QUANTUM MIND

 QUANTUM MIND

                                                                   

                                    https://www.techtarget.com/searchenterpriseai/definition/Turing-test

                                   

THE TURING TEST 

      (from Sir Roger Penrose's The Emperor's New Mind

Manufacturer's claim that computers/ devices can "think, feel, sensitive and understanding, conscious beings" in which case, we will face immense moral difficulties. But how can we believe such probably absurd claims? The operationalist viewpoint is that "the computer thinks if it acts indistinguishably from the way a human would act when thinking".

This viewpoint was argued in an article by Alan Turing entitled "Computing Machinery and Intelligence in 1950 in the philosophical journal mind. The Turing Test was designed to decide "whether  a machine can be reasonably said to think". 

In the Turing test, there are two subjects: a computer respondent and a human volunteer respondent, both hidden from a perceptive interrogator (this means that the interrogator is blinded from both the parties). The interrogator has to decide which one of the two is the computer and which one is the human by asking probing questions to both of them. The computer is programmed or set to lie to the interrogator and convince him that it indeed is the human; whereas the human subject truthfully responds to the questions to persuade the interrogator that he/ she is actually the human. During these several question-answer sessions, if the interrogator fails to identify accurately and consistently which one is the real human subject and which is the computer, then the computer/ programmer will be considered to have passed the Turing test. 

The Turing Test itself has inherent weaknesses:

1) For example what if the interrogator asks the subjects some complicated arithmetical questions? A human being will take sometime to think and answer, but a computer which is programmed or set to make such calculations will give away the answer instantly. But at the same time there also exists the human "calculating prodigies"  like Alexander Aitkin and Johann Martin Zacharias Dase, who could multiply two eight or two 20 digit figures in his head in a matter of minutes. The interrogator might mistake such "calculating prodigies" for as a computer. The interrogator, therefore, has to come up with more complex arithmetical questions or make the computer stupider which is not fair for the computer. 

2) What if the interrogator decides to design and asks some probing "common sense" questions, questions of original nature which requires actual understanding. A human being would easily answer such questions, whereas  a computer may be able to answer at the beginning just as a human would but with sustained questioning, especially questions of original nature which require actual understanding, it might become evident to the interrogator that the subject lacks actual understanding, and is indeed a computer/ machine. The interrogator might also occasionally decide to throw in some nonsense questions for example, "a rhinoceros flew along the Mississippi in a pink balloon, what you make of that? If the computer lacks a proper understanding, it might soon get trapped in such questions and reveal itself. 

3) What if the interrogator runs out of original questions? And once the computer is fed with all the answers that a human had previously replied, it would simply answer those questions exactly as a human subject. That would be considered cheating on part of the computer.

4) What if the interrogator is biased? There is a danger of being biased on the part of the interrogator. What if the interrogator somehow perceives that the particular subject is a computer, but still reluctant to accept to the computer's attribute of understanding even though the interrogator is convinced by the nature of the computer's replies and that there is a conscious presence under these replies.

Sir Roger Penrose writes that he is prepared to accept the Turing test as a valid one "If the computer were able to answer all questions put to it in a manner indistinguishable from the way that a human being might answer them- and thus to fool our perceptive interrogator properly and consistently-then, in the absence of any contrary evidence, might guess would be that the computer actually thinks, feel, etc." The writer wants to make it clear that contrary evidence do not mean that the computer will consists of neurons and blood vessels just as the human brain, but transistors, wires, etc. 

The writer believes that in future a coherent and appropriate physical theory might be developed which will be consistent in a way with the rest of physical understanding, and the predictions correlate precisely with human beings' claims, and the computer itself manifests signs and prove those claims that it actually "thinks, feels, sensitive and an understanding, conscious being"!