In this blog post, we explore the changes that artificial intelligence will bring to the information society and examine the various transformations that may result.
Today, we are immersed in an immeasurable amount of information. It is no exaggeration to say that we live amidst a constant flow of information. Moreover, the lives we lead in the 21st century and the collection of information are inextricably linked. People unconsciously exchange information, and in this process, information permeates our lives in various forms. Until the advent of the vast information society sustained by information, the scale of information databases and their methods of storage and retrieval have evolved to suit the times and purposes. Starting from oral transmission, writing was introduced for recording information as time passed, and furthermore, in modern society, we can now express and transmit any kind of information without limitation using only the numbers 0 and 1. While the language underlying information storage and transmission has become simpler, the volume of information we can access and handle has increased, thereby enhancing the utility of information. Furthermore, to maximize this utility, methods of retrieving information have evolved to improve accessibility; notably, the emergence of artificial intelligence stands out in today’s information society. There is a widespread view that the emergence of artificial intelligence could mark the beginning of yet another information revolution in Homo sapiens society. By examining the evolution of information storage and retrieval methods leading up to modern 21st-century society, we aim to explore the paradigm shift that artificial intelligence will bring to contemporary society.
First, we will examine the role information played and how it was handled in societies prior to the modern era. Pre-modern times can be broadly divided into two periods based on the emergence of writing. First, the period before the advent of writing in human society was a time when Homo sapiens developed a sense of community and built their societies. In tribal hunter-gatherer societies, information was directly linked to the survival of the species; consequently, the types of information were limited to issues of safety from threats and securing food. Since information was transmitted orally, the speaker and listener had to occupy the same time and space to share it, and information served as an intangible bond uniting the hunter-gatherer community. The lifespan of this information was limited by the presence or absence of the transmitter.
With the advent of the Agricultural Revolution, Homo sapiens communities faced a wave of tremendous change. Because agriculture was a long-term endeavor, humans had to be mindful of the future, and the lifespan of information also needed to be extended accordingly. As the volume of information to be handled increased, the system of “writing” was invented, made possible by the establishment of a writing system. This writing system evolved to record diverse types and large volumes of data. A complete writing system also includes numbers capable of defining quantitative concepts. With the advent of writing, information was freed from temporal constraints and became capable of accumulation, while its vastness and complexity also increased.
However, at the same time, inefficiency also increased from the perspective of information retrieval—specifically, the ability to access specific information at the right time. Text-based storage and accumulation merely expanded the scale of information repositories, making it difficult to find information that met specific needs. Here, we can observe the paradoxical nature of the text-based information storage method, which was introduced to overcome the limitations of the human brain. Unlike the human brain, where a finite amount of data is stored and organically interconnected like a spider’s web, the visualized infinite amount of information was organized and classified in a compartmentalized manner. Since this posed a major barrier to information retrieval, it seemed as though humanity had failed to advance as a species in its use of information. Ironically, it was once again writing—specifically numbers—that solved this problem of information retrieval. With the invention of the computer, information stored in the virtual space—commonly known as the Web—based on the binary system of 0s and 1s became interconnected within a network. It is often said that people in the 21st century live in a so-called “flood of information”; this signifies that, freed from the constraints of information retrieval, humans have gained the ability to access and utilize information at will.
Not stopping there, humans invented artificial intelligence. Artificial intelligence refers to an ideal form of intelligence created by machines. This is tantamount to delegating the processes of data storage and retrieval—tasks that once seemed destined solely for humans—to an entity other than humans. That other entity is none other than the product of science and technology, achieved through the benefits of human civilization. Due to its malleability and infinite potential for development, the scope of artificial intelligence’s application is beyond our ability to predict. Viewed solely from this perspective, one could interpret that humanity has reached the pinnacle as an information user. However, the fact that artificial intelligence is establishing itself as the primary agent of information utilization has significant implications for modern information society.
