Common Principles of Life, Intelligence, and Sociality: The Network of Identity
It seems that systems such as biology, intelligence, and society, which autonomously change while forming order, share common properties as a system.
In thinking about the commonalities of these systems, I initially focused on the nature of feedback loops. This is because feedback loops can retain changes with higher adaptability and eliminate less adaptable changes, thereby forming order amidst complex changes.
In addition to this, the property of identity has also become important to these systems. Identity is the property of distinguishing self from others.
Initially, I understood identity as a static structure that defines the boundary between the inside and the outside of a system. My thoughts progressed slightly from there, and I began to think that identity also includes dynamic and passive activities, such as whether something coming from the outside or produced inside is heterogeneous or homogeneous. And by delving deeper into this idea, I came to think that active activities, such as self-replication or reproduction, are also characteristics for the sake of identity.
In this article, I model the common properties of autonomously order-forming systems as a network of identities with feedback loops. In this process, I delve into the properties of identity and introduce the perspective of modeling it in terms of skewness and dispersion of distribution. Then, I examine various systems that form order autonomously, noting the presence of identity and feedback loops and their network structure.
This reveals that systems such as life, intelligence, and society form order autonomously based on common principles.
Methods to Maintain Identity
There are two methods to maintain identity: actively reproducing oneself and passively excluding heterogeneous elements.
In biology, self-replication and self-maintenance through genes are examples of internal self-reproduction, while the immune system corresponds to the exclusion of heterogeneous elements. In many cases, identity is maintained by combining these two methods.
Delving into Identity
Active self-reproduction is, in a sense, self-design. Organizational activities support this self-design through policies or guidelines; in communities, through education or socialization; and for individuals, through beliefs or successful experiences.
Passive exclusion is, in a way, security. The identification through whitelists or blacklists and the distinction between abnormal and normal allow security to function.
Skewness and Dispersion of Distribution
If gene sequence patterns were simply randomly combined, various patterns would be evenly distributed without any bias.
However, because genes that are suitable for the environment are more likely to remain, the pattern of sequences becomes biased towards certain types.
The beliefs of individuals, the culture of communities, and the policies of organizations also have various patterns, but those that fit better continue to remain and are regenerated, resulting in a specific bias.
Additionally, passively excluding heterogeneous elements through immunity or security narrows the dispersion of distribution and sharpens the bias.
In other words, identity is the skewness and dispersion of distribution.
Feedback Loops
If the bias and variance in distributions are inherent to identity and evolve and intensify over time, then there exists a feedback loop.
Whether it’s genetics, personality, culture, or policy, the bias and variance in identity affect the external world, and feedback from the external world, in turn, influences the bias and variance.
Through the repetition of this loop, the bias and variance are refined, and the distribution settles into a certain shape.
Networks
Surrounding a certain identity, there can be numerous other identities.
In this case, multiple identities are included within the feedback loop, and a single identity may belong to multiple feedback loops.
This means that numerous identities form a network.
Each identity has its unique bias and variance.
And through each feedback loop, the bias and variance of the identities contained within are refined in parallel.
Networks of Identities with Feedback Loops
Ecosystems form networks where various life forms, with genetic information as their identity, interact and create feedback loops.
Over time, the genetic information and immunity that constitute the identity of each species evolve and refine.
Similarly, multiple individuals form beliefs within societal networks, influencing and being influenced. Not only individuals but also organizations and communities refine and develop their cultures and policies within these societal networks.
Apart from ecosystems and societies, networks of identities with feedback loops can be observed in various layers, such as neural networks, chemical evolution, and industries and economies.
Neural Networks
The brain is a network of nerve cells called neurons, which is modeled as neural networks and applied in core artificial intelligence technologies.
Neurons convert multiple input signals into output signals.
This output, with its bias and variance, changes and refines in response to feedback, allowing both the brain and neural network models to learn.
This is a network of identities with feedback loops.
Chemical Evolution
Tracing back to the origin of life, there was a process where simple chemicals naturally synthesized into complex chemicals like DNA, RNA, and proteins before the birth of biological organisms.
This process is called chemical evolution.
DNA and RNA consist of four types of simple chemicals called nucleotides in a long chain, while proteins consist of twenty types of amino acids in a chain.
Each nucleotide or amino acid can connect in any sequence, potentially making the distribution of sequence patterns uniform.
However, feedback loops through protein or RNA catalytic actions cause biases and variances in the distribution, as DNA and RNA replicate themselves and proteins are reproduced from genetic information.
In water, various sequence patterns of DNA, RNA, and proteins could have formed a network influencing each other.
Thus, the process of chemical evolution is presumed to have networks of identities with feedback loops.
Even if the catalytic actions were minimal and the replication and reproduction were partial and imprecise in the early stages, over time, various identities refined and evolved.
Industries and Economies
When manufacturing products in factories, production is based on blueprints.
To improve quality, it’s important to minimize variations in products by controlling raw material variability and adjusting temperature and humidity, as well as stabilizing the work of craftsmen and machines.
If product variability cannot be controlled, products are checked post-production, and those exceeding standard ranges are discarded.
This shows a mechanism where bias and variance are adjusted through reproduction and exclusion.
Furthermore, the sales of high-quality products in the market provide feedback, enhancing this loop. Raw materials and parts for the products are also produced in other factories, where bias and variance are adjusted.
This implies that industries and economies can also be seen as networks of identities with feedback loops.
Conclusion
The model of networks of identities with feedback loops explains the commonalities among biology, intelligence, society, and more detailed layers such as the chemical evolution system at the origin of life, neural networks, and industries and economies.
Starting from the macro layers of complex and analytically challenging systems of biology, intelligence, and society, and applying the same model to these detailed micro layers, is considered a significant discovery.
This highlights the self-organizing systems, including chemical substances, neurons, and economic entities, forming a coherent order autonomously.
