How Should We Approach Generative AI? | Fumio Shimpo, Assistant to the Dean and Professor, Faculty of Policy Management
June 27, 2023
If we cannot effectively utilize generative AI, we will not be able to survive in the AI era. However, there are various debates about the pros and cons of using generative AI in education.
In the discussions surrounding its use, we often see debates about whether to permit the use of generative AI at all, but are such discussions not now meaningless?
This is equivalent to imposing conditions on a report assignment such as "all submitted assignments must be handwritten" and "research for materials must be done by visiting a library to check the original documents," and stating that failure to comply will result in not being graded—all this despite the fact that we can create documents by typing on a computer with word processing software and research various information on the internet.
Now that an environment where generative AI is available has emerged, what we should be considering in education is not a debate about prohibition or restriction but rather how both faculty and students can utilize generative AI. The ability to ask questions of generative AI will become a new form of "national language" proficiency, and the ability to communicate using generative AI will become a new "language skill." If that happens, the time may come when "AI language (prompt) proficiency" is added to entrance examination subjects, just like foreign language proficiency.
Now that a new technology has become available, is not what is required of education precisely to learn how to master it?
What is being questioned in education in the AI era is how to effectively use new technologies, including generative AI, and how both faculty and students can efficiently teach and learn while enhancing their own abilities through their use. At present, we are still in an exploratory stage of figuring out how to use generative AI in the first place, so everyone is having fun inputting various keywords and clumsy instructions (prompts), even entering their own names, and then laughing when the information is incorrect. Eventually, however, those who cannot master generative AI will be the ones who are laughed at.
When the use of generative AI becomes widespread and it becomes commonplace to use AI to create documents, text, images, and audio, it is conceivable that a low ability to use it—in other words, low generative AI utilization ability (AI literacy)—will become a fatal disadvantage in the coming AI era.
On the other hand, it is difficult for users to spontaneously consider legal and ethical issues. Nevertheless, we see opinions suggesting it is the user's own responsibility, but that is an unreasonable expectation. Researching these issues and indicating a direction is my specialty. Regarding the initial thoughts on this, I have published an article titled "Generative AI and AI Regulation" in the June 2023 issue of Mita-hyoron (official monthly journal published by Keio University Press) , which I encourage you to read.
Fumio Shimpo, Assistant to the Dean and Professor, Faculty of Policy Management, Faculty Profile
*Portrait photo (on the Top Page ) by photographer Kikuko Ujuyama