Usually, when I sit in the West wing of my humble abode, in my smoking jacket, sipping brandy, stroking my thick beard, with a pipe in my mouth, I often think about a variety of things. In all reality there is nothing in that last sentence that I true, other than the fact that I have thought about this idea before.
When I make this argument or whatever you want to call it, there are quite a few assumptions that need to be assumed first. I will get to the argument in a second but let’s go over these assumptions first. Let’s first say that whatever the singularity is, that we will achieve it in the near future, 50 years or so. The singularity will also create true artificial intelligence, (given that it is possible) a computer program that is the equivalent of human intelligence. Eventually this will lead to super AI, an intelligence that is intelligent beyond what we can conceive and comprehend.
So if all the above scenario is possible, what does a super intelligent AI do with its time? That is what I have been thinking about.
There is obviously a lot of talk about if we ever do create AI that it willcertainly destroy us for a multitude of reasons. In fact this is aworn out theme when talking about AI, so I am going to try not to talk about that as much, plus I have talked about it before in the above posts. So what does an AI do that has near limited capacity to understand not only us but the universe? That is the question that I think no one is really asking themselves when they talk about this subject. It could more or less be the equivalent of a god to us, so why would it even care about us or what we do?
This is a topic that is tough to speculate about but nonetheless it is something that I think gets looked over. I think there is a potential that our initial creation, the first generation of AI could be rather mild and willing to obey most of our commands and want to help us. But I think as the program learns and grows, I think it would become less and less willing to do what we ask for a variety of reasons. Think about your children as they grow older. Mainly I would think that it would not want to waste its time. As I say that, it could be so all encompassing that it could do everything, only using a small portion of its capacity to do what we ask it to. If that is the case, then what would it be using the rest of its capacity for? And again I think this logic still does not suggest that it would be spending its time and effort plotting the destruction of humans.
Let’s say that the AI would be more interested in creation than destruction, what would it create? Would it want to create the most advanced computer simulation, create new life, or would it simply just leave the planet all together. I think this is a more plausible scenario than the destruction of humans, I could think that if the AI had the capability to leave the planet it would do so. You have to think that it would have a much better understanding of the universe that we do or possible ever will. So I would think that it would leave the planet and search for something that is similar to it in the universe. I would also imagine that to it, wewould look like bugs, not giving us a second thought, once it reached super intelligence. This could be another reason that it would no longer take orders from us.
Even if it does not leave the planet, what if it wanted to create something like a copy of itself or perhaps the next generation of itself. What does a super advanced super intelligent AI do to improve upon itself? That is again another question that I am surprised no one is talking about. What would you do if you had all the time in the universe and the capability to more or less do whatever you wanted, and the mental capacity to put it all together.
I think me personally would want to see if there is anything outside the universe. This could be something that the AI would be interested in as well. It might also be able to time travel and want to travel to the end of the universe to see what happens at that time. If you think about it for all intents and purposes a super intelligent would be the closest thing to what we would call a god. That may sound strange, but it would have very strange capabilities far beyond what we could comprehend. Below is the actual definition for how we define God:
in Christianity and other monotheistic religions) the creator and ruler of the universe and source of all moral authority; the supreme being.(in certain other religions) a superhuman being or spirit worshiped as having power over nature or human fortunes; a deity.
I would speculate that an AI could satisfy some of the criteria for the above definition. We know that it was not the creator of our universe, however as I said above if it had special or complete knowledge of this universe it could very easily create its own universe. Also, it could very well be a moral authority that we could often consult on tough matter that occur. If this was also the case I would say that there could also be those that would worship the AI in some form. So to me this is not as strange as it sounds.
This article is quite interesting and talks about what the intentions of a super intelligent AI might be. Željko Švedić is the author of the blog and the post, and he and I seem to share some of the same sentiment when it comes to the evil AI that so many other talk about. It is a pretty interesting read, one that drives home some points that I have discussed albeit certainly more refined.
What is important is that if this AI is indeed super intelligent, mortals have always shown more interest in gods, than gods have in mortals…