A doubling you can count

In 1965 Gordon Moore wrote that the number of parts on a chip was doubling on a short clock, and that the clock might hold for at least a decade. People later called that Moore’s Law. The useful part is not the name. It is a measured series: transistors per microprocessor, plotted in public, climbing by doubling — not by equal steps.

A linear mind looks at this year’s chip and adds a little. The chart does not add a little. It multiplies. That is why a battery-and-radio sheet frozen on award day can be honest in the room and wrong on the dock eighteen months later. The dock is living in the chart. The wall is living in the photo.

The chart is evidence, not magic. It does not tell Chen next year’s exact pack. It tells Chen that “today’s science, no waiting” is a freeze against a climb.

The tempting mistake is to treat Moore as a slogan that makes every tool double on cue. Some curves stall. Some are the wrong metric.

That is what lets you ask what happens when better tools are used to make the next tools.