Herbert Simon: A Foreigner Who Visited the Juku

2017/06/06

Dr. Simon giving a lecture during his visit to the Juku

A pioneer of Artificial Intelligence (AI). A computer scientist and psychologist. He also received the Nobel Prize in Economics as an economist (1978). Originally, he earned his PhD in political science. He also influenced management science, linguistics, and sociology.

"What is your specialty?"

When asked that, Dr. Simon would always laugh and say, "Well now."

On August 18, 1979, he gave a lecture at the Mita Campus titled "The Birth of Cognitive Science," moderated by Yūichirō Anzai. It was an ambitious attempt to merge multiple fields and open up a new academic discipline.

While Dr. Simon was involved in various academic domains, he consistently pursued the question of "how humans solve problems." In the 1950s, he stunned researchers of the time by announcing the "General Problem Solver," a system capable of solving any problem. He replaced every problem with a "search problem"—finding one correct answer among many candidates—and proposed a general-purpose computer program for that purpose. In other words, no matter what the problem in the world was, as long as it could be properly defined, a solution could be found using this program.

The important thing here is that Dr. Simon did not necessarily aim to find the "optimal solution." This is because humans are happy if they find a "satisfactory solution," even if it is not the optimal one. For example, in a supermarket parking lot, you look for a space as close to the entrance as possible, but if a reasonably close spot is open, you will likely park there and be satisfied. Even though a closer space might be available, you do not try to find a better spot. This is because you are satisfied. He argued that to find a satisfying solution, humans do not use sophisticated mathematical optimization methods, but rather engage in rough, ad-hoc thinking. It is a theory that strikes at the very essence of human nature, and being told so clearly is actually quite refreshing.

Stanford University, the Massachusetts Institute of Technology (MIT), and Carnegie Mellon University (CMU), where Dr. Simon was based, were considered the "Big Three" of AI research. While research at Stanford and MIT was an extension of computer science, the mainstream at CMU was research that merged psychology and computer science, placing importance on the rough sensations of humans. Dr. Simon later named this "Cognitive Science."

Looking back at history, when computers emerged in the 1950s, everyone thought that their overwhelming calculation speed and memory capacity would surely surpass human intelligence in the near future, and great expectations were placed on the field of Artificial Intelligence (AI). For example, it was thought that automatic translation from French to English would be easily achieved in two to three years, and regarding chess, Dr. Simon predicted that "it will defeat the human world champion within 10 years."

However, even after the mid-1960s, automatic translation showed no signs of completion, and computer chess programs were far from reaching even amateur levels, let alone the world champion. As a result, Dr. Simon was at one point called "Simple Simon" (the braggart).

Thus, in the late 1960s, AI was effectively branded a "failed discipline." It was deemed that there was no hope for practical application in the near future, and as a "useless discipline," research funding was significantly cut, leading to a winter period. However, while many AI researchers changed fields, CMU, led by Dr. Simon and others, did not lose heart and continued to vigorously pursue AI research into the 1970s. To begin with, Dr. Simon did not use usefulness or the likelihood of practical application as his motivation for research. His driving force was an intellectual curiosity to understand exactly how the human thinking process works.

I myself felt sympathy and interest in the philosophy of CMU, and after graduating from the Faculty of Science and Technology at the Juku, I went to study abroad at the CMU graduate school. I traveled to the U.S. with what was, for me at the time, the incredibly large dream of "getting a PhD in the field of Artificial Intelligence."

The first time I met Dr. Simon was in the summer of 1981, when I entered CMU. First, I sent an email to the doctor to introduce myself and make an appointment, and the next day, a reply came from Dr. Simon himself. To my surprise, the first line was written in Japanese using Roman letters.

"TOMITA-SAN, YOKU IRASSHAIMASHITA" (Mr. Tomita, welcome)

The rest was in English, but it said:

"Please come to my office next Tuesday at two o'clock." The following Tuesday, I visited Dr. Simon's office with a somewhat nervous expression, and a gentleman with a fine build in his early 60s greeted me with a smiling face. Then, in broken Japanese, he said,

"KONNICHIWA, HAJIMEMASHITE, Herb Simon desu" (Hello, nice to meet you, I am Herb Simon) and held out his right hand. After that, in English, he said,

"I used to be able to speak more Japanese, but I haven't used it for a long time, so I've forgotten a lot. It's very unfortunate," he said with a shy, wry smile. On a section of the wall behind where the doctor was sitting, a certificate was displayed. Looking closely, it was the Nobel Prize certificate. It was impressively modest. It was in a small frame, hanging casually on the wall. As I stared at it,

"Let me show you something more interesting," he said, and showed me something. It was a nameplate that said "Saimon Hābāto" (written in kanji as 宰問翔人). He said he received it when he gave a lecture in Japan. It is read as "Simon Herbert," and it supposedly means "a person who takes flight toward great problems."

Listening further, Dr. Simon said that in addition to Japanese, he spoke several other languages, such as German and Chinese.

"No, I should say I could speak them," he hurriedly corrected himself, then added,

"Studying foreign languages is fun. Besides, the mechanism by which humans speak is still full of mysteries. If we examine the structure of human language closely, we will find the key to unlocking the process of human thought," he said, in words befitting a master of AI. Suddenly, I noticed some handouts written in Chinese on his desk.

"What is this?" I asked, and Dr. Simon said shyly,

"I am taking Chinese classes at the University of Pittsburgh (adjacent to CMU)." It was so unexpected that a Nobel Prize-winning scholar would take classes at a university that I instinctively asked,

"Is there a special lecture for faculty?" He replied that it was a subject for second-year undergraduates and that about 20 other college students were enrolled.

"Since it's aimed at young people, the pace is fast and it's difficult. This handout is homework for tomorrow. If I don't do it by tomorrow, the teacher will be angry with me," he laughed mischievously. He was a teacher of a different sort.

Then, in the 1980s when I was studying at CMU, the boom in AI research arrived once again. In Japan, projects like the Ministry of International Trade and Industry's Fifth Generation Computer Project acted as a catalyst; the "Japanese Society for Artificial Intelligence" was born, practical AI systems called "expert systems" appeared, and various companies began to put serious effort into AI. Eventually, "AI products" such as "air conditioners with AI" and "electronic rice cookers with AI" appeared on the market. The strength of computer chess also defeated the British women's champion in 1988 and later beat the world champion. Dr. Simon's prediction, though it took time, was proven essentially correct, and he gave up the nickname "Simple Simon."

However, in the late 1990s, excessive expectations for AI turned back into disappointment. Although the fictional birthday of Astro Boy in the anime is in 2002, and AI technology was steadily progressing, it was far from the completion of an intelligent robot like Astro Boy. Research funding was significantly cut, and the second "AI winter" arrived.

Then, entering the 2010s, expectations rose once again with the appearance of "Deep Learning" and "AlphaGo," and the third AI boom arrived. Futurists predicting that "AI will surpass human intelligence in 2045" appeared, and AI is currently attracting a great deal of attention. Whether a third AI winter will come is unknown, but regardless, AI is certainly an "academic field with intense ups and downs in popularity."

In the midst of all this, Dr. Simon consistently and steadily advanced AI research, whether during booms or winter periods. This was because his motivation was not based on being useful or receiving research funds, but rather on pure curiosity: "I want to unravel the mystery of the human thinking process."

Dr. Simon's life, which came to a close in February 2001 at the age of 84, suggested very important and valuable lessons for a researcher: "Do not be bound by academic fields" and "Do not be misled by trends." And his attitude of never being arrogant toward anyone and continuing to learn throughout his life made him a truly great mentor whom I can respect from the bottom of my heart.