When AI Made "Good" Free

In 1976, a wine merchant in Paris staged a small contest he assumed nobody would notice. He poured French wines and California wines into identical glasses and asked a panel of French experts to rank them. These were serious people. They could taste breeding. They could tell a chateau from a pretender across a crowded room. And with the labels hidden, they were wrong. The California wines won, and the experts never saw it coming, because the thing they were so sure they could see lived entirely on the label, and the moment the label was gone, so was their gift. I think about that test constantly now, because AI has quietly poured every one of us the same anonymous glass.

For all of history, the quality of a piece of work was a signal. A clean report meant a careful person. Elegant code meant an experienced hand. A warm, well-structured email meant someone had bothered. The polish was the proof, and we read character straight off the surface of things the way those judges believed they could taste a vineyard.

Then AI cut the wire. Polish was a signal, and the signal now means nothing at all.

Good got cheap, and cheap things stop being prizes

This is the quiet earthquake nobody is dressed for. For five thousand years, good was expensive. Good writing took a decade. Good design took a gift and a thousand hours on top of it. Good code took scars. So when you saw something good, you knew something true about the person who made it, because they could not have faked the cost. That is over. Good is free now, instant and infinite, and the machine prints it by the mile without breaking a sweat.

And here is the cruel arithmetic of abundance: when everyone can make it good, good stops being worth anything. Competence was a wall that kept most people out, and it has been knocked down into a turnstile that anyone can push through. The average just became free. Which means the average just became invisible.

The scarce thing was never making. It was choosing.

We always assumed the rare, valuable, hard-won skill was the ability to produce good work. We were wrong in the most expensive way possible. The rare thing was always the ability to tell. To look at a hundred competent options, all of them fine, all of them shippable, and know which single one is right. Anyone can fill the glass now. Almost nobody can taste what is in it.

We have a word for that, and we have spent a long time treating it as decoration. Taste. The instinct to say, of two perfectly good things, no, not that one, the other one, and to be right, and to know why. When the floor rises until it touches the ceiling, taste is the only thing left standing in the gap.

When everyone's glass is full, the rarest person in the room is the one who still knows what good actually tastes like.

And the good news nobody tells you

Here is where the despair is supposed to live, and here is where I refuse to leave it. Taste sounds like a birthright, a thing you either inherited or you didn't, the wrong-line lottery of the gifted. It isn't. Taste is not a gift you are born with. It is a debt you pay in attention, one careful look at a time.

The connoisseur was never magic. The connoisseur was just someone who paid attention for an unreasonable length of time. The machine can hand you the wine. It cannot hand you the palate. That part is still earned, the slow way, the only way, and that is the most hopeful sentence in this entire strange moment.

"But good enough is good enough"

Here is the fair pushback. Most work does not need a connoisseur. The status email, the routine report, the boilerplate function, all of it just needs to be fine, and fine is exactly what the machine now produces for nothing. So who cares about taste?

I write about AI, data, and learning regularly at pinaldave.com, and I have been teaching this hands-on in my AI workshops.