"Astro Forever" wrote:In the 1980 episode Outer Spaceport R-45, which was also in the manga and in the 1963 series I believe, there is the story of a man who created a robot son. To simulate the son getting older, he throws away one body after another, replacing it with a slightly more aged one. The son realized he was a robot only as an adult and has a hard time coming to terms with that, which is where the story is interesting.
I can't even imagine how he never realized he was a robot. He couldn't have hurt himself like a human would have, nor could he have caught a cold? Perhaps if he followed the human development, he could have thought everything was normal until age 3 or 4, but after?
What about Abullah in PLUTO?
Has anyone ever seen "the 13th floor"? I think it may be possible, though, perhaps unlikely, to program the AI such that anytime it thinks about such things, a satisfactory answer is always provided, though anyone else would not accept it. Even humans can deceive themselves & believe utter nonsense. Those who believe in bigfoot, will always find ways to explain away the lack of evidence, as will those who believe in alien abductions. Yet, they regard each other as fools, both because the other believes in something so silly, & because the other does not believe in what he believes in.
This, psychologists call the confirmation bias, & there are other terms that other fields use. We ignore dis-confirming evidence, and only consider confirming evidence. Moreover, We always find RATIONALIZATIONS that satisfy us that our beliefs are valid. I read a book called FLIM-FLAM, it detailed the attempts of dowsers, astrologers, & others to demonstrate supernatural powers under James Randi's scrutiny. They all failed, and many gave silly excuses for the failure. But they all agreed before the tests what results should indicate failure or success. Some were con artists, but some truly believed in their powers.
Randi even demonstrated the naiveté of those who believe in astrology by giving a group the exact same astrological profile. Each one read it, thinking it was specifically made for him only. RANDI then told them to exchange papers with the next person. They were all astonished to see that everyone had the same profile, though each one was certain that it was exclusive. Randi then identified himself, as he had been going about as a famous astrologer, told them they had been duped, etc. But many still believed in the famous astrologer, rationalizing that this was just a test of their faith.
Thus, I think it entirely possible to create an AI with a built-in system that deals with any scrutiny that might occur that it is not human. This might not even be necessary, if it is made to assume that some authority figure is infallible. As it grows from infancy, it is constantly reassured of these 'truths'; by the time it reaches maturity, it will have created its own rationalizations to protect its own beliefs.
hmy:
[/QUOTE]
