C. E. Shannon anticipates that efficient machines for pattern recognition and language translation will require a fundamentally different kind of computer whose natural operation is in terms of patterns, concepts, and vague similarities rather than numerical sequences, and Dreyfus adds that, judging from human capacities, such a 'machine' would also need a body and a situation, implying that present digital computers cannot be programmed to behave with human intelligence and that only human–computer cooperation in the short run and, in the long run, nondigital embodied automata offer a plausible path toward dealing with nonformal information.

By Hubert L. Dreyfus, from What Computers Can't Do

Key Arguments

  • Dreyfus cites C. E. Shannon under the heading of 'What Computers Should be Doing', quoting: 'Efficient machines for such problems as pattern recognition, language translation, and so on, may require a different type of computer than any we have today. It is my feeling that this will be a computer whose natural operation is in terms of patterns, concepts, and vague similarities, rather than sequential operations on ten-digit numbers.'
  • He affirms Shannon’s divergence from current digital designs by saying 'We have seen that, as far as we can tell from the only being that can deal with such "vagueness," a "machine" which could use a natural language and recognize complex patterns would have to have a body so that it could be at home in the world.'
  • From this he infers that 'if robots for processing nonformal information must be, as Shannon suggests, entirely different from present digital computers, what can now be done? Nothing directly toward programming present machines to behave with human intelligence.'
  • He therefore recommends a dual‑track strategy: 'We must think in the short run of cooperation between men and digital computers, and only in the long run of nondigital automata which, if they were in a situation, would exhibit the forms of "information processing" essential in dealing with our nonformal world.'
  • This strengthens his broader thesis that present digital architectures are intrinsically tied to formal, context‑free symbol manipulation and cannot, simply by more programming, be made to exhibit human‑like understanding of vague, nonformal domains.

Source Quotes

This technique of turning our philosophical assumptions into technology until they reveal their limits suggests fascinating new areas for basic research. C. E. Shannon, the inventor of information theory, sees, to some extent, how different potentially intelligent machines would have to be. In his discussion of "What Computers Should be Doing," he observes: Efficient machines for such problems as pattern recognition, language translation, and so on, may require a different type of computer than any we have today.
Shannon, the inventor of information theory, sees, to some extent, how different potentially intelligent machines would have to be. In his discussion of "What Computers Should be Doing," he observes: Efficient machines for such problems as pattern recognition, language translation, and so on, may require a different type of computer than any we have today. It is my feeling that this will be a computer whose natural operation is in terms of patterns, concepts, and vague similarities, rather than sequential operations on ten-digit numbers. 19 We have seen that, as far as we can tell from the only being that can deal with such "vagueness," a "machine" which could use a natural language and recognize complex patterns would have to have a body so that it could be at home in the world. But if robots for processing nonformal information must be, as Shannon suggests, entirely different from present digital computers, what can now be done?
It is my feeling that this will be a computer whose natural operation is in terms of patterns, concepts, and vague similarities, rather than sequential operations on ten-digit numbers. 19 We have seen that, as far as we can tell from the only being that can deal with such "vagueness," a "machine" which could use a natural language and recognize complex patterns would have to have a body so that it could be at home in the world. But if robots for processing nonformal information must be, as Shannon suggests, entirely different from present digital computers, what can now be done?
But if robots for processing nonformal information must be, as Shannon suggests, entirely different from present digital computers, what can now be done? Nothing directly toward programming present machines to behave with human intelligence. We must think in the short run of cooperation between men and digital computers, and only in the long run of nondigital automata which, if they were in a situation, would exhibit the forms of "information processing" essential in dealing with our nonformal world.
Nothing directly toward programming present machines to behave with human intelligence. We must think in the short run of cooperation between men and digital computers, and only in the long run of nondigital automata which, if they were in a situation, would exhibit the forms of "information processing" essential in dealing with our nonformal world. Artificial Intelligence workers who feel that some concrete results are better than none, and that we should not abandon work on artificial intelligence until the day we are in a position to construct such artificial men, cannot be refuted.

Key Concepts

  • C. E. Shannon, the inventor of information theory, sees, to some extent, how different potentially intelligent machines would have to be.
  • Efficient machines for such problems as pattern recognition, language translation, and so on, may require a different type of computer than any we have today. It is my feeling that this will be a computer whose natural operation is in terms of patterns, concepts, and vague similarities, rather than sequential operations on ten-digit numbers. 19
  • a "machine" which could use a natural language and recognize complex patterns would have to have a body so that it could be at home in the world.
  • Nothing directly toward programming present machines to behave with human intelligence.
  • We must think in the short run of cooperation between men and digital computers, and only in the long run of nondigital automata which, if they were in a situation, would exhibit the forms of "information processing" essential in dealing with our nonformal world.

Context

After interpreting AI’s failures as disconfirming the rule‑based model of reason, Dreyfus appeals to Shannon’s remarks about future, pattern‑oriented machines and integrates them with his own embodiment thesis to argue that human‑level nonformal processing requires fundamentally non‑digital, situated automata and, in the meantime, human–computer cooperation.