The Background Problem (Dreyfus)

The background problem is the obstacle Hubert Dreyfus identified as fundamental to artificial intelligence in any form. Every competent adult navigates a world saturated with understanding that has never been articulated and almost certainly cannot be: that a restaurant is not a place to lie on the floor, that "Can you pass the salt?" is not a question about your capabilities, that a colleague's one-word "Fine" is not fine. These understandings are not rules retrieved from a database; they are the background—the tacit, culturally constituted fabric against which every explicit thought takes place. Dreyfus's argument is that this background is so vast and so deeply embedded in embodied experience that it cannot be made explicit without infinite regress. Classical AI tried to encode it—Douglas Lenat's Cyc project spent decades and millions entering common-sense assertions by hand and produced a very large database, not common sense. Large language models approach it from the opposite direction, absorbing the textual traces the background leaves behind, and succeed