He was in charge of a computer called SEAC, which was the Bureau of Standards Eastern Automatic Computer. I forget who had designed this... this machine, but this machine had an actual dynamic memory; mercury delay lines in fact... put a pulse of sound into a pipe and it would take a fraction of a second to go around this long liquid path of mercury, and you could put maybe 500 zeroes or ones into that thing, so it... it was moderately fast. And there were four of these machines made for different places, and that’s the first computer I ever programmed, because – why am I having trouble remembering his name? – Russell Kirsch! Because Kirsch sat me down at a desk and said I couldn’t get up till I’d written a program for the SEAC. So I wrote a program which actually recognised the shape of a couple of graphical letters, and Russell Kirsch in fact wrote some of the very first graphic programs for any computer, including a little picture of a person’s face. So I didn’t actually touch a computer again till maybe 1956 when there were at last some real computers, and I wrote a couple of programs for the one out at RAND.