This tale of the Chinese room By John Searle (a contemporary philosopher, specializing in issues related to language and mind) is best known to a thought experiment, called "
Chinese Room", originally formulated in opposition to the idea of \u200b\u200bthe opportunity to program a computer to think like a human being.
summarize the terms of the situation, you have a man imprisoned in a room that receives daily an envelope containing the symbols have no meaning for him, looking in a book at his disposal, which are coupled to other, equally incomprehensible to him. He copies the new symbols, and then pass the sheets that they occupy a slot in the door (note the obvious analogy with
input and output
a computer), after a few minutes, receives food.
What really happens is that humans are sent phrases in Chinese, then he uses a manual with ready-made answers and then brings forward a kind of conversation. The question is, "the prisoner knows the Chinese"? Searle says no, but it deserves more consideration, for various reasons.
First, the term can be misleading: lines of argument such as "but you can not complete a manual so" are not really relevant, in the premises dell'apologo is the fact that from the point of view of sends symbols to man conversation credible.
In the second instance, define "Learn Chinese" differently from "being able to properly communicate in Chinese is quite complex, if not impossible. It follows that it can be argued that the brain behaves like a Chinese room
: after all, there is no mechanism to know the Chinese do not pass to the brain, which is subject to the laws of physics as the rest of the real (in fact one of the arguments in favor of the possibility of artificial intelligence that apparently says nothing prohibits a computer sufficiently powerful to fully simulate the operation of a human brain, neuron by neuron, is clearly a theoretical possibility, as there are those who
there is working ).
Now computers are more skilled chess players in human beings, they play differently, but if the result is valid, how is this relevant? A computer with "modules" so functional for things such as conversation or spatial perception is not smart?
forecasts of those who had hoped for faster results in artificial intelligence have been proven wrong, but believe that the reason lies in a theoretical impossibility I find is a sign of very scientific anthropocentrism, exceptionalism in relation to the human brain, ultimately
carbon chauvinism. The problem is in software engineering and related to the algorithms to identify the fact that chess is vastly easier than doing it for other human activities.