For a junior high school science fair I built a machine that played Tic Tac Toe — well actually “Reverse Tic Tac Toe”: the first player to get three in a row loses. It played against a human opponent. It worked. My guess is that the best a human opponent could do was tie.

My shop teacher helped me build a big box that stood on end to house the contraption. On the back were a bunch of wires and switches. On the front were light bulbs arranged in a Tic Tac Toe grid. I moved a switch — a strip of metal — to enter the opponent’s move, and a bulb lit up indicating the machine’s move.

This would have been in the late 50’s. Shannon’s paper on “Programming a Computer to Play Chess” was only published in Philosophical Magazine in 1950 (though it was first presented at the National IRE Convention a year earlier). The “birth” of AI at Dartmouth was in 1956. So a ’thinking machine’ at a junior high science fair in the late 50’s makes me pretty proud of my younger self. I’d be more impressed if I thought I came up with the idea from scratch, but I suspect I got the idea from a suggested project in a library book.

I made a crucial bad design decision though. The switches to enter the opponent’s moves were on the back of the box, and I stood behind the box to work the switches when the opponents told me their moves. That naturally gave rise to the assumption/suspicion that I was actually just making the moves for the machine myself. Maybe that’s why I didn’t get first prize. 🙂

So, the moral? Presentation may be as important as implementation. 🙂