From Cambridge to the World

John Horton Conway

Born in Liverpool in 1937, John Horton Conway was one of the most original and playful mathematicians of the twentieth century. A fellow at Gonville and Caius College, Cambridge, and later a professor at Princeton, Conway made foundational contributions to group theory (co-discovering the Monster group), knot theory, number theory (surreal numbers), and combinatorial game theory. He was prolific, irreverent, and legendary for working at a blackboard with no notes.

Colleagues describe a man who could hold an entire complex proof in his head while simultaneously entertaining the room with juggling, backgammon hustling, or explaining why some number's properties were secretly beautiful. His office at Princeton was famously covered floor-to-ceiling in mathematical ephemera and Go sets.

But for all his achievements, Conway spent the last decades of his life with something approaching ambivalence about the thing that made him most famous: a recreational simulation he invented in 1970 while avoiding faculty meetings.

The Problem: Self-Replication

In the late 1940s, John von Neumann — mathematician, physicist, and architect of modern computing — posed a fundamental question: could a machine reproduce itself? Von Neumann designed an extraordinarily complex cellular automaton with 29 possible states per cell to demonstrate that self-replication was theoretically possible. His universal constructor was proven but never actually implemented — the rules were too complex.

Conway's goal was to find the simplest possible rule set that could produce interesting behavior, including the potential for self-replication. He wasn't trying to model anything physical. He was doing mathematics for its own sake.

Working on a Go board during lunches at Cambridge through the late 1960s, Conway and colleagues experimented with various birth/death rules. Most produced boring outcomes — everything died, or the grid filled up immediately. After extensive testing, the B3/S23 rule set emerged: a cell is born with exactly 3 live neighbors; a living cell survives with 2 or 3 neighbors; all other cells die. The rules were deceptively minimal. The results were anything but.

Martin Gardner and Scientific American

Conway's invention might have remained a Cambridge curiosity if not for Martin Gardner. For twenty-five years, Gardner wrote "Mathematical Games" — the most influential recreational mathematics column ever published — in Scientific American. He had an uncanny ability to identify ideas that were simultaneously deep, accessible, and playful.

In October 1970, Gardner devoted his column to Conway's new cellular automaton, calling it the "Game of Life." The article ran only a few pages. The response was unlike anything Gardner had seen in his career.

Within weeks, hobbyists across North America and Europe were simulating Life by hand on graph paper, or — for those with access — on the minicomputers and terminals that were just beginning to appear in universities. The column consumed significant fractions of computing time at institutions including MIT's Project MAC, where systems were reportedly left running Life overnight.

Gardner followed up with additional columns in February 1971 and April 1971 as readers mailed in discoveries. The $50 prize Conway had offered for the first proof that some initial pattern could grow forever without limit — eventually claimed by Bill Gosper's Glider Gun in November 1970 — generated extraordinary community engagement.

Timeline

Conway's Complicated Legacy

For all the joy it brought others, Conway reportedly found his association with the Game of Life more burden than blessing in later years. Visitors to his Princeton office invariably wanted to talk about Life. Journalists called him not for his work on the Monster group or surreal numbers — achievements he considered far more significant — but for a lunchtime recreation.

The irony is that Conway's attitude reflects a truth about creative work: the things that resonate most widely are often not the things their creators value most. Life spread because Gardner's column reached the right audience at the right moment — as computing moved out of institutions and into reach of motivated amateurs. The game gave non-mathematicians a way to do mathematics, to make a discovery, to participate.

Conway's real masterworks — the surreal numbers, the classification of finite simple groups, On Numbers and Games — require years of mathematical training to appreciate. Life required only graph paper and patience. And that accessibility, not the game's mathematical depth, is what made it immortal.