Artificial Intelligence and Society: Philosophy of Fallibility
Column: Philosophy of Fallibility and Pragmatism

Faculty Fellow, RIETI

Regarding the “theory of innovation-driven justice,” discussed from Part 1 to Part 22, Shigeki Uno pointed out its proximity to pragmatism, an important school of modern philosophy (Uno and Kobayashi, 2019). According to Uno (2013), pragmatism is a philosophy that welcomes individuals’ freedom to conduct various experiments on the premise of human fallibility, and democracy is the institution that makes that freedom possible. It is John Dewey who presented the idea that the essence of democracy is protecting the freedom to conduct experiments under the premise of fallibility. In that sense, it may be said that the theory of innovation-driven justice is an idea that is close to Dewey's pragmatism.

The theory of the universe that was advocated by Charles Sanders Peirce, one of the founding fathers of the philosophy of pragmatism, is based on a worldview that is very similar to the theory of “strong isomorphism” (see Uno [2013] and Ito [2006]).

Peirce believes that the universe is comprised of three elements. The first element is “chance (chaos),” and the second is “law.” The universe starts from a state of chaos—a state in which chance is the dominant factor— and then it gradually develops into a state that is governed by laws. Ultimately, when everything has come to be governed by laws, the universe comes to an end. Under Peirce’s theory of the universe, it is the third element, “habit,” that crystalizes the first element, chance, into the second element, law.

According to this theory, chance ultimately comes to be governed by laws, and laws are created through the habit-formation process. The idea that the laws of the universe (laws of physics and chemistry) are created through habit formation may sound nonsensical. However, it makes sense if habit formation is understood to mean the process of learning on the part of observers of the world (humans and artificial intelligence). The habit formation as conceived under the philosophy of pragmatism is a process similar to the identification of features (fixed patterns) by AI systems, particularly through deep learning. If we interpret Peirce’s theory of the universe in that way, it may be said that the theory presents a worldview that is very similar to the theory of strong isomorphism.

The relationship between the habit formation as conceived under the philosophy of pragmatism and deep learning, which has led to the arrival of the theory of strong isomorphism, must be analyzed in detail. By linking the theory of strong isomorphism to the philosophy of pragmatism, it may become possible to conceive a new philosophy suited to the era of AI. Realizing that vision is a challenge for future research.

October 23, 2023

Article(s) by this author