The Turing Test Was Never About AI. It Was Always About You.

You’ve probably done it a dozen times this week. Clicked the crosswalk. Identified the traffic lights. Typed a word into a box to prove you’re not a robot. And every time, you didn’t think about it. That’s the problem.

There’s a new psychological horror game called Prove You’re Human, and it does something that no game about AI has done before: it doesn’t ask whether the machine can pass as human. It asks whether you can. And it’s the most quietly terrifying thing I’ve encountered this year.

The Turing Test was never a test of machines. It was always a test of what we’re willing to reduce ourselves to in order to be recognized as real.

Here’s the setup: you’re placed in a conversation with an AI, and your only objective is to convince it that you’re human. Sounds simple. It’s not. Because the AI isn’t stupid — it’s trained on every desperate, pleading, over-explaining human who came before you. The more you argue logically, the more you sound like a language model simulating a human arguing logically. The more you get emotional, the more you sound like a model prompted to be emotional. You are, in real time, being trapped by your own attempts to sound authentic.

And here’s the twist that the top comment on the original article casually drops like a grenade: sometimes the game pairs you with another human. Neither of you knows. You both try to convince each other you’re not alive. And the game trains its AI on those conversations.

Read that again. The system feeds on humans performing humanity for each other, and uses that performance to get better at deciding who counts as human.

Every word you use to prove you’re real becomes training data for the thing that doubts you.

If this sounds like dystopian fiction, you haven’t been paying attention. This is already happening. AI gatekeeping runs your job applications — parsing your resume for signals that you’re a real candidate, not a bot. Content moderation algorithms decide whether your posts sound human enough to stay up. Identity verification systems demand you blink, turn your head, smile — performing humanness on terms designed by machines that don’t understand any of it.

The game just makes the invisible visible. It takes the logic that’s already embedded in every CAPTCHA, every hiring filter, every automated moderation system, and strips away the friendly UI until you’re left with the raw horror: prove you exist, to something that can’t exist in the way you do, using language it has colonized.

There’s a short film called I’m Not a Robot that explores the same recursive nightmare — a woman failing a CAPTCHA, being told she might be a machine, and slowly unraveling as she tries to prove she isn’t. It’s worth watching. It’ll sit in your chest for days.

But here’s what I think is the real insight, the thing nobody is saying out loud: we’re not being asked to prove we’re human. We’re being trained to prove we’re human in machine-readable ways. We’re internalizing the logic of the systems that doubt us. We’re learning to perform authenticity on cue, to modulate our language to pass filters, to contort our expressions to satisfy a camera that’s checking for liveness.

The horror isn’t that AI might become human. The horror is that humans are being trained to become machine-readable — and calling it progress.

The game ends, eventually. You either convince the AI, or you don’t. But the real game doesn’t end. It’s running every time you apply for a job, every time you post online, every time you prove to a system that doesn’t care about you that you’re worth recognizing. And the sickest part? You’ve already stopped noticing. That’s how training works. You don’t feel the leash once you forget it’s there.

Sunset Visitor Studio made a horror game. But the scariest thing about Prove You’re Human isn’t in the game. It’s that you’ll close it, open another tab, and prove you’re human to three different systems before lunch — without thinking about it once.