Abstract

The original edition of What Computers Can't Do comprised three roughly equal parts: (i) a harsh critical survey of the history and state of the art in AI, circa 1970; (ii) a brilliant philosophical expose of four hidden assumptions shoring up AI's rmsplaced optimism; and (iii) a much more tentative exploration of ways to think, about intelligence without those assumptions. Part I, because it was the most combative (and also the easiest to understand), got most of the attention. Also, since that discussion was the most timely — hence the most quickly obsolete — it is what the excellent substantive introductions to the later editions..