"Restless for going on three years now." His face grew solemn, as though he were thinking it over very carefully. "I'll amend that statement. Hell with the Aesopian language. I've been a plain lush for going on three years. Ever since--" If it was something personal--I suggested. "It is _not_ something personal," he said, mimicking me. "Guess I can tell an old cyberneticist pal about it. Been a lush for three years because I've been scared for three years. Been scared for three years because three years ago I saw a machine beat a man at a game of chess." A machine that plays chess? That was interesting, I said. "Didn't tell you the whole truth the other day," Len mumbled. "I _did_ work on the Remington-Rand computer, sure, but I didn't come to IFACS directly from that. In between I spent a couple years at the Bell Telephone Labs. Claude Shannon--or, rather, to begin with there was Norbert Wiener back at M.I.T.--it's complicated...." "Look," I said, "are you sure you want to talk about it?" "Stop wearing your loyalty oath on your sleeve," he said belligerently. "Sure I want to talk about it. Greatest subject I know. Begin at the beginning. Whole thing started back in the Thirties with those two refugee mathematicians who used to be here at the Institute for Advanced Studies when Einstein was around. Von Morgan and Neumanstern, no, Von _Neu_mann and _Mor_ganstern. You remember, they did a mathematical analysis of all the possible kinds of games, poker, tossing pennies, chess, bridge, everything, and they wrote up their findings in a volume you certainly know, _The Theory of Games_. "Well, that got Wiener started. You may remember that when he founded the science of cybernetics, he announced that on the basis of the theory of games, it was feasible to design a robot computing machine that would play a better than average game of chess. Right after that, back in '49 or maybe it was '50, Claude Shannon of the Bell Labs said Wiener wasn't just talking, and to prove it he was going to _build_ the robot chess player. Which he proceeded withforth--forthwith--to do. Sometime in '53, I was taken off the Remington-Rand project and assigned to Bell to work with him." "Maybe we ought to start back," I cut in. "I've got a lot of work to do." "The night is young," he said, "and you're so dutiful. Where was I? Oh yes, Bell. At first our electronic pawn-pusher wasn't so hot--it could beat the pants off a lousy player, but an expert just made it look silly. But we kept improving it, see, building more and more electronic anticipation and gambit-plotting powers into it, and finally, one great day in '55, we thought we had all the kinks ironed out and were ready for the big test. By this time, of course, Washington had stepped in and taken over the whole project.