How to recognize the development of artificial intelligence

Artificial intelligence is a very broad field. At present, artificial intelligence covers many major disciplines. Because the disciplines are relatively scattered, most doctors, professors and other professionals engaged in related research are often only involved in one of the above disciplines, and even long-term focus on specific issues in a certain discipline. In the next article, Gerststein conducts a comprehensive analysis of artificial intelligence.

Six categories covered by artificial intelligence. Computer vision (temporarily including pattern recognition, image processing and other issues), natural language understanding and communication (temporarily including speech recognition and synthesis, including dialogue), cognition and reasoning (including various physical and social common sense) , Robotics (mechanics, control, design, motion planning, task planning, etc.), game and ethics (interaction, confrontation and cooperation of multi-agent agents, integration of robots and society, etc.), machine learning (various statistical modeling , Analysis tools and calculation methods). These fields are still relatively scattered, and they are currently in the process of cross-development and unification.

For example, face recognition is a small problem in the subject of computer vision; deep learning belongs to a popular genre in the subject of machine learning. Many people now equate deep learning with artificial intelligence, which is equivalent to describing a prefecture-level city as the whole country, which is definitely not appropriate. After reading this, students engaged in deep learning must be unconvinced, or very angry. Don't worry, after reading the following content, you will find that no matter how many layers the CNN network has, it is still very shallow and the tasks involved are still small.

The chronological phenomenon of the development of artificial intelligence. Due to historical development, since the 1980s, artificial intelligence has been divided into the above-mentioned university disciplines and developed independently, and these disciplines have basically abandoned the previous 30 years of research methods based on logical reasoning and heuristic search, and replaced them with Methods of probability statistics (modeling, learning). This kind of field differentiation and historical dating have objectively caused the current academic and industrial thoughts and opinions to be quite "chaotic", and the chaos in the media has been amplified. However, from a positive perspective, this situation does provide a good opportunity and a broad stage for the younger generation of researchers and graduate students to make contributions.

Put aside the fantasy of science fiction, and talk about a few recent specific applications. Everyone has heard a lot about unmanned driving, let's talk about military use first. A squad or action group in the army, for example, now requires seven people, but in the future it can be reduced to five, and the other two will be replaced by machines. Secondly, robots can be used in disaster relief and some dangerous scenes, such as the scene of a nuclear leak. People cannot enter and must rely on robots. There are many medical examples: smart prostheses or exoskeletons connect with human brain and body signals to enhance people’s ability to control their movements and help people with disabilities live a better life. In addition, there are service robots such as home care.

The research of artificial intelligence, in simple terms, is to extend and augment the ability and efficiency of human beings in various tasks of transforming nature and governing society through intelligent machines, and finally realize a society in which humans and machines coexist in harmony. . The intelligent machine mentioned here can be a virtual or physical robot. Different from the various tools and machines created by humans for thousands of years, intelligent machines have autonomous perception, cognition, decision-making, learning, execution, and social collaboration capabilities, which conform to human emotions, ethics, and moral concepts.

Guess you like

Origin blog.csdn.net/gstemG_maker/article/details/108663156