Artificial intelligence as science

Artificial intelligence is the sphere of computer technologies that investigates the formalization of tasks that remind the tasks solved by people. The algorithm of the tasks solving is unknown beforehand. Sometimes there are also the tasks with the images taking that are traditionally included into the sphere of artificial intelligence investigation. As you see the definition doesn't contain any "thinking instead of people" and "the analogue of people's brain", etc.
There is no exact definition to this science as the problem of the human intelligence essence and its status is not solved in philosophy. There are also no criteria of the computer's intelligence; however there were some theories and hypothesis on the computer intelligence evaluating. Nowadays there are a lot of principles in the tasks of the artificial intelligence understanding and the creation of the intelligent systems.
According to one of classifications there are two main approaches to the artificial intelligence development. One of them is semeiotic that means the creation of the analytical systems that model the developed psychological processes such as thinking, talking, emotions, art, reasoning, etc. The second approach means the investigation of the neural networks and evolutionary computation that design the intellectual behaviour on base of the small "non-intellectual" elements.
This science is connected with such sciences as psychology, neurophysiology, medicine, etc. As all computer sciences it uses the mathematical apparatus. Philosophy and robotics are especially important for this science.
