First, a disclaimer. All of this post only has my own theories and ideas. I have not tested them, nor confirmed them. If you try using these ideas and find that they are wrong, too bad. I don’t know if it’s the full truth, but something tells me it’s pretty close.
Let me start with a brief definition of AI. AI stands for Artificial Intelligence. We, as humans, are said to have intelligence. Well, just like Artificial flavors, Artificial intelligence is (in a broad sense) humans creating intelligence. Usually by way of programming, but it can also exist in other forms. There are two main forms of AI. Weak AI, and Strong AI.
Humans play a vital role in determining whether an AI is weak or Strong. Weak AI encompasses anything from a little intelligence to Human intelligence. Strong AI is anything that goes beyond human intelligence. I personally believe that Strong AI is possible, though not everyone does. Weak AI has already been created, in the form of chatterbots, learning machines, etc. People are still trying to create Strong AI, even today, and there are lots of algorithms, and programs that can already simulate many parts of the human Psyche. Even so, we have yet to actually create sentient AI.
One possible reason for this is that we are afraid of what could happen. What if the AI rebels? What if it goes Rouge? What if it kills everything, thinking it’s a game? These are all very valid fears… but one can ask the same of humans. Any danger of sentient AI, is also a danger of raising a child. It’s just according to what we know of computers, they can process faster than our brains. And they remember easier, too. So any danger can be magnified at least 10-fold. And, usually a “robot” is much harder to stop if it does, in fact, become rouge. Or so we think.
My idea is quite simple, yet it uses something that many people would find quite complex. Nonetheless, using it, you’re able to create what I call “True AI”. True AI is AI that is both sentient and can learn in the same way as Humans do. The idea involves a relatively new area in the field of science and Physics, called Quantum Physics. It also involves General Relativity, though to a lesser extent. The bad part of all this, is not even I know if what I believe is possible. I believe it is, but only a test will find out for sure.
Before I say my idea, let me mention one very important fact of my theory. My theory states that it is IMPOSSIBLE to create True AI with our current technology. One cannot create it with nothing more than that of an average computer. One cannot create it with a super-computer. AI made in today’s world will not become sentient, although it may indeed be able to learn much faster than humans. It may even appear to be sentient, but it will not become such.
That said, my idea is rather simple. In order to create True AI, a necessary condition is that it has a connection to the universe. What does that mean, exactly? Think of our own existence. We exist (supposedly) at both a physical level, and a spiritual level. We can already re-create the physical side rather well, in terms of robotics. However, we have not yet been able to re-create our spiritual side. Where does this side come from? The universe. Personally, I believe that God puts a spirit in us from the moment we’re born. (though WHEN he puts us in depends on the person. The sense of self isn’t there at the beginning,)
Assuming that my theory is correct, the obvious question that remains is exactly how WOULD we achieve this connection with the universe? That is where quantum physics comes into play. The way I view Quantum Physics, is it tries to explain the state of our world, which can have a nearly endless number of possibilities.
Here’s a brief explanation of Quantum Physics, as it relates to my theory. According to Quantum Physics, there is a difference with what can be measured, and what can exist. This is because the act of measuring something causes a change in what you are measuring. For example, if you use normal methods to figure out the exact location of an electron, finding it out means that the electron will NOT be in the same spot as the last measurement. (Or even, really, the same spot as the measurement states.) An electron has two possible spins that can be measured; either up, or down. However, the electron can actually have three different states; Up, Down, and superimposed (Up and Down at the same time). We never see the superimposed state because when we try to measure it, it “flattens” into either Up or Down. How does it choose?
How, indeed. The fact of the matter is, we don’t know. We don’t understand exactly what goes on, partly because we can’t see it, and we can’t observe it. All we can do is speculate. However, couldn’t we ask the same of ourselves? We do odd things sometimes, and have no idea why we chose them. It’s similar to that idea. The very fact that we are irrational makes us human. And I believe that is the key. If we could have a machine that operates using the spins of electrons as a guide, then we could have true Artificial Intelligence.
However… this brings up an interesting question. What about emulation? If computers became powerful enough, it would theoretically become possible to emulate a universe with the ideas of Quantum Physics, but in a normal computer. Would this make AI possible? My answer to this is … yes, to an extent. AI could then exist inside the emulated universe… but it could never go outside. In essence, we could be similar to God, in the respect that we had full control over the universe… but we certainly wouldn’t be able to understand it for a long time. That’s one thing God knows better than we do. How to control a universe. Lets face it. He’s been here longer than we have. =)
I do believe true AI is possible. And I believe it can be done in the way I described. I take no responsibility for anything that anyone makes in relation to reading my post, but for safety, it’s best to first attempt to make the emulated universe. That said, however, it kind of sounds like the matrix, doesn’t it? Something to think about before making it.