Tokens are the new transistors

In 1965, Gordon Moore had a handful of data points and a hunch: the number of transistors on a chip would double roughly every two years. He was right for the next sixty. But I doubt if anybody in that era really understood what they were looking at. They thought they were arguing about how many components you could etch onto silicon. It'd be hard to imagine how quickly transistors will become the substrate of the world 60 years from then — every phone, car, thermostat, pacemaker, and pair of earbuds. The transistor didn't just get smaller. It got everywhere.

The scale is hard to hold in our heads. A leading chip went from ~2,300 transistors in 1971 to 58 billion in 2021. We now manufacture close to 10²¹ of them a year — more than the industry made in every year before 2017 combined. The transistor is the most-produced object in human history.

I think we're watching the same movie again — except this cut runs at roughly 8x speed.

The new unit isn't the transistor. It's the token: the unit by which we measure AI consumption, the thing a model ingests and emits. And the curve looks eerily familiar. It's hard to come by consistent data sources, but we have some stats from Google I/O's. In early 2024, Google was processing 9.7 trillion tokens a month across its products. A year later, 480 trillion. A year after that, 3.2 quadrillion. That's about 330x in two years — and that's one company.

Transistors doubled every couple of years. Tokens are doubling every few months. What took transistors seventeen years of compounding, the tokens are doing in two.

However, this time is different in two important ways.

First, speed: the transistor explosion was slow enough that society got to adapt — from mainframe to PC, and then to the smartphone. The token explosion is compressing that magnitude of adaptation into just a few years.

Second, ceiling: a transistor is a fixed physical thing bumping into the laws of physics. A token also ultimately has underlying physical limitations, but one way to think of it is that it's just AI pointed at a problem, and there's no obvious limit to how many problems we'll point it at.

So here's the bet I'd make: tokens will end up exactly where transistors did — invisible, ambient, in literally everything. Not a few areas that benefit from it, but every single thing. The teams that win the next decade are the ones building as if tokens are already infinite, because, on the timelines that matter, they will be.

The transistor took fifty years to vanish into the substrate. We know the token is getting there within the next decade.