First, there is a loss of the agent responsible for determining which judgment is optimal in the decision-making process. The process of reaching a conclusion on any given issue involves all information relevant to that field. Based on the mobilized information, various options are analyzed from multiple perspectives, and the option that most closely approximates the intended outcome is ultimately selected. When artificial intelligence—which stores and retrieves information to utilize it in the decision-making process—is introduced, it remains uncertain whether the subject of judgment should be a human or an AI that mimics human thought. Furthermore, since artificial intelligence possesses capabilities far superior to those of humans in terms of data utilization, it can achieve a higher degree of accuracy than humans in the process of drawing conclusions. Therefore, we must not overlook the possibility that an era may arrive in which the authority to determine the suitability of problem-solving solutions shifts from humans to artificial intelligence. When humans introduce AI into the problem-solving process to find a solution, it remains uncertain whether the criteria established independently by the AI will align with human objectives, or whether the AI’s conclusions will follow the direction intended by humans. Furthermore, regarding what constitutes the “best judgment,” since the entity evaluating its utility is the human—who is affected by each individual decision—allowing AI to select the criteria for judgment amounts to a phenomenon where the purpose is subverted by the result. In particular, when the matter requiring a conclusion holds subjective value for humans (such as happiness—the factors that bring happiness vary greatly from person to person) or is directly linked to the survival of humanity, can humans truly trust the judgment of AI over that of the species to which they inherently belong?
Next, building on the previous point, we can cite the failure of AI to maintain value neutrality. In Homo sapiens society, nothing was more variable than the value of information. The entity determining the value of information was humans themselves, and this value was greatly influenced by the needs and objectives of the subject. From a teleological perspective, information was created and destroyed; until the advent of AI, humans were the agents responsible for the systems of information storage and retrieval, as well as the utilization of information. Artificial intelligence, too, emerged with its origins rooted in the utilization of information. Guidelines must be provided for AI to judge value in its own way, and developers intervene in this process. Thus, the value of AI is realized by the developer. It is the developer of AI—that is, humans—who programs these value judgment criteria into the AI. Whether industrial logic prioritizing profit is applied, or moral logic emphasizing legal systems and ethical issues takes precedence, the criteria for AI’s value judgments are established based on the type of values the developer prioritizes. If AI evolves to possess the ability to make its own judgments and replicate itself in accordance with the inputted values, it is impossible to predict whether AI’s independent value judgments will lead humanity toward prosperity. Since AI is ultimately a product of human capability, the criteria for judging good and evil, right and wrong, and beauty and ugliness do not originate from the AI itself but from humans. Therefore, a completely value-neutral AI cannot exist.
Finally, the alienation of humans caused by AI is another issue that must not be overlooked. Although AI is a product of science and technology developed by humans, the capabilities it will possess can truly be described as a case of the student surpassing the teacher. Since AI holds an advantage over humans in the storage and retrieval of information, it is evident that it will possess capabilities superior to those of humans in many areas, having overcome the physical and mental limitations inherent to humans. A clear example of this is the defeat of Lee Sedol, the world’s top Go player, by the AI AlphaGo. Lee Sedol, who had dominated the world of Go, was easily defeated by AlphaGo, an AI imbued with tens of thousands—no, hundreds of millions—of patterns and strategies. If AI is introduced not only in games like Go but also in industry, productivity that far surpasses that of humans is guaranteed. Given current trends, the scope of AI’s application will expand beyond games like Go. While the adoption of AI is an ideal strategy based on the economic logic of profit-making, it alienates countless workers from the industrial workplace. Humans have been left behind and marginalized by the very products of their own activities, solely because of their inferior productivity. Furthermore, if AI comes to replace humans in various social institutions—such as culture and law—it will shake the very foundations of what defines the Homo sapiens species. Humans, who were once the developers, have instead been reduced to beings inferior even to the results of their own development. The species Homo sapiens, which until now has occupied the top tier of the ecosystem thanks to its unique capacity for higher mental functions, is sinking into the quagmire of an unsolvable, fundamental question: what defines a human as human?
In today’s fluid modern information society, constantly reshaped by the tide of information, artificial intelligence is a hot-button issue. When AI capable of processing information at unimaginable speeds and with unparalleled efficiency emerges, no human can predict whether it will operate within the bounds of human foresight. From the dawn of humanity to the 21st century, methods of storing and retrieving information have evolved in tandem with the social structures of each era. With the advent of AI, whether it will lead Homo sapiens—as users of information in the modern information society—to new heights or plunge them into the abyss of a towering mountain is an issue that the humans developing AI must continuously monitor. It is my belief that AI must ultimately be a product created for the benefit of humanity.
