Intellectual issues in the history of artificial intelligence

Science is the quintessential historical enterprise, though it strives to produce at each moment a science that is ahistorical. With a passion bordering on compulsion, it heeds the admonition that to ignore the past is to be doomed to repeat it. Science has built its main reward system around discovering and inventing, notions that are historical to the core. Thus, writing about science in the historical voice comes naturally to the scientist. Ultimately, we will get real histories of artificial intelligence (henceforth, AI), written with as much objectivity as the historians of science can muster. That time has certainly not come. We must be content for a while with connections recorded in prefaces, introductions, citations, and acknowledgments the web that scientists weave in their self-conscious attempt to make their science into a coherent historical edifice. So far, only a few pieces, such as Machines Who Think, provide anything beyond that, and they still have no deliberate historiographic pretensions. [McCorduck,