The first thing my company ever built was a money bomb

The first thing my company ever built was a money bomb, and we built it badly.

This was 2007, back when a campaign website was a brochure. You put the candidate's face on it, posted the press releases, printed the walk lists. Reading material. Then the Ron Paul campaign pulled six million dollars out of the internet in a single day. Fifty-eight thousand donors. People got around to calling that era Web 2.0; at the time it just looked like the internet had suddenly grown hands. A client wanted a money bomb of their own, so my brand-new company said sure, we can build that. We could not. Nobody could yet. There were no rules and nobody to copy, so we wired something together and learned the rules one broken thing at a time.

And nothing new had been invented that morning. The internet had been sitting there the whole time, like lightning before anybody built a wire. What changed was what it was for. A couple of cycles later, online fundraising was a line item in every budget in America, and nobody remembered it ever being a stunt.

A few months ago, my own machine interrupted me with a pitch. It had read the morning's news, spotted a story worth a donor email, and drafted the ask in the candidate's voice, with its reasons attached. I sat there looking at it, and I recognized the feeling from the money-bomb numbers ticking up in 2007: the thing in front of me had quietly changed categories.

What doesn't change categories is who decides. The machine can draft a hundred asks; somebody with scar tissue still has to say yes to one. Took me twenty years to build that judgment. Both times, it turned out to be the part worth paying for.