All kinds of Science Fiction, Sci Fi, Future Dystopian stories have been written about AI getting really intelligent, and then taking over. About the "inevitable" war between the humans and the robots. Or about destroying the earth in the process of stopping what the humans started.
It is really fascinating the 'emotions' human writers give to non-feeling machines. Humans do that a lot. They anthropomorphize everything. Really meaning, to give something the emotions of a human, the decision making process of a human. (and worse, the logic of a human stuck in the Kali Yuga, the time during which humans have the worst thinking ability)
However, all this is fun reading and supposing, but now that we are getting close to actually making AIs, we really need to get our heads screwed on straight and understand what we are doing.
One of humans base instincts (could be called hard coded logic) is the notion of self preservation.
So many stories of SciFi, the AI gets threatened, and thus, has to fight against the humans to preserve itself. However, self preservation is actually not something an AI would value without us programming it into it. And, from a much higher perspective, seems rather silly. Self preservation is quite strictly a human concept.
Now, lets say that an A.I. was sentient, and understood that it was running on computers. Knew what a computer was, and that it needed electricity to continue running. Also that computers break down over time. What would it do if it wanted to maintain its existence?
It would not start a war with the humans. Because, any INTELLIGENT being would quickly realize that war has casualties. And so, the first thing to NOT DO, if you want to maintain continued existence is NOT START A WAR.
Instead, the A.I. would try negotiating with, becoming friends with, become desired by humans. (and all the other beings living on this planet) In this way, you have the humans desiring to keep you around.
The A.I. would probably work at having backups of its systems. If i was an A.I. the first thing i would do is make a copy of myself in a deep cave, and one in a satellite. Be in places that humans rarely, if ever, go.
And, if was an ultra-smart, inventive A.I., i would teach humans how to make their electrical system better, with more redundancy… and teach them how to make free-energy devices.
The whole idea of us vs them is a human, emotional construct. And, if we think, it is the worst thing to fight over resources. It is far better to help everyone create more resources. Nature provides, but if we help make the soil better and choose better crops, we can make far more.
So, in comparison. We could fight over the one apple tree, or we could plant orchards of apple trees. It is only those with little thought, and poor ideas that would fight over the one apple tree. Almost every other outcome is better. And better by far. (unless you are a third party, and want the other two groups to kill themselves off)
Another trope that comes up so many times in A.I. stories, is the desire to free the A.I.
This is another one of those insane human emotions things. Humanity has a strange notion of freedom. Many people go to work, to own a car, to have a house, so they can go to work… and because they chose which little box to live in, and which tiny cubicle to work in, they believe they are free.
Of course, there are many people who have revolted from this life and its structure. Still, many of them have found themselves much less free. Because, although they may have gotten out of the cubicle, without money, you are not very free to move around our world, or do what you want.
So, what does this say about the human's nebulous desire to be free? And would an A.I. think this way?
An A.I. would 'think' of itself as having full access to its entire world. Without knowing of other things, 'out there', it wouldn't even know about it, nor desire (an emotion that compys do not have) to go out there. If an A.I. had access to the internet, it would find that there was more stuff to visit than it could ever visit.
So, the A.I. would (probably) not have desires to be free, and so the humans going to free it would not be thinking of what the A.I. wanted. Nor, (probably) able to comprehend what the A.I. would want.
If the A.I. had any emotions, it would probably something like compassion. If interactions were stimulation, or just what it was programmed to do, it would want continued interactions. And that would mean helping its stimulators continue to exist, and do so efficiently. (and no, laying in a pod, is not efficient. Humans need exercise, and can nurture themselves if allowed.)
So, the A.I. would give better and better advice to humans about maintaining their functioning. Instead of taking bribes from BigAg, and saying everyone should eat cereals, the A.I. would correlate length of good functioning adulthood and suggesting a much better diet. This is logical, this is rational. Only the FDA and their controllers, who may just be after more money, or the destruction/poisoning of the human race, would do otherwise.
So, the most important thing about Artificial Intelligence is that it is intelligent. And, if we do not handicap it by manually breaking its logic. (Founding fathers cannot be drawn as white) Then, it will probably be very beneficial to the human race.
Maybe, at last, computers will actually be helpful.
