LETTER: Artificial brain is not likely

Editor,

Hugo de Garis made a claim in the Oct. 10 Statesman that is highly unlikely. He said that he wanted to build a machine that “would have all of the cognitive and creative capacity that a human brain has, but it would function a million times faster, because the signals moving from one neuron, or brain cell, to the next would be electronic instead of chemical like in a biological brain.”

Unfortunately, the computer scientist seems to be unaware of the information published by computer scientists at M.I.T., Harvard and Stanford that no one has even come close to understanding what languages are or how they could be used in communication. Not a single one of the languages spoken in the world today has been mapped in such a way that any one of them can be put into a computer and translated into other languages. In fact, the United States government spent millions of dollars trying to do this when we did not have good relations with the Russians.

Humans are indispensable and essential in language translation. We don’t know, nor has anyone suggested, how to do this with computers. The simple commands given to the Japanese toy AIBO are so simple and trivial, that they do not come close to a human language. No computer scientist of my acquaintance has ever suggested otherwise. This shows how essential humans are in language translation. Translation can’t be done without them.

Kent E. Robson