Invented by mathematician John Conway in 1970, the Game of Life is a "cellular automaton" — a grid of cells that live or die according to four simple rules based on their neighbours. Despite the extreme simplicity of its rules, the Game of Life produces astonishingly complex, unpredictable behaviour, and has been mathematically proven capable of performing any computation a general-purpose computer can perform.
Imagine an infinite grid of square cells, like graph paper. Each cell is either "alive" or "dead." Starting from any initial pattern of living cells, the entire grid updates simultaneously, generation after generation, following just four simple rules based on how many of each cell's eight neighbours are alive:
Despite its playful name and origin, the Game of Life is a serious and extensively studied mathematical object. It has been rigorously proven to be Turing complete — meaning that, in principle, with a sufficiently large and cleverly-arranged starting grid, the Game of Life can be configured to perform absolutely any calculation that any general-purpose computer is capable of performing.
The Game of Life has become a classic and widely-cited example of emergence — the broader phenomenon in complex systems where remarkably intricate, organized, higher-level behaviour arises spontaneously from a set of extremely simple, purely local underlying rules, with no explicit central controller or planner directing the overall outcome.
Some patterns never change from one generation to the next. The simplest is the "block" (a 2×2 square of living cells) — every living cell has exactly 3 living neighbours (satisfying survival), and no dead cell adjacent to it has exactly 3 living neighbours (so none come alive).
Some patterns cycle repeatedly through a fixed sequence of states before eventually returning to their original starting configuration. The simplest oscillator, called the "blinker," is a row of 3 cells that alternates every single generation between a horizontal line and a vertical line.
Some patterns translate steadily across the infinite grid over time while otherwise repeating their shape. The "glider" — a small 5-cell pattern that shifts diagonally by one cell every 4 generations — is the smallest known spaceship and has become an iconic, widely recognised symbol representing hacker culture and computational thinking more broadly.
The "Gosper glider gun," discovered by Bill Gosper in 1970 (responding to a $50 prize Conway had offered for finding any pattern demonstrating truly unbounded growth), periodically emits a fresh new glider every 30 generations, forever — providing the first known concrete proof that Game of Life patterns can grow completely without any theoretical bound over time.
Some very small, simple-looking starting patterns take a surprisingly long time to stabilize into a final steady, repeating, or empty state. The "R-pentomino," just 5 living cells, doesn't fully stabilize until generation 1103 — evolving through an extraordinarily complex, chaotic-looking, and thoroughly unpredictable-seeming intermediate sequence of transformations along the way.
Paul Rendell and others have constructed explicit, functioning patterns within the Game of Life that correctly simulate the fundamental logical components of a computer — including working AND, OR, and NOT logic gates, as well as complete, fully operational Turing machines built entirely out of Game of Life cell patterns. This rigorously demonstrates that the Game of Life is Turing complete: any computation performable by any conventional general-purpose computer can, in principle, also be performed by a sufficiently large and carefully engineered Game of Life pattern, though such constructions can require an enormous number of cells and computational steps to carry out even comparatively modest calculations.
Because the Game of Life is Turing complete, several important general questions about it are provably undecidable — meaning no possible algorithm can correctly answer them for all conceivable input cases. For example, in general, there is no possible algorithm that can always correctly determine in advance whether an arbitrary given starting pattern will eventually die out completely, stabilize into a fixed repeating pattern, or instead grow forever without bound — this is a direct consequence of the general undecidability of the Halting Problem, transferred into the specific context of Life's own dynamics.
The Game of Life's demonstrated capacity for extraordinarily complex emergent behaviour realized this earlier theoretical vision. John von Neumann had earlier (in the late 1940s and early 1950s, published posthumously in 1966) designed a considerably more complicated cellular automaton specifically intended to demonstrate that self-replicating machines were theoretically possible in principle. Conway's Game of Life, though originally designed with a completely different, much simpler purpose in mind, was later also shown (by various researchers) to be capable of supporting genuinely self-replicating patterns, connecting Conway's playful recreational puzzle directly and substantively to serious foundational questions in the theory of computation and artificial life.
Stephen Wolfram's extensive study of simple 1-dimensional cellular automata (published prominently in his 2002 book A New Kind of Science) proposed a broader four-class classification scheme for their characteristic long-term behaviour: uniform, periodic/repetitive, chaotic, and — the most interesting class — complex behaviour capable of universal computation. The Game of Life is a particularly famous and well-studied 2-dimensional example squarely within this fourth, computationally universal class, and has substantially influenced the entire subsequent field of cellular automata theory more broadly.