Simulating the Human Mind: The Greatest Possible Technological Achievement
I think if you were to ask people “What is the greatest technology the human race could possibly develop?” you might get a variety of answers ranging from faster-than-light travel, terraforming alien planets, or the ability to harness the power of an entire star. Someone might also suggest certain god-like powers such as eternal life, raising the dead, creating whole universes, or reshaping our own reality to our liking like the elves in Warhammer 40K.
In my opinion, however, the greatest technology the human race could ever develop, at least hypothetically, would be an artificial human being (or at least an artificial human mind) that was indistinguishable from the real thing. (I suspect that anyone with the expertise to create an artificial human mind would have no trouble manufacturing an artificial human body, since the human mind is probably several orders of magnitude more complicated than the external qualities of the human body). Such an artificial person would hypothetically be capable to achieving anything else the human race might eventually achieve. If you create something that can do everything you can, you’ve reached a plateau.
Of course, there is a spiritual aspect to the human mind that I do not think could ever be fully simulated through coding. I think it’s more likely that someone would create a tyrannical AI monster that exemplifies the worst qualities of whoever programmed it long before honest-to-goodness androids start appearing.
Certainly algorithms could be developed to simulate human emotions. Given specific, pre-programmed sets of conditions, a computer could spurt out a behaviour meant to simulate some emotion or another. Receive a compliment, express happiness; get ignored, express dejection; and so forth. This sort of thing has been around since the days of Dr. Sbaitso.
I have not studied this in any particular depth, but I think the primary obstacle right now for creating something that even vaguely approximates a human mind is how to simulate creativity and problem-solving. If this is possible, it would probably require massive memory storage capabilities, processing power much greater than anything that currently exists, and the AI’s ability to not only write new algorithms, but the ability to constantly reassess those algorithms and modify them on the go. The memory storage would be necessary for storing records of how other people solved problems, and the processing power would be necessary for adapting how other people solved those problem to whatever situation the AI is currently facing.
It is possible for a computer to “write its own code” or "write new algorithms", but the limitation is that it can only rearrange the code it already has. Just to be clear, I think the human mind probably operates in the same way--- so in order to get an AI to be indistinguishable from a human being, you would probably have to create enough algorithms to approximate all or most of human experience. I won’t say this is impossible, but it would be prohibitively time consuming to create all these algorithms. I think it is more likely that the Singularity would occur (and the human race consequently gets wiped out) long before anyone achieves this. The computers that emerged from the Singularity would be operating on whatever limited algorithms their human progenitors thought important to put into them, and given the leftist propensities (i.e. destructive, demon-inspired propensities) of the people working on AI projects in Silicon Valley, these supercomputers would probably destroy the world.
***
One tangential issue that I find fascinating is whether the AI programmers of the future would try to create androids that were capable of sin; or would they try to create sinless beings? If they were trying to perfectly simulate the human mind, then it would of course be capable of sin, or at least “simulated sin”. Creating an android race capable of sin, and consequently able to make war on the human race, would be extremely stupid; which would probably mean that the manufacturers of androids would try to make them "sinless".
This raises some interesting question: How much would you have to modify the algorithms that define human behaviour in order to eradicate sin? Or would you merely have to delete or suppress certain algorithms? Would the AI be capable of being “tempted”? Would the sinless android be in essence a simulation of Jesus, more like Adam before he fell, or more like Elrond?
Sound like inspiration for a good science fiction story to me!
