John Conway set a black Go stone on the board in the mathematics common room at Cambridge, then a white one beside it, then another black. He was testing rules.
It was 1968. Conway, a thirty-year-old algebraist with wild hair and a taste for games, had taken up a question John von Neumann had answered the hard way. Find the simplest set of rules that could produce complex, unpredictable behavior. Von Neumann's own answer needed twenty-nine possible states per cell and thousands of rules. Conway wanted something a child could follow.
For a year and a half he and his colleagues ran the experiments at afternoon tea. They tried dozens of rule sets, tracked each generation by hand, and swapped stones for pennies when the patterns outgrew the board. Most rule sets died at once, the cells freezing into still shapes or blinking out within a few generations. Others spread without limit and filled the board with noise.
Then Conway settled on three rules. A cell with fewer than two neighbors dies. A cell with more than three neighbors dies. An empty cell with exactly three neighbors comes alive. Everything else stays as it is.
Richard Guy, a colleague who had come to Cambridge to help track the patterns, looked up from the board one afternoon and told Conway, "My blinker is walking."
Five cells had shifted one square on the diagonal. Nobody had written movement into the rules. The rules covered birth, death, and survival, and nothing else. Conway named the pattern a glider and spent the following weeks crashing gliders into each other in forty different configurations to see what the wrecks produced.
