Chess Metaphors: Artificial Intelligence and the Human Mind (MIT Press)
Format: PDF / Kindle (mobi) / ePub
When we play the ancient and noble game of chess, we grapple with ideas about honesty, deceitfulness, bravery, fear, aggression, beauty, and creativity, which echo (or allow us to depart from) the attitudes we take in our daily lives. Chess is an activity in which we deploy almost all our available cognitive resources; therefore, it makes an ideal laboratory for investigation into the workings of the mind. Indeed, research into artificial intelligence (AI) has used chess as a model for intelligent behavior since the 1950s. In Chess Metaphors, Diego Rasskin-Gutman explores fundamental questions about memory, thought, emotion, consciousness, and other cognitive processes through the game of chess, using the moves of thirty-two pieces over sixty-four squares to map the structural and functional organization of the brain. Rasskin-Gutman focuses on the cognitive task of problem solving, exploring it from the perspectives of both biology and AI. Examining AI researchers' efforts to program a computer that could beat a flesh-and-blood grandmaster (and win a world chess championship), he finds that the results fall short when compared to the truly creative nature of the human mind.
functioning of the brain, but here I focus on the neurons, since they are involved in generating cognitive processes. In addition to their classic functions of support, protection, control of pH in the environment, and nutrition (thanks to their intimate relationship with the blood vessels), many more functions for the glial cells have been discovered. This new, more active role for the glial cells includes inﬂuencing communication between neurons by regulating the ionic concentrations on each
experiences. In fact, experiments in which neurons have been made to grow within silicon circuits to form a hybrid connection between a computer and an organism have already taken place. But until then, this fundamental difference between the brain’s neural networks and a conglomeration of chips seems an insurmountable obstacle to building a machine that can carry out intelligent activities. The type of intelligence that can be simulated depends on the software and not the hardware, with the
farther from mathematical algorithms. Finally, chess has a science-like special attraction since it lets the player ﬁrst propose hypotheses of different strategic plans that are based on the game rules and possible moves of the pieces and then refute those hypotheses after careful investigation of the different lines of play. This process is analogous to the everyday work of a scientist. A second class of metaphors—mathematical algorithms, heuristics, and models—brings us closer to the world of
with its billions of galaxies, each with its billions of stars. Every civilization has been fascinated to discover one inﬁnitesimal part of its mysteries in the form of regularities: the sun rises in the east; the moon has a cycle of different phases from full moon to new moon; comets return to pass near the earth periodically; the earth is located in one of the arms of a small spiral galaxy. We could continue enumerating numerous discoveries that ﬁll the shelves of specialized libraries and that
operators (concrete actions that bring the system’s state closer to the given objective). For example, the gain of a pawn can be achieved by bringing in another attacking piece, exchanging a defending piece, creating another threat to attract a defending piece from another area of the board, and so on. The process is repeated recursively until it comes as close as possible to the ﬁnal objective. NSS used a selective search method that was supported by a heuristic based on operators that responded