The headline of the January 5, 2025 Nikkei Veritas was as follows.

“The market survey said that stock price is likely to reach new highs this year.”

There are some market proverbs related to the Chinese zodiac. Readers might already know them, but I will leave them here for your information.

The Chinese zodiac has a saying like this: “The stock prices hit the ceiling in the year of the dragon and snake, and will then decline in the year of the horse. Be patient in the year of sheep, and the prices will go up and downs in the year of monkey and rooster. The year of the dog makes you laugh. Then, the year of the boar will be calm. In the year of the rat, the market tends to rise, but stumbles in the year of the ox, and the market is rough in the year of the tiger, but in the year of the rabbit, the market bounces up.”

According to the Nikkei Veritas survey answered by 72 market participants, “AI (68%)” was the most popular theme of the year. The expectations for AI-related issues, which drove global stock market gains in 2024, are likely to continue.

The words “artificial intelligence” and “AI” have become such common terms. Since the release of ChatGPT from OpenAI in November 2022, the term “Artificial Intelligence (AI)” has become a part of our society. The term “Artificial Intelligence” itself appeared in a proposal submitted to the Rockefeller Foundation by American researcher John McCarthy at a 1956 conference at Dartmouth College. The term “Artificial Intelligence (AI)” was used for “Logic Theorist,” a program designed to mimic human problem-solving abilities.

Artificial Intelligence (AI) research has been ongoing for more than 60 years, from the 1950s to the present. In the early years, the focus was on symbolic reasoning, or pre-programming rules for solving problems, but this was not able to solve real-world problems, and this led to the “AI winter”. In the 1980s, expert systems (computer systems that have knowledge in a specific area　and can reason and judge events like an expert) succeeded in some areas, but faced obstacles such as the inability to express tacit knowledge that humans normally can obtain from experience. Then the AI ​​industry once again entered a period of stagnation. However, as machine learning (repeatedly making inferences based on data and finding patterns without explicit programmatic support) developed from the 1990s to 2000, and especially the boom in deep learning (an advanced version of neutral networks) had a great impact on AI industry.

(By the way, when you visit an online shopping site, you may see recommendations such as “Would you like this product as well?” This is a typical machine learning algorithm called “collaborative filtering” that learns information about preferences and usage trends, and presents products that are likely to lead to purchases.)

(Figure 1: History of Artificial Intelligence (AI))

(Source: Wang, Lihui. (2019). From Intelligence Science to Intelligent Manufacturing. Engineering. 5. 615-618. 10.1016/j.eng.2019.04.011.)

The examples of services and technologies utilizing “generative AI,” such as ChatGPT, Gemini, and Canva, seem to be relatively new models in the AI field. The word “AI” used to refer to a type of discriminative AI*. However, the definition of “AI” is expanding to include “generative AI” that has the ability to create new content from the data.

(*Discriminative AI, as seen in character recognition and AI cameras, learns from large amounts of data in advance and is used in areas such as product quality checks and image recognition.)

One thing that we shouldn’t forget is that “generative AI” is a type of technology that uses deep learning, and deep learning is a part of machine learning, which is also only one field of AI. The definition of “AI” seems to change depending on who is talking.

(Figure 2: Relationship between AI, machine learning, and deep learning)

(Source: Ministry of Internal Affairs and Communications, “White Paper on Information and Communications” (2016))

Returning to the topic at the beginning of this article, there is no disputing the fact that interest in this “artificial intelligence (AI)” is growing worldwide. Do you remember when ChatGPT was released in 2022 and reached 100 million users in about two months? The craze has not gone away, and people’s eyes on artificial intelligence (AI) are still very hot. At the same time, the subject “Information I” was introduced in the Common University Entrance Examination in Japan for the first time, and it is easy to imagine that the basic knowledge required in this field will increase. In order to follow this AI period, we just should keep learning.

・

【References】

・[Iwata 23] Iwata, Yosuke, “Zukyu Pocket: Latest Generation AI Service ChatGPT Wakaru Hon (Illustrated Pocket: A Book to Understand the Latest Generation AI Service ChatGPT),” Shuwa System, 2023.

・[Ministry of Internal Affairs and Communications 19] Ministry of Internal Affairs and Communications, “2022 White Paper on Information and Communications,” URL: https://www.soumu.go.jp/johotsusintokei/whitepaper/ja/r01/html/nd113210.html.

・[Nikkei 23] Nikkei BP Mook, “ChatGPT & Generating AI: Saikyou no Shokugyutsu (The Strongest Work Technique of ChatGPT & Generating AI),” Nikkei BP, 2023.

・[Nikkei Veritas 25] Nikkei Veritas, No. 878, January 5, 2025.

・[Wang 19] Wang Lihui, “From Intelligence Science to Intelligent Manufacturing”, Engineering. 5. 615-618. 10.1016/j.eng.2019.04.011. 2019.

※The statements in this blog are not the official views of the Institute, but rather the personal views of the author.

Chancellery Unit, Group for Project Pax Japonica, Maria Tanaka