When a computer became something you sat down at

Transistors changed who could sit at a computer. The machines built in the early 1960s were still expensive, but a new class of them was small enough to fit in a room rather than fill a building, cheap enough for a single department to own outright, and meant to be used directly by one person at a console with a screen in front of them. Waiting two days for a printout is a completely different relationship than watching something happen as you type. Once a display could refresh fast enough to show motion and a bank of switches could be read while a program was running, real-time interaction stopped being a research problem and became something students could simply do.

They did. Around university computing centers a culture formed in which writing an elaborate, useless, delightful program was a respectable way to spend the night, and the best known product of that culture was a two-player space duel with a star tugging the ships off course. Nobody sold it. It travelled instead as a courtesy copy handed from one installation to the next, partly because it made an excellent demonstration for salesmen and visitors. That distribution model was accidental but formative. It established that a game could spread on its own merits, with no packaging, no promotion and no price attached, and it gave a whole generation of engineers a shared reference point.

Commerce was moving on a separate track, and mostly outside the United States. Japanese manufacturers were building elaborate electromechanical amusements full of lights, sound and moving parts, and discovering that these exported well. A submarine game shipped abroad in the middle of the decade helped settle a quarter as the American price of a single play, which is the number every arcade designer would work against for the next twenty years. In New Hampshire, meanwhile, an engineer at a defense contractor argued in a short internal proposal that the tens of millions of television sets already installed in American homes sat idle most of the day and could be doing something interactive instead.

None of this looked like an industry yet, and the decade's wider preoccupations were elsewhere: a space program that made engineering glamorous, a television schedule that had become the country's shared culture, and a youth politics deeply suspicious of the institutions that owned all the computers. Two threads ran quietly underneath that noise. Prototypes in a contractor's laboratory proved that a household television could be turned into a playfield with fairly modest circuitry, and a research network carried its first message between two California sites. Neither was aimed at entertainment in any way. Both would be load-bearing within a decade.