Artificial intelligence, quotes from books

Artificial intelligence, quotes from books

The founders of the theory of “Artificial Intelligence” at the Dartmouth seminar in the summer of 1956. Group photo: Oliver Selfridge, Nathaniel Rochester, Nathaniel Rochester, Marvin Minsky, John McCarthy, John McCarthy, Ray Solomonov, Ray Solomonov, Claude Shannon.

There is another, more interesting reason why we should not focus only on human intelligence. Some characteristics of human intelligence in the process of evolution indicate that most of them carry a purely functional load. This is because the human mind can be like a magnificent and impressive peacock’s tail. This tail does not help the peacock to fly, fight, or eat. From the point of view of aerodynamics, this is a heavy load. His role is only to help the peacock find a female. Peacock females choose their males by the size of their tails. It’s possible that most human behavior—including music, poetry, etc.—is just a way to find a mate. Most of what we think of as human intelligence is just a peacock’s tail. These qualities do not help us fight, find food, survive. They only draw attention to us. And in this case, the only reason to study them is to satisfy the curiosity about ourselves. There is no reason to imitate these qualities in technology.

Quote from the book: Artificial Intelligence. Is The Matrix Real \ Artificial Intelligence A Beginner’s Guide, author: Blay Whitby \ Blay Whitby

Indeed, articles and books often focus on the idea of ​​anthropomorphic artificial intelligence, in some sense a simulation of the brain, so the author of such a text believes that it is easier to copy than to create.

But there is another opinion:

A computer can be made at least from wires and switches, at least from pulleys, blocks and ropes; all that matters is how each part changes its state in response to changes in other parts to which it is connected.

Quote from the book: Emotion Machine \ Emotion Machine: Commonsense Thinking;

Modern man knows that computers – computing devices work on electricity, it was not always like that. In the 18th and 19th centuries, they were analog devices that used gear mechanisms that performed a fixed calculation algorithm. Charles Babbage’s differential machine is an attempt to change this state of affairs, to change the fixed algorithm to an arbitrary one, by analogy with the Jacquard loom using a punched card, but the attempt to implement such a mechanism was not entirely successful. With the development of mathematical apparatus and physics, mankind was surprised to find that electricity is not a decisive factor in the construction of computing devices. It turns out that logic gates are basic elements that can be built on anything, as Minsk wrote about.

Binary pneumatic paper adder:

Mechanical half-adder on Lego:

Logic gates in Conway’s cellular automaton of life:

Logic valves on scales and ropes:

If a computer can be built on anything that interferes with making the device “think”, there are many reasons for this question. With automated computing just under a century old, humanity has yet to develop a mathematical apparatus to solve “unsolved questions.”

One of these tasks: combinatorial explosion or, in other words, p, np tasks.

The equality of P and NP will mean that we have a universal efficient algorithm for all NP-problems. The world will change so much that the development of the Internet will turn into a secondary historical fact. It is impossible to describe these changes in detail now, or at least to predict the main consequences of the introduction of new technologies.

Quote from the book: The Golden Ticket. P, NP and the Limits of the Possible \ The Golden Ticket, P, Np, and the Search for the Impossible by Lance Fortnow \ Lance Fortnow

Another problem is related to neural networks:

All feature scoring machines have serious limitations in that, while capable of evaluating the presence or absence of different features, they can only take into account the correlations between these features. For example, in the book Perceptrons, Seymour Papert and I proved mathematically that no feature-scoring machine would be able to distinguish between the pictures below, no matter how carefully we defined the “weights.”

The two figures on the left depict connected structures, that is, those that can be drawn with a single line. The drawings on the right, unlike those on the left, require two separate lines. Here is a way to prove that no feature scoring machine is capable of recognizing such a difference. Suppose we divided each picture into many tiny fragments. It is impossible to tell which fragments belong to the left pictures and which belong to the right, simply because each group of fragments contains identical segments of images! Each will have four fragments with turns at right angles and two “line carriage marks”, and the total length of the horizontal and vertical segments will be the same. Therefore, it is impossible to distinguish one group from another by weighting the evidence, because any information about the connections between the fragments is lost

Quote from the book: The Society of Mind, author: Marvin Minsky

Below will be a list of classic books on artificial intelligence. The first two authors have a sequel without translation into Russian.

Name: What Computers Can’t Do Criticism of Artificial Intelligence \ What Computers Can’t Do: The Limits of Artificial Intelligence
Author: Hubert Dreyfus \ Hubert L. Dreyfus
Year of publication: 2009 \ 1972
Number of pages: 336\354
ISBN: 978-5-397-01007-8\9780060906139

Name: What Computers Still Can’t Do: A Critique of Artificial Reason
Author: Hubert Dreyfus \ Hubert L. Dreyfus
Year of publication: 1992
Number of pages: 429
ISBN: 9780262540674

Name: Gödel, Escher, Bach, this endless garland \ Gödel, Escher, Bach, An Eternal Golden Braid
Author: Douglas R. Hofstadter \ Douglas Hofstadter
Year of publication: 2001 \ 1979
Number of pages: 752\756
ISBN: 5-94648-001-4\9780465026562

Name: I Am a Strange Loop
Author: Douglas R. Hofstadter \ Douglas Hofstadter
Year of publication: 2007
Number of pages: 436
ISBN: 9780465030781

Name: The Society of Mind \ The Society of Mind
Author: Marvin Minsky \ Marvin Minsky
Year of publication: 2018 \ 1985
Number of pages: 592\336
ISBN: 978-5-17-106639-0\9780671657130

Name: The Emotion Machine \ The Emotion Machine
Author: Marvin Minsky \ Marvin Minsky
Year of publication: 2020 \ 2006
Number of pages: 592\400
ISBN: 978-5-17-114660-3\9780743276634

Related posts