Print

10 Outsourcing War to Machines of philosophers, mathematicians, physicians, biologists, and computer scientists for centuries. While we have learned a great deal about the bio- chemical processes associated with the human brain, we still do not have a unanimously accepted model of how thinking works in humans, much less a meaningful way to replicate it in machines. Perhaps our greatest mistake has been the attempt to duplicate organic processes in an inor- ganic environment. Machines designed to focus upon a particular aspect of thinking, such as arithmetic calculations or pattern recognition, have proven far more capable than humans in these hyper-specialized func- tions. Thus far, the most useful “thinking” machines have been those dedicated to very specific tasks, at which computers can excel. Because computers do not get tired of repetitive tasks, bored with inactivity, or too exhausted to concentrate, they offer unique possibilities with regard to functions that have traditionally remained human responsibilities. How- ever, turning many military operations over to machines carries its own risk, as the conclusions drawn by computers regarding the same input available to humans may differ markedly from those of humans. Further, because computers have a much wider variety of potential sensory inputs, and because they are typically much better at reconciling conflicting infor- mation from varied sensors, they may act far more decisively than their human counterparts, even in situations when decisive action may not be warranted. Early modern philosophers contemplating the nature of cognition devel- oped a wide variety of explanations for human thought and how individ- ual ideas might be stored within the mind. Thomas Hobbes considered “thinking” to be a form of mental discourse—and he made no differentia- tion between the process of thinking and the act of writing. To Hobbes, the most rational form of thinking followed methodical rules, which led to predictable patterns and reasoned conclusions. Rene Descartes divorced thought, which he considered inherently symbolic, from the objective world. In this regard, he essentially created the modern concept of the mind. Descartes believed the mind was entirely separate from the physi- cal world, and thus reason should be considered entirely divorced from physical laws. David Hume attempted to discern the laws of the mind, applying scientific processes to the notion of mental mechanics. To Hume, thoughts were akin to movements of matter within the brain—an idea that offered the possibility of sharing thoughts directly if the said matter could be isolated and transferred to another individual. Each of these classical philosophers not only expanded the human conception of intelligence but also offered guidance for the development of artificial forms of thinking. For pioneers of AI, it was the Hobbesian notion of methodical patterns that drove the concept of programming. Descartes’s conception of sym- bolism rather than literalism within the mind advanced the AI concept of internal programming languages that need not conform to the rules of