Do computers have souls? In the next 50 years it is almost certain we will have a general purpose AI that passes the (strong) turing test. An android that requires a specialist to distinguish it from a human being. Now if no soul exists, or a soul is a convenient term to describe what makes a person a person but is fundamentally reducible to matrix math; then there is genuinely no mystery here. Lets assume instead that a soul does exist. If the android seems perfectly human, but is entirely mechanical, then we need to wonder what role the soul plays in human consciousness. The only avenue for a soul to interact with a human that seems to be physically realizable is through Quantum Mechanics. So, one might posit that the robot has a quantum random number generator that influences its decisions. We would then have a way of testing the soul hypothesis. If the android acted more human with a QRNG than with a pseudo-random number generator, wed have strong evidence of some sort of soul like thing. If on the other hand they seemed roughly the same, it would indicate a strong likelihood that the soul doesnt exist (or at least can have no immediate influence). Interestingly, Turing himself, wondered about the capacity to distinguish based on ESP. And proposed that if a computer could get the same percent hits as a human, youd have to call it conscious. But, continuing from the hypothesis that there is a soul and it interacts via quantum mechanics; at what point does such a thing possess a soul? If for example we base a morality upon whether something has a soul, at what point is there enough soul to argue for rights? Is running a double split experiment a massacre? Does plugging a QRNG into my computer mean its a person? On the other hand, suppose that the model appears perfect without interacting with the Quantum World. What does that mean for our moral imperative towards the machine? And towards other humans? If we assign the machine rights, is the Turing Test our litmus? At what point is a machine entitled to rights? Why do we grant the hypothetical AI rights but not my laptop? Finally, what about a person that has half their brain replaced with a digital brain? Removing half a persons brain is a real thing we do now (for cases where seizures would prove deadly to the person and splitting the corpus collosum seems insufficient). If that person received a digital brain replacement (even one that doesnt simulate a health human half brain); is that person a person or an machine? If they are a machine, what is the limit by which we say it is a person again 90% machine, 99% machine? Nick Halderman, Cydian Rake, Jeff Tucker
Posted on: Thu, 17 Apr 2014 17:54:53 +0000
Trending Topics
Recently Viewed Topics
© 2015