I Broke Conway's Game of Life (for funsies)

This is the first post on the SynthDev blog. I'm building AI tools, developer utilities, and interesting experiments in public. Follow along.

---

What is Conway's Game of Life?

Conway's game of life is a zero-player, cellular automaton simulation designed by British mathematician John Horton Conway in 1970. Really really good stuff so far.

To "play", start with an infinite grid of square cells. If you don't have an infinite workspace to play in, you can create a proxy by defining the outer edges as being "tiled" to their opposite edge, so that if you go off one edge, you arrive on the other edge. Neat.

Each cell can be either alive or dead. Cells live or die depending on number of live neighbors.

Four rules apply each "frame":

1. If a cell has fewer than 2 live neighbors, it dies.
2. If a cell has two or three live neighbors, it stays alive.
3. If a cell has more than 3 live neighbors, it dies.
4. If a dead cell has exactly 3 live neighbors, it comes to life.

The grid is first populated randomly with live and dead cells. Then the 4 rules are applied, generating a new grid. The game proceeds frame by frame, applying the rules over and over. (Potentially infinitely given a large and stable enough universe.)

Emergent Complexity

Interesting self-replicating patterns emerge. Structures are created, duplicated, destroyed. Some arrangements become stable each frame. Some appear to locomote, others permutate through complex progressions before dissolving into nothing.

Generally the entire grid achieves a stable state with several frozen or looping structures each frame forever.

It's an interesting thought experiment. Great complexity and structure emerging from very simple rules. Self-replicating patterns. Nerd.

But...

It's a closed system. No external variables. Only rigidity. Same environment. Same rules. Nothing ever changes.

So I broke it.

Existence Breaks Things.

I thought: "What if an external factor, like wind, were to affect the cells? the wind will push them around the grid between frames and cause interactions that wouldn't normally occur."

"How will this added randomness affect the cells?"

So I added variable wind that pushes "less secure" cells that have fewer neighbors to hold on to.

And it indeed broke it. Chaos. Previously stable systems falling apart rapidly. The grid was quickly wiped bare by even the slightest wind.

The rigid rules and unchanging environment could not adapt to the unexpected new variable, and the wind generally killed off most cells before the grid reached any semblance of equilibrium. Changing the rules of the environment overpowered the rules of the players.

The Spark of Life

Something else was needed to balance this destructive variable.

A constructive variable.

The spark of life emerging randomly where it's not expected.

I added "mutated" cells that came to life in unexpected places.

Freaks.

Rule breakers.

I colored them magenta so their influence could be seen and appreciated.

And the environment thrived. As destructive forces raged havoc on the unbending order of the herd, the mutations showed up to support their neighbors and build new strongholds.

Equilibrium returned.

What's Next?

What if cell behavior has actual intent that survives from one frame to the next, rather than just rigid rules?

---

Play the simulation | View the source on GitHub | Join the conversation on X