Philosophical Thinking In The Development Of Artificial Intelligence

From John McCarthy proposed the concept of artificial intelligence the on Dartmouth Conference in1956, within short50years, the paces of development of the artificial intelligence shocked the world. In the21st century, the achievements of artificial intelligence and the application has been throughout our daily life, all aspects of industrial production and scientific research. Small as voice search software in smart phone, large as intelligent robots on industrial production line; simple as automatic washing machine in millions of households, complex as the supercomputer which could make over10million operations average per second, artificial intelligence greatly enhance human cognition and the ability to change the world in all aspects. Powerful artificial intelligence manifested far more than human in some ways, and become an integral part of the human society, continue to change and promote the process of development of human society.Although development of artificial intelligence is fast, but research of artificial intelligence also encountered some difficulties witch is bound to affect its future progress if not be resolved. In order to improve the development of artificial intelligence, to make it better to serve human, we must recognize artificial intelligence essentially, analysis the trace of its development, found problems on the road ahead, to find the way to solve these problems. This paper a comprehensive analysis of previous research results, from a philosophical point of view, the use of the method of dialectical materialism and system analysis, starting from "what is the nature intelligence", discuss the origins of intelligence; analyze the Development of Artificial Intelligence status quo from faction of artificial intelligence, weak artificial intelligence and strong artificial intelligence, artificial intelligence, the development of technical basis. I think the current artificial intelligence development bottlenecks encountered:how to make machines with human recognition knowledge, learning, reasoning and correction to reflect human intelligence, creative thinking-such as human at all times in the initiative amendment, under natural conditions, the ability to adjust their behavior, or take in the case of insufficient information obtained certain tentative choice, or that the information obtained to the conflict between how the trade-offs in order to take the appropriate action.From the philosophical point of view, the future development of the Artificial Intelligence (AI) will focus on the following aspects, if it calls for essential breakthroughs. Firstly, machines must be made to use natural human languages, to communicate with human beings and to identify the natural language in different contexts. A variety of ambiguity apparently exists in natural languages. Until now, the AI has been able to analyze the significance of the single word, while it shows some weakness in identifying the true meaning of long sentences of natural human languages. To solve this problem, it needs to refine and logically analyze human languages, which cannot stick to the surface of the language but to understand the connotation of it. In the20th century, philosophy has given birth to such significant theories as Idealism and Speech Act Theory. This paper aims at analyzing the problems that the AI must solve from the perspective of Pragmatic Reasoning——the core of communicative logic of human languages. Secondly, it is another difficulty to enable machines to possess creativity——the essence of human intelligence——and to make uncertain reasoning. Creativity is the essential characteristics and the highest form of human intelligence. As long as human beings participate in creative activities, they must exert deductive reasoning. More importantly, they must have capacity to make inductive reasoning, analogical reasoning and fuzzy inference. As a vital methodology of understanding and reforming the world, incomplete inductive reasoning has achieved success in human practice. How to equip machines with this kind of ability is bound to be one of the priorities of the future research. On the other hand, there are countless vague phenomena in the world which become fuzzy concepts and propositions in human brains. To achieve the level of human beings, the AI must be able to deal with various kinds of fuzzy concepts and propositions. Fuzzy logic lays a foundation for it and relative studies on this field have achieved momentous results. In spite of the accomplishments, it is still strenuous to get algorithm extraction and the choice of Membership function. This is definitely the hot topic of AI research in the near future.