Artificial intelligenceThe question whether machines can compete with human intelligence has concerned scientific community since 1950, when Alan Turing published his historical study.

The science of artificial intelligence provides a series of challenges that must be overcome before the dream of Turing is realized. If you are put the question: “Which element has the atomic number 27 and precedes “blue” and “green”?” The human brain is able to combine language concepts and raise information from a huge database in memory before you answer: “Cobalt”.

Computers have become very proficient in it. Think that the Internet’s search engines can find your question just after you type a few letters.

However, modification of the mathematical algorithm behind the search engines, so that they can give chess problems answers, was the beginning of the domination of machines. This was happened in 1999, when the IBM computer Deep Blue won over the world chess champion Garry Kasparov. Twelve years after it appeared the IBM supercomputer Watson which has the ability to analyze the importance and content of human language and quickly process enormous amounts of information.

The analysis of the logical implications of each movement on a chessboard is the easiest task for a computer.

Turing test

The Turing test from the study of 1950 has to do with communication. Turing asked: while talking over the network with a man and a computer can you distinguish one from another?

Since the only way to judge a person’s intelligence is to communicate with him, should not we attribute intelligence with a computer if it was able to impersonate a human?

However the modern science of artificial intelligence believes that this definition is not reliable. Even if a computer manages to pass the Turing test, it does not have human perception. This is easily proved if we think that although computers are able to translate languages, they can’t digest complex concepts.

The biggest challenge of artificial intelligence is to emulate the human ability to manage visual images. People are able to decode malformed characters and find hidden messages behind them. Computers are not.

The importance of this ability is manifested clearly if we think about the operation of security cameras. Security companies would love to have software capable of recognizing suspicious persons or behavior. So far, only the human eye can do that.

Anna LeMind, B.A.

Copyright © 2012-2020 Learning Mind. All rights reserved. For permission to reprint, contact us.

Leave a Reply