208, 168, 199, 381, 979, 984, 699, 478, 633, 344, 862, 770, 286, 522, 453, 884, 530, 548, 425, 639, 456, 820, 927, 419, 612, 738, 015, 378, 525, 648, 451, 698, 519, 643, 907, 259, 916, 015, 628, 128, 546, 089, 888, 314, 427, 129, 715, 319, 317, 557, 736, 620, 397, 247, 064, 840, 935.
The rules of Go are simple and take only a few minutes to learn, but the possibilities are seemingly endless. The number above depicts the number of potential legal board positions - which is greater than the number of atoms in the universe - and was only determined in early 2016. Because there are so many directions any given game can move in, Go is a notoriously difficult game for computers to play. It has often been called the “Holy Grail” of artificial intelligence.
A computer therefore has to change how it plays over a brute-force approach – a strategy that can be used for games with less potential moves, such as chess, where it can evaluate all possible moves before making a decision. With Go, it has to break down the next most likely moves and their possible options, but it can only do that for a few steps at a time as there are so many combinations. Because AlphaGo has studied millions of real played games, it has an idea about how the game will proceed…but only up to a point. It has to constantly reevaluate.
How much of what we call creativity is about relying on or referring to past patterns? Certainly a syntax is needed to make sure a given innovation will serve a purpose.
Does the AlphaGo strategy demonstrate the seeds of creative thinking? Or just episodic brute-force and pattern-matching against outcomes learned from historic games…played by humans…who did the original creative thinking? At what point does it test hypothetical moves that no-one has played? And if it can play moves we never thought of, what else could it do?