Principles of Role Division: Amidst the Sea of Intelligence and Life
When we consider artificial intelligence, it often leads us to reflect on the nature of intelligence itself.
To unravel the nature of intelligence, it seems a shortcut to consider the differences between ordinary events occurring within inorganic physical laws and those conducted by intelligence, and what mechanisms create these differences.
Continuous Role Division
Approaching intelligence with this method, it becomes apparent that intelligence, compared to general physical phenomena, characterizes itself by the ability to autonomously think about what should be processed and in what sequence over continuous time, and to actually execute these tasks.
This can be seen as assigning tasks to different versions of oneself over time. Normally, division of labor means accomplishing tasks that cannot be done alone by sharing them with others, but a similar concept is applied by a single intelligence over continuous time.
Parallel Role Division
Furthermore, biological intelligence including humans, as well as artificial intelligence realized by computers, can generally be described as information processing by networks of numerous neurons. From this perspective, the parallel processing of these groups of neurons can also be interpreted as a division of labor.
Therefore, it is believed that sophisticated intellectual processing is achieved through both continuous division over time and parallel division among groups of neurons. This ability can also be described as the capacity to organize continuous and parallel processing units.
Learning Role Division
In practicing such division, there are methods where each processing unit decides how to divide tasks and processes them, and methods that involve role division through communication and instructions.
Both between parallel processing units and continuous ones, the judgments of each unit are acquired through learning and training processes. This applies to the learning of artificial intelligence, as well as to individual and team practices in sports. As more learning and practice are accumulated, the quicker and more accurate the task division becomes, enabling more advanced processing.
Dynamic Role Division
Moreover, in cases of high intelligence, the learning and practice not only fix the division of roles but also endow each processing unit with the ability to dynamically decide on role division based on the situation. This allows for organized processing to be effectively executed even in unprecedented situations.
Role Division Based on Memory and Communication
Additionally, communication between processing units over continuous time is carried out through memory, and among parallel processing units through communication. Moreover, the information transmitted ranges from very simple signals like zeros and ones, to symbols and words that can convey various meanings instantly, to sentences and documents that communicate more complex meanings.
Language as a Mode of Communication for Role Division
From this perspective, the hypothesis that the memory and communication abilities of neurons, and language as a mode of communication, have evolved for the purpose of role division seems plausible.
When decisions on role division among continuous and parallel processing units are made through memory and communication, even unknown situations can be managed more precisely. The signals of neurons alone can realize quite powerful intelligence. Additionally, using language, an advanced form of information, allows a single intelligence to perform more refined processes and accurately handle vast operations like those in large corporations or national projects.
The Evolution of Intelligence for Advanced Role Division
Thus, viewing intelligence from the perspective of organized role division suggests that intelligence has indeed evolved for this purpose.
The fact that neurons, appearing to be the smallest units of intelligence, have the ability to solidify role division through learning, and adjust role division with other neurons through memory and communication, explains this. As neurons form networks and their numbers increase, the simple increase in processing units that can handle role division enables more sophisticated processing, making the increase in brain cells significant in the evolution of biological intelligence. Furthermore, the evolution of communication styles used by neurons for memory and communication — ranging from signals to symbols, words, sentences, and documents — and the creation of language can also be explained from the perspective of advanced role division.
The Extensibility of the Basic Principle of Role Division
Moreover, the basic principle of learning and information transmission-based role division, which applies to a single neuron acting intelligently over time, also applies to multiple neurons in parallel, and extends to societies formed by organisms with networks of neurons, as well as to intellectual activities that continue beyond lifetimes and generations.
This means that the basic principle of role division by intelligence has multilayered extensibility, expanding and becoming more sophisticated across both time and space scales.
Multicellular Organisms from the Perspective of Fundamental Principles
Neurons are individual cells, but combined with non-neuronal cells, multicellular organisms function. From the perspective of role division, it might seem that cells other than neurons do not possess the flexible and advanced capabilities for learning and information transmission.
However, if we consider the evolutionary timescale of biology, the process by which multiple cells have evolved to adapt to their environments can be seen as a form of learning role division. Furthermore, the fact that an individual organism can somewhat manage situations that its evolutionary predecessors have not encountered suggests a mechanism for dynamic role division among multiple cells.
Thus, not limited to neurons, the cells of multicellular organisms possess the ability for dynamic organization of continuous and parallel role division. Here, the same fundamental principles that govern intelligence are at work.
