Herbert A. Simon (June 15, 1916 - February 9, 2001) - American economist, political scientist, and theorist of social sciences. In 1978, he was awarded the Nobel Prize in economics for being one of the most important decision-makers in organizations.
short biography
Herbert A. Simon was born in Milwaukee, Wisconsin. He studied at the University of Chicago, who graduated in 1936 and received his doctorate in 1943. He worked as an assistant at this university (1936-1938), as well as in organizations related to the management of state bodies. Including the International Association of City Managers (1938-1939) and the Bureau of Public Administration from the University of California at Berkeley (1939-1942), where he managed the administrative measurement program.
After this professional experience, he returned to the university. He was an assistant professor (1942-1947) and professor (1947-1949) of political science at a technological institute. In 1949, he began to teach administration and psychology at the Carnegie Institute of Technology. And after 1966 - computer science and psychology at Carnegie Mellon, which is located in Pittsburgh.
Herbert Simon also spent a lot of time advising public and private institutions. He, along with Allen Newell, received a Turing Award from ACM in 1975 for his contribution to the development of artificial intelligence, the psychology of human perception, and the processing of certain data structures. He won an award for outstanding scientific contribution to the development of the American Psychological Association in 1969. He was also appointed an honored member of the North American Economic Association.
Theory of Bounded Rationality
Consider Herbert Simon's theory of bounded rationality . She points out that most people are only partially rational. And the fact that, in fact, they act in accordance with emotional impulses, which are not completely rational in many of their actions.
Herbert Simon's theory states that personal rationality is limited to three dimensions:
- Accessible information.
- Cognitive limitation of the individual mind.
- Time available for decision making.
In another work, Simon also points out that rational agents experience limitations in formulating and solving complex problems and in processing (receiving, storing, searching, transferring) information.
Simon describes a number of aspects in which the “classical” concept of rationality can be made more realistic to describe the economic behavior of real people. He gives the following tips:
- Decide which utility functions to use.
- Recognize that there is a cost to collect and process information and that these operations require time that agents may not want to refuse.
- Assume the possibility of having a vector or multidimensional utility function.
In addition, limited rationality suggests that economic agents use heuristic methods to make decisions, rather than strict optimization rules. According to Herbert Simon, this method of action is due to the complexity of the situation or the inability to process and calculate all the alternatives when the processing costs are high.
Psychology
G. Simon was interested in how people learn, and together with E. Feigenbaum he developed the EPAM theory, one of the first theories of learning, which was executed in the form of computer software. EPAM was able to explain a considerable number of phenomena in the field of verbal education. Later editions of the program were used to formulate concepts and gain experience. With F. Gobet, he supplemented the EPAM theory with the computer model CHREST.
CHREST explains how elementary pieces of information form building blocks that are more complicated structures. CHREST was mainly used to implement aspects of a chess experiment.
Artificial Intelligence
Simon was a pioneer in AI, developing the Logic Theory Machine and Common Problem Solver (GPS) with A. Newell. GPS is perhaps the first method designed to isolate a problem-solving strategy from information about specific problems. Both software was implemented using a data processing language developed by Newell, C. Shaw, and G. Simon. In 1957, Simon stated that AI-based chess would surpass people's skills in 10 years, although this process lasted about forty.
In the early 1960s, psychologist W. Neisser stated that while computers can reproduce “stale cognition” behavior, such as thinking, planning, perceiving, and concluding, they can never reproduce cognitive behavior. Excitement, pleasure, displeasure, lust and other emotions.
Simon reacted to Neisser's position in 1963 by writing an article on emotional cognition, which he published only in 1967. The AI community has largely ignored Simon's work for several years. But the next work of Broken and Picard convinced the community to focus on the work of Simon.