Hubert Dreyfus
Hubert Dreyfus, PhD, is a professor of Philosophy at the University of California, Berkeley, whose main interests include phenomenology, existentialism and the philosophy of both psychology and literature, and philosophical implications of artificial intelligence.
Background
Professor Hubert Dreyfus is an American philosopher and the author of the 1972 book What Computers Still Can't Do and in 1979 its revision What Computers Still Can't Do. Professor Dreyfus taught at the Massachusetts Institute of Technology between the years of 1960 and 1968. He also taught Frankfurt University and Hamilton College.
In 1964 Dreyfus published Alchemy and Artificial Intelligence, a destructive criticism on the work of Allen Newell and Herbert Simon, two of the leading researchers in the field of Artificial Intelligence. As a contemporary Socrates, Dreyfus not only questioned the results they had so far obtained, but he also criticized their basic presupposition that intelligence consists of the manipulation of physical symbols according to formal rules, and argued that the AI research program was doomed to failure. In 1965, he spent time at the Rand Corportation, whilst work on artificial intelligence was in progress there.
Dreyfus's criticism of AI
Dreyfus's critique of artificial intelligence (AI) concerns what he considers to be the four primary assumptions of AI research. The first two assumptions he criticizes are what he calls the "biological" and "psychological" assumptions. The idea is that the brain is analogous to computer hardware and the mind is analagous to computer software; that is, they both perform discrete computations on discrete representations. A simplistic sort of picture is that humans form a representation of external reality which they manipulate according to the computational rules that define intelligence in order to figure out what to do.
Dreyfus claims that the plausibility of the psychological assumption rests on two others: the epistemological and ontological assumptions. They assume that all non-arbitrary behavior can be formalized and that reality consists of a set of mutually independent, atomic (indivisible) facts. It's because of the ontological assumption that it makes sense to think about forming an internal representation of reality, and because of the epistemological one that correct reasoning could be equated with formal rule-following.
What these two assumptions have in common, in Dreyfus's formulation, is they are an objective, context-free form of understanding (of the sort the physical sciences use). The reasonable-seeming idea is that we can understand our own behavior just like that of anything else (say, planetary motion); by considering ourselves as objects which behave in prescribed ways to universally definable conditions.
Dreyfus's argument against these assumption boils down to using phenomenology (careful observation of the presumably universal structure of experience) to state that our being is highly contextually bound, which is why the two context-free assumptions are false. He concludes that, while this isn't a knock-down argument that AI is impossible, he thinks it presents a case that it's very unlikely; and that to get a device with human-like intelligence would require it to have a human-like being in the world, which would require a body more or less like ours, and social acculturation more or less like ours (see also: embodied philosophy).
Achievements
Erasmus University awarded Dreyfus an "for his brilliant and highly influential work in the field of artificial intelligence, and for his equally outstanding contributions to the analysis and interpretation of twentieth century continental philosophy".
External Resources
Online
-
-
Paper-based
- Alchemy and Artificial Intelligence
- Continental Philosophy: An Introduction
- What Computers Can't Do: The Limits of Artificial Intelligence (1972 - ISBN 0060906138)
- What Computers Still Can't Do: A Critique of Artificial Reason (1979 - ISBN 0262540673)
- Mind Over Machine (1986)
- Being in the World: Division 1 (1991)
- On the Internet (2001)
- Internet (2002)
