What Computers Still Can't Do: A Critique of Artificial Reason (MIT Press) Revised ed. Edition, Kindle Edition

The dream of artificial intelligence is still alive and well despite its repeated failures. Why?

Dreyfus traces the program from its early successes in the middle of the last century to the abandonment of all the major players at the end. In doing so he provides an excellent explanation into the philosophical misconceptions that led (and still lead) many to overestimate the prospects of this program. He draws on Heidegger and Wittgenstein to correct a flawed picture of what it is to "know" that is still prevalent in Western thought.

This book and Dreyfus' life work are very notable accomplishments and demonstrate the powerful of insightful humanistic enquiry in our increasingly scientific and technical age.

Dreyfus brings philosophy to the rescue of humanity, namely, and ironically, Martin Heidegger. Though Heidegger was mostly indebted to his teacher and mentor Edmund Husserl, Dreyfus treats him as some sort of savior of mankind. No need, for William James’ Principles of Psychology, which predates Being and Time by almost forty years, could have done just as well in proving why computers can’t think. James’ work on consciousness and attention greatly influenced Husserl and the birth of phenomenology. Had Dreyfus deferred his argument to James instead of the Nazi Heidegger, he might have gotten more respect from the AI researchers. James’ criticism of the “automaton theory” of the mind is much more convincing than Heidegger’s dull notions about hammers.

In drawing attention to his muse Heidegger, Dreyfus is very critical of the philosophical tradition, which relied primarily on reason. Yet reason, which Dreyfus does not define, is the central focus of his argument for being human. His method is primarily his ability to doubt that computers can think, i.e., I can’t think, therefore, I can’t be human. Is this the tradition in disguise or what?

Of course, his argument goes much deeper than just thinking, which computers do best in the form of calculating, etc. It has to do with their inability to function in the world like humans, which is more of a problem for robots than computers per se.

Humans are prone to make mistakes, especially when confronted with new environments or handling complex systems like spacecraft. Engineers must design systems in ways to question the appropriateness of human behavior. Is it safe to do such a thing at this time? Of critical importance in the whole study of AI is not whether computers can do what humans do, but whether computers can support, correct, and enhance what humans do.

Fifty years from now when science moves information processing from the digital to the molecular level, new forms of knowledge will emerge, which will be a call to philosophers to reinvent the tradition and reawaken interest in Plato, Descartes, and Kant. Then some philosophy professor will write a book titled Why Computers Can Think And Man Still Can’t and end the debate.

Mr. Shallit, in a previous review, sounds exactly like a throwback to the previous era when glorified computer technicians started to take themselves seriously and bestowed upon themselves the bogus title of "computer scientists". When Dreyfus critique of strong AI came out they were enraged that someone was trying to punch the air (and as it turned out it was full of hot air) out of their over inflated clown balloon. After decades of sheer nonsense, and of confusing metaphor with the real thing (their preposterous proposition of the "complexity critical mass" for intelligence is akin to thinking that a sufficiently complex computer simulation of an atomic explosion will actually explode: I guess we should all be really worried...) strong AI proponents just gave up. Now they are much quiet and many have given Dreyfus some concessions after being proven completely wrong (old habits die slowly). No serious GCT (Glorified Computer Technician) of today would profess faith in strong AI.

On the 1st front, the book argues that machines can't replace people. On the second front, I highly recommend to anyone interested in Big Data. Take the points given by Dreyfus as a methodology to interpret and understand how a machine thinks.