Single-Celled Organisms from the Perspective of Fundamental Principles
Tracing back through the evolutionary history of life, before the advent of multicellular organisms, only single-celled organisms existed. A single cell maintains life functions alone, equipped with the capacity to adapt to the environment and to some extent, unknown situations.
This capacity stems from the presence of diverse intracellular organelles within a single cell, which perform organized role division both continuously and in parallel. Therefore, even single-celled organisms are based on the same fundamental principles as intelligence, undergoing a form of environmental learning and adaptation.
Cells are not merely advanced intracellular organizations, but can also be viewed as complex systems centered around DNA, RNA, and proteins.
DNA can store information learned by an organism about its environment, which is then transmitted to intracellular structures via RNA to synthesize proteins. Proteins are agents that perform a variety of processes necessary for maintaining life. Within this mechanism, highly organized processes necessary for sustaining life occur, capable of learning and adaptation, involving both continuous and parallel role division.
The Origin of Life from the Perspective of Fundamental Principles
As a personal research theme, I am exploring the origin of life from the perspective of systemic mechanisms.
Before the appearance of single-celled organisms, a prototypical mechanism operated within cells, which evolved into single-celled organisms. From this viewpoint, even simpler mechanisms might contain the same fundamental principles of role division, leading to the evolution of life.
Considering dynamic organization of continuous and parallel role division, it’s natural for chemical reactions, even among inanimate substances, to occur continuously and in parallel. The challenge lies in learning this continuous and parallel role division and having the dynamic flexibility to adapt to unknown situations.
If the chemical substances on Earth could fulfill these characteristics and continuously expand and sophisticate, the emergence of life from non-life could be explained using the same fundamental principles and framework used to describe the sophistication of intelligence.
In the Sea of Chemicals and Neuronal Cells
The chain-like structures of nucleotides in DNA and RNA, and the chain-like structures of amino acids in proteins, interestingly parallel the symbolic structures of language in intelligence.
The hypothesis that language has significantly contributed to the evolution of intelligence’s role division capabilities might be analogously supported by the evolution of DNA, RNA, and proteins as central systems in life.
Furthermore, if the emergence of life from non-life in the sea of chemicals can be explained using the fundamental principle of advanced role division, then the emergence of consciousness from the unconscious within the sea of intelligence woven by neuronal cells might also be explainable by the same principle.
In Conclusion
In this article, we’ve focused on the coordination of processes over time from the perspective of differences between general physical reactions and intelligence, leading us to the principle of role division. This principle of role division can be applied not only temporally but also concurrently, and we’ve managed to organize learning and information transmission as key elements.
From the perspective of these fundamental principles, it becomes clear that the same principles apply whether it’s a single neuron, a neural network, an individual, or society, enabling the realization of active processing. Furthermore, tracing back through multicellular and single-celled organisms to the prebiotic world of chemicals, we see that advanced processing driven by this principle has evolved through the scaling of parallelism, and the sophistication of learning and information transmission.
Additionally, this leads us to hypothesize the evolution of mechanisms like language, DNA, and RNA that have enhanced role division, and speculate whether the origins of consciousness and life itself could be explained by these common principles.
By assuming this fundamental principle, we can incrementally understand how this principle is refined from the simplest models. Even a single neuron in continuous time can divide roles to an extent; expanding to two neurons, we explore what becomes possible, and how many neurons are necessary as memory and communication evolve from signals to symbols, and further to words and sentences. Throughout this evolution, we can see what structures and processes appear within neurons and how role division is sophisticated.
Even this preliminary contemplation provides many insights from this perspective. Being able to think about complex phenomena like life and intelligence from a small scale incrementally not only proves that they could evolve sequentially but also means that our understanding can progress step by step. Moreover, applying this approach to both the origins of life and intelligence may yield additional insights and clues.
Furthermore, the keyword “sea” used in the title and final sections of this article draws an analogy to the origins of life, thought to have occurred in Earth’s waters, and to cells maintaining life through chemical reactions in the liquid cytoplasm.
And the application to intelligence reflects my belief that the structure of chemical reactions occurring in liquids, where chemicals meet and chain-react, mirrors the processing structure of artificial intelligence systems like ChatGPT, which incorporate outputs back into inputs for repetitive processing. Thus, I anticipate that the mechanism by which role division becomes operational not only shares the common principles of learning and communication evolution but also has commonalities in the processing structure that occur in a “sea”.
