Conversation

AI Is a Mirror of Humans

Bonkwon Koo, Director at Human & Digital Lab (HDL; Department of Philosophy, class of 1984), and Eunsoo Lee, Professor at the Department of Philosophy

This new AI entity has been confounding us, and a conversation with intellectuals may be just the panacea for this confusion. With an agenda of "coexistence between humans and AI," HDL Director Bonkwon Koo and Philosophy Professor Eunsoo Lee shared deep and omnidirectional thoughts from the viewpoint of humanities and philosophy.

Bonkwon Koo

Director at Human & Digital Lab (HDL; Department of Philosophy, class of 1984)

Eunsoo Lee

Professor at the Department of Philosophy

Humans have acquired an unprecedented tool. I would like to inquire what the meaning of doing philosophy in the digital age is to these two experts who are exploring “digital humanities” by combining humanities and information technology (IT).

Professor Eunsoo Lee Digital humanities is a challenging field to define. In addition to various implications, there is even a dedicated website (whatisdigitalhumanities.com) because its definition is being updated whenever an issue arises. This field reflects an epochal trend that includes the task of incorporating humanistic reasoning into technology development as well as the mere utilization of technologies.

Director Bonkwon Koo A book I published in 2015, Robotic Age , Human Work , brought me a nickname, "digital humanist." This is because AlphaGo, which appeared in the following year, became a big issue, and I was recognized as having a philosophical, critical mind in the digital reality. The pre-digital world had changed sequentially and continuously. Our perceptions, worldviews, and social relationships have expanded in the same manner, whereas digital technologies have brought discrete and disconnected shifts. Nevertheless, our philosophical inquiries and thoughts have remained the same as those 2,500 years ago.

Professor Eunsoo Lee The latest discussions surrounding philosophy appear to repeat what the director has remarked. My external lectures are mainly related to the topic of “digital humanities as anthropology.” In the midst of still pursuing the correct answer, I view digital humanities with high expectations as a tool for properly learning anthropology. This is because we may be able to acquire various human thoughts and emotions that could not be previously captured with an enormous amount of data. Thus, I have recently been planning a project to convert the critical thinking of Seoul National University humanities students into data. I intend to re-describe human behavior, such as reading, seeing, hearing, knowing, and making through data collection. In the long term, I would like to determine the philosophical meaning of the AI revolution we are undergoing in the context of whether it is a mere mechanical invention or an anthropological transformation. Moreover, admitting that it is impossible to master all scientific discoveries and engineering technologies under meteoric development, I am striving to incorporate the unique viewpoint of a humanist by diligently participating in AI-related discussions. In the same context, I am making an effort to upgrade my curriculum related to the convergence of humanities and technologies in a systematic manner.

Director Bonkwon Koo Because all knowledge was divided into certain branches in the past, many experts have gathered and cooperated with functions within one field, while we are in an era of consilience. South Korea is no exception. It is difficult to understand problem awareness and methodologies related to AI without involvement in many necessary fields regardless of expertise, such as engineering, natural sciences, business, and humanities. We cannot create anything. The reason we teach ethics to engineers is that we fear AI.

As you have just mentioned, I am wondering about your thoughts regarding what we fear about AI,
what we would lose as we use AI, and what we need in response.

Director Bonkwon Koo I believe that the fear of AI arises because we misunderstand it. When a new tool emerges, it becomes an object of fear and awe as we feel ignorant of it and we deem it out of control. We even reach the sci-fi imagination that we would be ruled by the tool we created and soon destroyed by it. However, in fact, rather than AI ruling and destroying humans, those who know how to use the powerful tool called AI will rule and destroy those who do not. Ultimately, this inquiry leads to the matter of ownership, the question of who has the ability to control this powerful force.

Professor Eunsoo Lee I would like to add an issue of creativity to this story. Even when AlphaGo beat Pro Lee Se-dol, I thought that the ‘AI would excel in computational ability.’ However, ChatGPT created opportunities for AI to be creative if quality data are offered. However, just because we collect data that can capture human creativity for AI, and it creates something similar to what we call creative activity, we should not define this product as creative. Humans must be at the center of results. From now on, both imitation and productive activities generated by AI will become novel tools to enhance human creativity. For example, suppose that the only genre of art that I can express with my abilities is text. If this text could be converted into images or music, it would offer tremendous opportunities. Conversely, what concerns me most is diversity. Globalization has enabled people to reach a similar level of experience, while AI could lead people to a similar level of knowledge. As for now, the reports submitted by students would be similar. As a humanist, I worry whether many sub-optimal thoughts would disappear. It would be a horror story for all people to produce similar answers and ideas.

AI is a tool created to overcome various human limitations and flaws.
What would be the unique characteristics of humans in the AI era?

Professor Eunsoo Lee Humans with the possibility of making mistakes are finite beings approaching the obliteration called death. That is why humans are special beings. From the viewpoint of creating AI, a machine that resembles humans and ensures that there are no mistakes, it is necessary to know the unique capabilities of humans in more detail and to mull over what to add or subtract from these capabilities. With the advent of ChatGPT, AI has become a universal interest tightly interwoven with our lives. Building AI that greatly resembles humans may be an initial step forward. This may be our first stage of experience in the possibility of creating another being with intelligence. Furthermore, this experience provides our generation with the prime responsibility of creating AI that is essential to humans. Thus, the current phase is an exhilarating moment to me, when a completely new being can be born as we explore and design.

Director Bonkwon Koo A classical question raised by philosophy, that is, what is the meaning of being human, should be a renewed inquiry in this digital environment for contemporary philosophy and humanities. To create a being that strongly resembles a human being, humans must know everything about themselves. For this reason, humans are given the task of accurately understanding everything about themselves to create AI that is essential for themselves. This is why various observations and accurate investigations into humans must come first. AI that is generated from now on will become incredibly diverse depending on the utilization. All people will then use AI for their absolute desire and highest value. However, any approach for individuals to properly control or use AI has yet to be devised. In this situation, the task that has been entrusted to us today is whether we can create this powerful tool according to the promises and ideas of all human beings.

<   Previous Next   >