Everyone's saying "taste" is the new currency in the age of AI. But nobody explains how it's built. It just floats as an abstract gift you're supposed to "develop" and "own." Or were just born with.

I reject this deeply. Taste is not magic.

Curiosity is the most important ingredient. It pulls you toward the material. Practice builds the pattern recognition and dexterity. Literacy is the depth of embodied inputs. Taste is just the essence of the whole sequence running.

John Coltrane practiced until his reed cut his lip, then kept going. He shedded for years before anyone heard the results. A Love Supreme was thousands of hours of harmonic study and physical repetition distilled into forty minutes.

J Dilla played the MPC like an instrument. Countless days and nights pushing beats slightly off-center until the swing became a feel no quantize setting could reproduce. Donuts was made by a man who had practiced rhythm until his body knew where the pocket lived.

Gilles Peterson became one of the most trusted ears in music by digging through crates for decades, programming thousands of sets, learning how placing a Brazilian bossa cut next to a broken beat track reshapes both. That's how Talkin' Loud and Brownswood were built.

Andreas Gursky spent decades learning what to remove from a photograph. He composites, strips, duplicates, restructures. Rhine II had its power plants, walkers, and factories deleted until only the structural essence of a riverbank remained. That image isn't documentation. It's a decision only practice could produce.

Tadao Ando never attended architecture school. He drove trucks to fund years of travel, studying Le Corbusier buildings and concrete firsthand. Then he built the Church of the Light.

Daniel Humm worked restaurant kitchens for years, moving through every station, breaking down thousands of techniques, learned what happens when you reduce a carrot jus for an extra forty minutes and why that patience becomes the dish, resulting in Eleven Madison Park.

Pierre Overnoy rejected chemicals when every other Jura producer was adopting them. He spent decades making wine the way he believed it should be made before anyone cared. His Pupillin bottles became the most sought-after natural wines in the world.

Matthew Carter cut punches, drew for phototypesetting, designed for screen at 72dpi. Every technological shift forced him back into the material. Bell Centennial, Big Caslon, Galliard, Miller. Each one works because he practiced type design across five decades of production constraints.

Wim Crouwel didn't arrive at the New Alphabet by theorizing about the future of type. He designed exhibition systems for decades, learned how grids behave under pressure, and built a practice so systematic that when digital screens arrived he already understood the constraints.

The list goes on...

Same sequence. Every time.

None of these people perform “taste.”

Conviction and curiosity drove them to do the work.