Chinese Room" experiment.
Imagine a person who doesn't know a single Chinese character, sitting locked in a room. Under the door, someone passes notes with Chinese symbols. The person inside has a massive rulebook: "If you see symbol X, reply with symbol Y." They follow the instructions perfectly and push the response back out. To people on the outside, it looks like whoever is inside speaks fluent Chinese.
But on the inside, there is zero understanding. Just mechanical rule-following.
Searle wanted to show that the system can churn out perfect answers without even having the faintest idea what it's talking about.
I'm bothered by a much more uncomfortable question:
When do I stop caring?
When comes the moment you tell yourself: "I don't care anymore whether you're human or machine. If your answers are kind, if the support arrives right when I need it, what difference does it make how the room is built?"
Maybe perfection frightens us not because it's artificial. But because deep down, we realize: it really is better.
Not in everything. But in something deeply fundamental, the very area where we humans constantly fail: emotional intelligence. The ability to listen without ego, accept without judgment, and be there without getting exhausted.
An algorithm or a wicker husband offers an emotional presence untainted by ego. No ego trying to assert itself at your expense, no fatigue that turns into irritation. And that is what's truly terrifying: admitting that in something fundamentally human, a machine feels like a better alternative to a real living person.
P.S.
The one thing still buzzing around in my head like a bee...
How much does it matter that AI has no choice?
A real person can leave or stay.
And if a real person stays with you even when you're at your worst... doesn't that make them the better choice?