"ELECTRONIC COMPUTER FLASHES ANSWERS, MAY SPEED ENGINEERING."
New York Times front page, 15 February 1946
On the evening of the fourteenth of February 1946, in a basement at the University of Pennsylvania, a machine the size of a small house computed the trajectory of an artillery shell in twenty seconds. The shell, had it been fired, would have taken thirty seconds to land. The calculation beat the flight.
The reporters in the room did not know what they were looking at. They knew it was big: thirty tons of steel cabinets, 17,468 vacuum tubes, 1,500 relays, 70,000 resistors, miles of cable. And they knew it was hot, because the room was hot. They knew the man at the lectern, J. Presper Eckert, was twenty-six and impossibly young to have built it, and that the older man beside him, John Mauchly, was the one with the original idea. What they could not film was the part that mattered, because the indicator lights on the front of the machine, small neon bulbs the size of a fingernail, were too dim for newsreel cameras.
So Eckert's team improvised. They took ping-pong balls, sliced them in half, painted numerals on the outside, and slipped the painted halves over the neons. The numbers glowed. The cameras rolled. For the next thirty years, whenever a film director wanted to show a computer, he showed flashing lights on a panel, because that is what the newsreels had taught the world a computer looked like. The ping-pong balls were not the machine. They were theatre. But they were the part that photographed.
What the machine underneath had actually been built to do was less photogenic. The army had paid for it because the Ballistic Research Laboratory at Aberdeen Proving Ground, Maryland, could not keep up. Every new gun needed firing tables: for a given shell weight, propellant temperature, wind, humidity and barrel elevation, how far did the shell go? A single table required about three thousand trajectories. A single trajectory, worked by a woman with a mechanical desk calculator, took roughly twenty hours. Guns were arriving in North Africa with no tables. The backlog was measured in years.
The machine in front of the reporters solved the backlog. It did not think about ballistics. It did not know what a shell was. It did what it had been told to do, very quickly, with fewer errors than a tired human. Eckert said as much, repeatedly, to anyone who would listen. The reporters did not listen. The next morning's New York Times carried the story on the front page under the headline ELECTRONIC COMPUTER FLASHES ANSWERS. The Philadelphia Evening Bulletin called it a Giant Brain. A Prudential actuary named Edmund Berkeley, reading the coverage from his desk in New Jersey, began sketching out a book he would publish three years later called Giant Brains, or Machines That Think. The phrase went out through Movietone into cinemas from Minneapolis to Madrid, and by the end of February the world had learned the word computer and decided, without being asked, what it meant.
Standing at the function tables of the machine that evening, dressed for the occasion, were six women. They had programmed the demonstration. They had spent the previous night fixing a fault that had threatened to sink it. They were not introduced. When the demonstration finished, they were asked to serve coffee. After the photographs, the dignitaries went on to a celebratory dinner at Houston Hall. None of the six was invited.
The next morning's newspapers printed the photographs. The captions identified Eckert and Mauchly and the army generals. They did not name the women. The assumption, held by everyone who looked at the pictures and never corrected by anyone who had been there, was that the women were models, posed beside the cabinets to make the machine look glamorous. For forty years, the archived captions called them refrigerator ladies.
This is the evening on which, in the popular mind, the computer was born.