Your dentist's blue paper is a masterclass in agent design

I was at the dentist settling an old Skittle debt. After the filling, she placed a strip of blue paper between my teeth and asked me to bite down, then checked where the color remained and ground down that spot.

I asked her if she just needed to remove the colored spot.

“Only the filling,” she said with a defensive smile. “Not the tooth.”

I asked how long dentists had been doing this.

“As long as there have been fillings, I think.”

She was wrong by about 13,000 years.

The gap

Dental fillings are an ancient art. Every generation of them has the same problem: if a filling sits a hair too high, the patient feels it in every bite. You have to check the bite and grind down the high spots. Always have. Still do.

For thirteen millennia, dentists did this badly. They listened to the patient tap their teeth. They used inked ribbons that smeared the moment they hit saliva. They waited a week and looked for shiny wear spots. The function was ancient. The tool was garbage.

Bausch’s paper

In 1953, a father-and-son pair of dentists named Jean and Hans Bausch made a small piece of paper that changed everything. Dentists worldwide adopted it as soon as they got their hands on it.

Dentists had been trying to mark bite contacts for generations, but every material they used smeared in saliva. The Bausch coating transferred color only under bite pressure, not from ambient wetness. That was the materials breakthrough. The concept was old. Making it survive a wet mouth was new.

It also gave dentists gradient. Old papers left a mark or didn’t. Bausch’s coating left darker marks under harder contact, so darker meant higher. Grind the dark first. A priority map instead of a binary yes/no.

And the paper kept getting out of its own way. A 200-micron strip inserts 200 microns of bulk between teeth trying to contact at micron precision. The tool distorts what it’s trying to measure. Bausch’s line now spans 200 microns down to 8-micron polyester films, thin enough to nearly disappear from the measurement.

The real evolution

The paper mattered because it removed the need for skill where skill was never the point.

Before Bausch, finding the high spot was its own craft. The dentist had to read tap sounds, watch the patient’s face, inspect wear patterns, and synthesize all of it into a guess. After Bausch you just see the color and grind the color. The detection moved out of the dentist’s head and onto the surface of the tooth.

But you can’t blindly remove every marked spot. The color lands on natural tooth too. The dentist still has to know what’s filling and what’s enamel, and that part is real expertise. The paper didn’t eliminate the dentist. It eliminated the part where being a dentist was wasted. It turned a talent problem into a judgment problem, and judgment is what dentists are actually for.

The dentist as an agent

The dentist is an agent. The paper is the harness. A good harness compresses the skill floor. It lets an average agent do what previously required a gifted one.

That’s true for AI agents the same way it’s true for dentists. A well-built harness closes the gap between a frontier model and a smaller open-source one on every task where the harness does the seeing. The model doesn’t get better. The room becomes more legible.

We’ve been building harnesses for human agents forever. A circuit breaker panel is one. A syntax-highlighted editor is one. Bausch’s blue paper is one. Each one moves perception out of the agent’s head and onto the surface of the world. Build rooms where mediocre agents do excellent work.

The 13,000-year lesson

Fillings are ancient. Articulating paper is 72 years old. Every dentist who ever placed a filling knew the bite needed checking. Nobody built the tool that made the checking trivial.

We’re in the 1952 of agent tooling. Most of us are still asking the patient “does that feel high?” Some of us are trying to train better dentists. The move is to build the paper.

See the color. Grind the color.