What Computers Can't Do
From Wikipedia, the free encyclopedia
- See also: Philosophy of artificial intelligence
What Computers Can't Do: The Limits of Artificial Intelligence is a controversial work on artificial intelligence, authored by Hubert Dreyfus, a professor of philosophy at the University of California, Berkeley. The book deals with the idea that thought, intelligence or reason can be reduced to computation. After a summary of the idea's history Dreyfus proceeds to attack this project, and show why it is impossible, regardless of the claims of the Artificial Intelligence (AI) research community. No other book has ever produced as much controversy and emotion in the Artificial Intelligence community. [1]
Contents |
[edit] The book's argument
Dreyfus summarizes the History of artificial intelligence both as a conceptual aspiration (up to the 20th century) and as a technology (since). The book takes great exception to the unwarranted optimism that seems to permeate the field, for example. Herbert Simon, following his General Problem Solver (1957), predicts that within 10 years a computer will
- Be world champion in Chess and
- Discover and prove an important new mathematical theorem. In the same timescale
- most theories in psychology will take the form of computer programs.
Dreyfus also ridicules the public's credulity regarding anything robotic or Artificial Intelligence-related, e.g. quoting the following claim from a US newspaper, in 1968:
Cosmos, the West German publishing house... has come up with a new idea in gifts... It's a genuine (if small) computer, and it costs about $20. Battery operated, it looks like a portable typewriter. But it can be programmed like any big computer to translate foreign languages, diagnose illnesses, even provide a weather forecast.
Within this atmosphere of wild optimism no critical discussion of any underlying conceptual difficulties is allowed. Dreyfus points out major differences between Human's and machine's thought that Artificial Intelligence research stumbles across time and again:
- Fringe consciousness vs. Heuristically guided search
- Ambiguity tolerance vs. context-free precision
- Essential/inessential discrimination vs. trial-and-error search
- Perspicuous Grouping vs. Character Lists
These difficulties are not coincidental. All work in Cognitive Simulation and in Artificial Intelligence is predicated on one basic assumption: that Humans in some fundamental way process information in ways that computers can emulate. This is no small assumption, because all computer-based information is explicit, discrete, linear, rule-based and definitive, while we have no evidence that human thought is so. The assumption that human minds function like general-purpose symbol-manipulating machines amounts to:
- A Biological Assumption - That at some level people operate in a digital manner
- A Psychological Assumption - All thought is calculation
- An Epistemological Assumption - That all knowledge can be formalized
- An Ontological Assumption - That our world consists of context-free facts
All these assumptions would be very convenient if they were true, however there is little evidence to support them and much to refute.
Dreyfus argues for the role of the body in intelligent behavior, for the importance of the overall situation for orderly behavior, and the inapplicability of any rule-set. He also stresses the importance of specifically Human needs as determinators of the situation (in the Heideggerian sense) which is the context in which the interpretation of the world occurs.
The book concludes with a chart classifying intelligent activities (quoted):
| I Associationistic | II Simple-Formal | III Complex-Formal | IV Nonformal | |
|---|---|---|---|---|
| Characteristics of Activity | Irrelevance of meaning and situation | Meanings completely explicit and situation independent | In Principle, same as II; In practice internally situation-dependent, independent of external situation | Dependent on meaning and situation that are not explicit. |
| Innate or learned by repetition | Learned by rule | Learned by rule and practice | Learned by perspicuous examples. | |
| Field of activity (and appropriate procedure) | Memory games, e.g. "Geography", (association) | Computable or quasi-computable games, e.g. nim or tic-tac-toe (seek algorithm or count out) | Uncomputable games, e.g. Chess or Go (global intuition and detailed counting out) | Ill-defined games, e.g. riddles (perceptive guess) |
| Maze problems (trial and error) | Combinatorial problems (non-heuristic Means-ends analysis) | Complex combinatorial problems, (planning and maze calculation) | Open-structured problems (insight) | |
| Word-by-word translation (mechanical dictionary) | Proof of theorems using mechanical proof procedures (seek algorithm) | Proof of theorems where no mechanical proof procedure exists (intuition and calculation) | Translating a natural language (understanding in context of use) | |
| Response to rigid patterns (innate releasers and classical conditioning) | Recognition of simple rigid patterns, e.g. reading typed page (search for traits whose conjunction defines class membership | Recognition of complex patterns in noise (search for regularities) | Recognition of varied and distorted patterns (recognition of generic or use of paradigm case) | |
| Kinds of Program | Decision tree, List search, Template | Algorithm | Search-pruning Heuristics | None |
[edit] Bibliographical Data
The book initially appeared under this title in 1972 (ISBN 0-06-090613-8), and a second edition with a new introduction was published under the same name in 1979 (ISBN 0-06-090624-3). A third edition was published under the name What Computers Still Can't Do (ISBN 0-262-54067-3) in 1992.
[edit] See also
- A detailed summary of the book's argument
- Computer Chess
- Church–Turing thesis
[edit] References
The book itself, and
