Human intelligent behavior arises from skillful bodily activity in satisfying needs, which generates… — Hubert L. Dreyfus

Human intelligent behavior arises from skillful bodily activity in satisfying needs, which generates a human world that pre‑organizes what counts as relevant and significant facts; by contrast, artificial intelligence starts from already produced, decontextualized 'objective' facts and attempts to simulate intelligence by accumulating and processing this neutral data, a strategy that leads to intractable data‑handling problems and is unlikely ever to succeed given the potentially infinite data and the impossibility of fully formalizing our form of life.

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

Key Arguments

Dreyfus characterizes his alternative conception of intelligence as 'an analysis of the way man's skillful bodily activity as he works to satisfy his needs generates the human world,' indicating that intelligence is grounded in embodied coping rather than detached representation.
He insists 'it is this world which sets up the conditions under which specific facts become accessible to man as both relevant and significant, because these facts are originally organized in terms of these needs,' so relevance and significance are world‑ and need‑structured, not added later by rules.
He contrasts this with AI: 'Artificial intelligence must begin at the level of objectivity and rationality where the facts have already been produced. It abstracts these facts from the situation in which they are organized and attempts to use the results to simulate intelligent behavior.'
Once facts are abstracted from their organizing situation, 'these facts taken out of context are an unwieldy mass of neutral data with which artificial intelligence workers have thus far been unable to cope,' highlighting that decontextualization destroys the pragmatic organization that makes facts usable.
He notes empirically that 'All programs so far "bog down inexorably as the information files grow."' indicating a practical manifestation of combinatorial explosion in existing AI systems.
He observes that 'No other data-processing techniques exist at present besides the accumulation of facts,' so current AI is methodologically committed to this accumulation‑and‑retrieval approach.
Once the 'traditional philosophical assumptions underlying work in artificial intelligence have been called into question there is no reason to suppose that digital data storage and retrieval techniques will ever be powerful enough to cope with the amount of data generated when we try to make explicit our knowledge of the world,' tying the empirical storage/search problem to his broader philosophical critique.
He suggests that 'the data about the world may well be infinite and the formalization of our form-of-life may well be impossible,' so any attempt to fully explicitate human know‑how into facts will generate an unbounded or non‑formalizable body of data.
From these considerations he concludes that 'it would be more reasonable to suppose that digital storage techniques can never be up to the task,' i.e., that a fact‑accumulation approach to simulating intelligence will in principle be inadequate.

Source Quotes

Conclusion This alternative conception of man and his ability to behave intelligently is really an analysis of the way man's skillful bodily activity as he works to satisfy his needs generates the human world. And it is this world which sets up the conditions under which specific facts become accessible to man as both relevant and significant, because these facts are originally organized in terms of these needs.

Conclusion This alternative conception of man and his ability to behave intelligently is really an analysis of the way man's skillful bodily activity as he works to satisfy his needs generates the human world. And it is this world which sets up the conditions under which specific facts become accessible to man as both relevant and significant, because these facts are originally organized in terms of these needs. This enables us to see the fundamental difference between human and machine intelligence.

This enables us to see the fundamental difference between human and machine intelligence. Artificial intelligence must begin at the level of objectivity and rationality where the facts have already been produced. It abstracts these facts 1* from the situation in which they are organized and attempts to use the results to simulate intelligent behavior.

Artificial intelligence must begin at the level of objectivity and rationality where the facts have already been produced. It abstracts these facts 1* from the situation in which they are organized and attempts to use the results to simulate intelligent behavior. But these facts taken out of context are an unwieldy mass of neutral data with which artificial intelligence workers have thus far been unable to cope.

It abstracts these facts 1* from the situation in which they are organized and attempts to use the results to simulate intelligent behavior. But these facts taken out of context are an unwieldy mass of neutral data with which artificial intelligence workers have thus far been unable to cope. All programs so far "bog down inexorably as the information files grow."2 No other data-processing techniques exist at present besides the accumulation of facts, and once the traditional philosophical assumptions underlying work in artificial intelligence have been called into question there is no reason to suppose that digital data storage and retrieval techniques will ever be powerful enough to cope with the amount of data generated when we try to make explicit our knowledge of the world.

But these facts taken out of context are an unwieldy mass of neutral data with which artificial intelligence workers have thus far been unable to cope. All programs so far "bog down inexorably as the information files grow."2 No other data-processing techniques exist at present besides the accumulation of facts, and once the traditional philosophical assumptions underlying work in artificial intelligence have been called into question there is no reason to suppose that digital data storage and retrieval techniques will ever be powerful enough to cope with the amount of data generated when we try to make explicit our knowledge of the world. Since the data about the world may well be infinite and the formalization of our form-of-life may well be impossible, it would be more reasonable to suppose that digital storage techniques can never be up to the task.

All programs so far "bog down inexorably as the information files grow."2 No other data-processing techniques exist at present besides the accumulation of facts, and once the traditional philosophical assumptions underlying work in artificial intelligence have been called into question there is no reason to suppose that digital data storage and retrieval techniques will ever be powerful enough to cope with the amount of data generated when we try to make explicit our knowledge of the world. Since the data about the world may well be infinite and the formalization of our form-of-life may well be impossible, it would be more reasonable to suppose that digital storage techniques can never be up to the task. Moreover, if this phenomenological description of human intelligence is correct, there are in principle reasons why artificial intelligence can never be completely realized.

Key Concepts

This alternative conception of man and his ability to behave intelligently is really an analysis of the way man's skillful bodily activity as he works to satisfy his needs generates the human world.
And it is this world which sets up the conditions under which specific facts become accessible to man as both relevant and significant, because these facts are originally organized in terms of these needs.
Artificial intelligence must begin at the level of objectivity and rationality where the facts have already been produced.
It abstracts these facts 1* from the situation in which they are organized and attempts to use the results to simulate intelligent behavior.
these facts taken out of context are an unwieldy mass of neutral data with which artificial intelligence workers have thus far been unable to cope.
All programs so far "bog down inexorably as the information files grow."2
No other data-processing techniques exist at present besides the accumulation of facts,
there is no reason to suppose that digital data storage and retrieval techniques will ever be powerful enough to cope with the amount of data generated when we try to make explicit our knowledge of the world.
Since the data about the world may well be infinite and the formalization of our form-of-life may well be impossible, it would be more reasonable to suppose that digital storage techniques can never be up to the task.

Context

Closing section of the book’s main Conclusion, where Dreyfus contrasts his phenomenological, embodied account of human intelligence and world-formation with the fact-accumulation and data-processing paradigm of AI, tying empirical failures (programs bogging down) to principled limits of digital storage and formalization.