What AI Actually Sorts
There is a dating app that floods Instagram with ads today. It writes your opening line and promises an unfair advantage. The funny part isn’t that it’s dishonest. It’s that it says the thing nobody says out loud. The words were never the point.
Flirting is a test. Not of your vocabulary. Of whether you can think on your feet when something is at stake. The wit is the thinking, made visible. An app that sends the words without supplying the thinking is a way of failing the test faster.
Initially it looks like a fact about dating. It isn’t. It’s a fact about signals, and signals are everywhere.
A signal is anything you do to convince someone of something they can’t check directly. She can’t read your mind, so she reads your sentences. An employer can’t watch you work, so he reads your resume. A signal works only when it’s expensive to fake. A diploma means something because you can’t get one in an afternoon. An opener means something because most people freeze.
But AI makes faking cheap. People worry this lets individuals cheat. The real damage is to everyone else. When anyone can generate the perfect opener, the opener stops meaning anything. Not for the cheater. For the honest man who happened to be witty on his own. The counterfeit doesn’t rob one victim. It debases the currency. After enough fakes, a good signal and a bad one look the same, and the receiver stops believing any of them.
This is happening to a lot of things at once. Cover letters. Student essays. Product reviews. Profile photos. Each one was a signal. Each one is becoming noise at the speed the models improve.
Get Thiyagarajan Maruthavan (Rajan)’s stories in your inbox
Join Medium for free to get updates from this writer.
So we’re losing something. That feels right and it’s half wrong, which is the half worth staying on.
The signal doesn’t vanish. It moves. AI eats the opener, the test moves to the date. It eats the essay, the test moves to the conversation where you defend it and can’t paste. It eats the resume, the test moves to the afternoon you spend working with someone. The un-fakeable layer slides downstream, to wherever you have to be present, in real time, with no chance to precompute. AI deletes the part of the test you could fake. That was never the part that mattered.
The people who are upset are mourning the opener. The opener was hazing. What survives is what the opener was a weak proxy for. Can you sustain this, in person, when the words have to come from you.
The other side is stronger than it looks. Printing. Photography. Spellcheck. A nervous man who genuinely likes someone and can’t find the first line isn’t faking quality. He’s paying an anxiety tax that has nothing to do with whether he’d be good to love. Remove the opener and maybe you removed a filter that screened for confidence, not character.
Both are true and they don’t contradict. They disagree about one thing. Where the real test sits. If it sits at the opener, the app is fraud. If it sits at the third month, the opener was always noise and the app skips a ritual. AI didn’t settle this. It made the cheap layer free, which shows how much of what we measured was cheap layer.
So what does AI sort. Not winners from losers. It takes everything we treated as proof and splits it into two piles. One was real. The other was expensive enough to look real. The second pile is free now, falling to zero. The first pile is judgment, taste, presence, the willingness to be in the room and own what you said. That pile is the only thing left worth measuring. It’s the only thing the machine can’t hand you.
A rule for the next ten years. When a tool offers to do something for you, ask whether the doing was the proof. If it was, the tool isn’t handing you an advantage. It’s removing your reason to have one. The people who win won’t be the ones who let the tool do the work. They already had the thing the tool can’t fake, and used the tool to skip everything else.
The rest of us will spend a few years confusing a cheaper signal for a better self. Then the machine finishes sorting, the way it always does, in the room where the words have to be yours.
