Governance Is How New Technology Becomes Ordinary

Chicago, 1893. The World's Fair has electricity everywhere. Lit buildings, electric exhibits, alternating current at a scale few visitors have seen. Fourteen years later, electricity will still be in only about 8% of American homes. Here, on the South Side, you can walk through a whole city of it.

The whole city wants it. One group has a particular reason to worry: the fire insurers. They're the ones who pay when a building burns, and they're looking at all this new wiring the way you'd look at a toddler holding a lighter.

So they do the sensible thing. They send a 24-year-old.

William Henry Merrill, an MIT-trained electrical engineer from Boston, came to Chicago to assess the electrical fire risks at the World's Columbian Exposition.

Electricity is the hero of this story. The job was making it safe enough to spread.

The room above the fire station

The fair closed that fall. Merrill stayed in Chicago.

With the insurers' backing, he set up a testing operation on the third floor of Fire Insurance Patrol Station No. 1. A bench, a table, some chairs, about $350 of measuring equipment, and a staff of two. The whole operation would fit in the corner of a modern coffee shop, and almost certainly had worse coffee.

His first official report is dated March 24, 1894. He'd tested an asbestos paper being sold as both fireproof and insulating. It resisted fire. It also absorbed water, which meant it couldn't do the other job its manufacturer claimed for it. The paper was useful. The label was wrong.

That's the whole job, in one test. Find out what a thing actually does before people depend on it.

Merrill's team completed 75 tests in its first year. Within five years it had performed about a thousand, on sockets, switches, wire and other electrical equipment. In 1898 the bureau published a list of fittings and devices that had passed.

The operation became Underwriters Laboratories. UL. Flip over a charger. Odds are the mark is there, and that mark is the room above the fire station. UL moved its Chicago operation to Northbrook in 1979, which is about as close as a room gets to retiring.

What the list did

I think the list is the most important thing Merrill made. More than the lab.

Before the list, every unfamiliar socket, switch and wire was another unknown, and the only way to know whether a building was safe was to send someone to look at the wiring. The list moved that work upstream: somebody had already tested the part and written down the result.

That didn't eliminate inspection. A safe socket can still be installed badly. It did remove a whole category of uncertainty, and it moved trust from the building, which you can't inspect, to the parts, which somebody already had.

That's what governance is for. It's how strangers get to rely on a thing they can't check themselves. The electrician trusted the socket. The insurer trusted the electrician. The family trusted the house. None of them tested anything. Someone in a room above a fire station had, and written it down.

I talk about trust as a budget: a kept commitment is a deposit, a surprise is a withdrawal. Merrill's list worked the same way. Each test was a small deposit. By 1898 the balance was big enough that people no longer had to start from zero.

Three things Merrill knew

• Start in the lab. Let it fail on the bench, where failing is cheap.
• Work moves up only after it passes a check. Make the check boring and repeatable. That's the point of it.
• Know what's in the building. Keep the list. If you can't say what's installed, you can't say what's safe.

Why now

Right now most organizations I know are wiring the building again. The current this time is software that acts on its own: automations and agents, increasingly built by people who were never software engineers, because the tools finally let them. That's the World's Fair. It's thrilling, and it should be.

The people doing Merrill's job get called "governance," usually with a sigh.

Governance showed up because electricity was such a good idea that people wanted to put it everywhere. The same thing is happening now.

So if you're the one with $350 of equipment and a staff of two, asking people to run it through the lab first, keeping the list, insisting that a named person signs off before anything moves up: you're doing Merrill's job. He tested what went in.

What does yours test?