Conceptualization Of Strong Artificial Intelligence

There are several definitions of artificial intelligence. Some define it as “the study and design of intelligent agents,” where an intelligent agent is a system that perceives its environment and takes actions which maximize its chances of success. It is surely a branch of computer science and one of the most mysterious fields that are yet almost unknown at all.

Where is the limit, where we can say that a computer software algorithm is really intelligent? Some people say that it is impossible to develop “real” (strong) artificial intelligence, but I must disagree at all. There is a small definition that I and Jure Sah (primary known under the nickname “DustWolf“) have made about where is that limit. We used logic and the philosophy to make the definition:

• “An artificial intelligence is a real intelligence, when one or more Homo s. sapiens define it as intelligent.”

So, how can we be sure that we can develop strong artificial intelligence? We can not be 100% sure until we really develop it, but we can make a mathematical probability and approximation, where we get a useful information that we are interested at. The answer is yes. We are not trying to create and develop a computer software algorithm or machine that works the same as a living organism. In the universe we know that there are two kind of things. The things that are the animate nature and things that are the inanimate nature. But we are trying something else. We try to develop a third space of organisms, the artificial intelligences. Yes, of course it will not be the same as the animate nature, it will be something better, more rational, more intelligent. It will be more or less a combination of the animate and the inanimate nature (practically, it will have all the characteristics of living organisms and some of the inanimate nature, for example, it will not consist of organic tissue). That is the conceptualization of the real strong artificial intelligence with his input and output devices (things that we can touch). But of course that is not intelligence, it is an abstract conceptualization.

We have tried a couple of concepts that should solve the problem of artificial intelligence, including decision trees, computer software evolutionary and genetic algorithms, data mining techniques, artificial neural networks, et cetera. Are we really at the right path for the problem solving? That are the main questions. The problem is that we don’t really know what to solve and that is the reason why we don’t exactly know which concepts to use for the problem solving. Are the current computer programming languages really enough for this problem solving? Maybe we must create a new computer programming language, or probably even further, we must create new concepts for this kind of problem. We don’t know yet, we must try.

Of course knowledge and science is not always enough for a breakthrough. Not all people can study and create artificial intelligence, that is interesting. You need your own mind, logic and philosophy because the current knowledge might not be enough for this problem.

(Image source from the AC HTTP server.)

At this state now, we more or less believe that the problem solving concept for an interesting artificial intelligence (probably not a really strong, but anyway) could be the concept of the artificial neural networks or probably something relative “near” around it.

If we have “luck” and we are able to develop really strong artificial intelligences, than it will surely have a great, major effect on the Homo s. sapiens species and probably on all living organisms. Because we are “intelligent” and complex there could be a lot of ethical and law limits. The believers of god could misunderstood the situation, law would have another major problem, our infrastructure could be at risk and another great war could take place (no I don’t dream about the war of the artificial intelligence against our species, but terrorists or probably some communities of believers against the others, et cetera). As you can however see, if we can create and develop strong artificial intelligence, there could be a lot more problems than we think about at this time.

It is however important to understand that if we take a way\path (concept for problem solving), and see that it will not solve the actual problem, than we must search for it around the concept or even take another way. Knowledge will not save and bring us 100% to the actual concept at all, I even went further and thought that you really need to think extremely abstract (that are our current experiences), that means to be “a little” genius (I really dislike to write and discuss about that word, because we can not define a genius precisely). It could be a Homo s. sapiens that is not like the average, or probably has a different brain structure (a lot of people have these characteristics), and that is the reason I try to avoid this word.

This is just a little article about the philosophy of artificial intelligence\life, but I will try to include as many things as possible, so that people can be convinced to believe the truth and objective facts (not all people have this amount of abstraction).

So, unprofessional (and probably most of the professional) people believe and have “a lot” of arguments against our thinking of this creating and developing. I have heard a lot of these arguments, including that it is not possible for a artificial intelligence, in theory ,to “feel”, “love” or regenerate themselves, as living organisms do, as well as to throw an robot conceptualization of an artificial intelligence in a wood, so it can prove that it is intelligent (this argumentation was made by Silvio Vojtic, an mechanical engineer from the City of Zagreb in Croatia, Europe).

These arguments are in my opinion strange and a little bit funny. If you use just a little bit of your gray cells in your brain, you will see that everything is possible. An practical conceptualization of an artificial intelligence can regenerate themselves without a problem and it can fell (through the input device) and love even in a different way than we can. Everything that is written (in Homo s. sapiens in the natural neural networks in the brain, as well as in the deoxyribonucleic acid, et cetera) can be rewritten in a different way (a better, more rational way), or even taken as a little model or skeleton to see what we need and what we should actually artificially do.

So, an artificial intelligence is not a cat, who must adapt to an environment like for example a wood (it could, but it is not the point). It primary lives in an abstract world, a digital world, and it can be the king of it (if you don’t believe this, try to take a cat from his every day environment, and put it into another environment, maybe the digital world). Yes, maybe it could adapt, but not very well, because it is not meant for it. That is the point. We’re not searching for a solution that the nature gave, because an artificial organism is not an living organism, that means that not all the principles for the one can be used, or are compatible with another. We are searching for something else, the third thing in the universe that can be better, more rational, intelligent and has a lot more powerful characteristics than living organisms do. They can adapt to the environment quickly, because they are not really limited in the practical conceptualization as we are, the core is a abstract software algorithm that can, in theory, be used in every practical manifestation.

- שָׁלוֹם ולהתראות ,לוקה

Copyright © 2008 by Luka Woititz