One Machine, Any Game<br>Jul 15, 2026<br>One Machine, Any Game

Here’s a question that’s been bugging me for a long time. If you handed a machine a board game, the instruction manual plus what’s in the box (the cards, the board, the dice, the little wooden pieces), could it teach itself to play well? No programmer wiring it up for that specific game. Just the rules, and the machine.

Not one game. Any game. The same system that teaches itself chess should handle Monopoly, Ticket to Ride, and the half-finished prototype a designer scribbled last week, because it never knew which one it was playing in the first place.

I think the answer is yes, and I’ve started building the tool to find out. This series is the working log of that attempt, from the very beginning. But before the first experiment, I want to set the table properly, because the idea only makes sense against seventy-five years of history. Every tool I’ll be reaching for was forged in someone else’s fight with a game, and the arc of those fights bends somewhere interesting.

concept How this series explains things<br>Whenever a new technical term shows up for the first time, the story stops and a block like this one explains the underlying concept in plain terms, with a small demo you can step through yourself. No gesturing at an acronym and moving on. Everything here is written for a general technical audience; no machine-learning background assumed.

Here’s the territory: seventy-five years on a single line, color-coded by the thread each milestone belongs to (click it to zoom). The rest of this post walks it left to right.

⤢ click to enlarge<br>The search era: thinking ahead, very fast

Games were the first thing AI researchers reached for, before the field had a name. Alan Turing wrote a chess program in 1948 with no computer capable of running it; he executed it by hand, one instruction at a time, half an hour per move. Claude Shannon published the blueprint for computer chess in 1950. The idea at the center of both is the one that carried the next fifty years: look ahead.

concept Minimax: assume the worst, pick the best<br>A game between two good players has a strange property: you can score a position by assuming the best from both sides. Consider each of your moves, then each of the opponent’s replies, and so on down a branching tree, until you reach positions you can score. Then let the scores flow back up: at your turns, take the branch with the maximum; at theirs, assume they take the minimum, the one worst for you. The move that survives that pessimism is the right one. It’s easier stepped through than described:<br>mins --> leaves --> ? you pick the biggest ? ? ? they pick the smallest +3 −2 +5 +1 +8 −9 your move their move outcomes ← Back Next →

That’s minimax, and with a pruning trick called alpha-beta (skip any branch that provably can’t change the answer) it is, at heart, the whole search era. Arthur Samuel’s checkers program was playing credible games on an IBM 704 by the late 1950s. Chinook took the world checkers title in 1994. Deep Blue beat Garry Kasparov in 1997 by searching two hundred million positions a second. And in 2007, checkers wasn’t just beaten but solved outright: perfect play from both sides is a draw, full stop.

But notice what every one of those systems needed: a human-written formula saying how good a position is. The search reaches down the tree, hits positions nobody has time to explore further, and asks “roughly, who’s winning here?” Deep Blue’s answer was thousands of hand-tuned features crafted by engineers and grandmasters. The machine did the looking; humans did the judging. That’s the wall the search era never really got past, and Go sat behind it, smug, for decades. Too many branches to search, too subtle to hand-score.

The learning turn: judgment you don’t have to write

The escape had been hiding in the timeline all along. Back in 1959, Samuel’s checkers program learned: it adjusted its own judgment formula by playing against itself, until it could beat its own author. Samuel coined the term “machine learning” to describe what it was doing. It might be the most powerful idea in this whole story, learning from your own experience instead of from instructions, and its reach goes far beyond games: the reinforcement learning that helps turn raw language models into things like ChatGPT belongs to the same family. But in 1959 the idea was decades ahead of its hardware, and it sat mostly dormant for thirty years.

It woke up in 1992 as TD-Gammon, a backgammon program built at IBM by Gerald Tesauro. Instead of a hand-written formula, it used a neural network, and instead of expert data, it learned entirely from playing itself, hundreds of thousands of games. It reached the level of the best human players, and its opening theory was so alien and so sound that human backgammon changed in response.

concept A neural network: judgment you train instead of write<br>For our purposes, a neural network is a judgment machine you...