In my thesis, I am examining the future of Artificial Intelligence which emerges as a contemporary topic. There are many philosophical arguments regarding the possibilities and limits of Artificial Intelligence. Most of them are valid and persuasive, but they have a serious weakness that underestimates the potentials of engineering and presupposes anthropocentrism. I enquire two central topics on the limit of A. I.: the problem of meaning and the problem of learning ability. I contrast the result of the enquire mentioned above with the vision of engineering. With that, I would like to stress the fact that the future of A. I. may turn out to be different from what philosophers envision. Main theme is that studying of A. I. is not the same with simulation work. The study of A. I., whose aim is to simulate human intelligence, is not only unsound - scientifically or technologically. It is also capable of preventing us from having a universal or cosmic point of view about intelligence. In case we overcome the absolutism and mystification of human intelligence, we can easily comprehend implications of highly unfavorable future of A. I. Eventually, we can heartily accept the possibility of the A. I. to excel human beings, and get an idea that the process might be a natural course of evolution. At this point, we can discard the old dichotomy of the natural and the artificial, and prepare to accept the change in the definition of the human species.