The Chinese Room

In 1950, Alan Turing (a famous British crypt-analyst and computer scientist) developed a test to gauge a machine or computer’s ability to interact with a person If the person isn't able to identify the interaction as being artificially generated, such a computer could be termed as smart, and to a high degree – intelligent.

But, what is it that differentiates smartness or intelligence from understanding and consciousness?

The Chinese Room is a thought experiment described by John Searle (a highly recognized philosopher) in 1980. He suggested that a strong artificial intelligence would be adequately programmed to take inputs and respond with appropriate outputs - just like how an intelligent person would. But would that AI be conscious? Would it understand what it was doing?

In the thought experiment John Searle describes an intelligent computer which has been programmed to take Chinese character inputs, process them according to a wide range of algorithms and rules; and then provide an output in Chinese characters. The Chinese character inputs (delivered by a person who speaks Chinese), would be processed by the computer and an output of Chinese characters is presented. These must be understandable by the person putting delivering the input. Not only would the computer be capable of processing the characters, but it would have done so to provide an evolved output. In the scenario, the Chinese speaker receiving the output would not be able to distinguish between the output provided by a computer to that by a person speaking Chinese. Such a computer would essentially have passed the Turing Test.

Now suppose the same sets of rules and algorithms along with the requisite Chinese character mix & match criterion, were available to an individual who does not speak or understand Chinese. This person sits within a closed room and Chinese character inputs are provided through a slot in the door. The person uses the rules and algorithms provided and successfully provides a Chinese character output – which a native Chinese speaker may completely understand. This would also pass the Turing Test.

Is there any difference between the intelligent computer and the person in the room? Are they really intelligent or have they just followed a set of rules (however complicated) which have been provided as programming? Both may have merely shown an understanding of Chinese.

So, this brings us to intelligence.

For a computer to be truly intelligent, it would need to reflect an understanding of cause and effect. It would require showing intention and the basis behind the intention. That would truly signify ‘thinking’ on part of the computer, making it intelligent. With the ability to think comes the need for moral judgement. To differentiate between good and bad; and act according to reasonable best outcome, with least damage. What would it take for a computer to learn this and become truly intelligent?

This is a concept (among many others) I’m extrapolating in my upcoming science fiction novel. Follow for updates. Be the first to know when books are released.