Feedback loops are a key feature of the systems the article is dedicated to, such as ecosystems and individual organisms. They also exist in the human world, communities, organizations and families.
Artificial systems of this kind include robots with control systems that use negative feedback to maintain desired states.
Key features
In an adaptive system, the parameter changes slowly and does not have a preferred value. However, in a self-regulating system, the value of the parameter depends on the history of the dynamics of the system. One of the most important qualities of self-regulating systems is adaptation to the edge of chaos or the ability to avoid chaos. Practically speaking, heading towards the edge of chaos, without moving further, the observer can act spontaneously, but without catastrophes. Physicists have proven that adaptation to the edge of chaos occurs in almost all feedback systems. Let the reader not be surprised at the elaborate terminology, because such theories directly affect the theory of chaos.
Pratopoiesis
Practical poetry as a term coined by Danko Nikolic is a reference to a kind of adaptive or self-regulating system in which autopoiesis of an organism or cell occurs through allopoietic interactions between its components. They are organized into a poetic hierarchy: one component creates another. The theory suggests that living systems demonstrate a hierarchy of four such poetic operations:
evolution (i) → gene expression (ii) → non-gene related homeostatic mechanisms (anapoiesis) (iii) → cell function (iv).
Practical poetry challenges the modern doctrine of neurobiology, arguing that mental operations mainly occur at the anapoetic level (iii), that is, that minds arise from fast homeostatic (adaptive) mechanisms. This contrasts with the widespread belief that thinking is synonymous with neural activity (cell function at level iv).
Each lower level contains knowledge that is more general than the higher level. For example, genes contain more general knowledge than anapoetic mechanisms, which, in turn, contain more general knowledge than cell functions. This hierarchy of knowledge allows the anapoetic level to directly store the concepts necessary for the emergence of the mind.
A complex system
A complex adaptive system is a complex mechanism in which a perfect understanding of the individual parts does not automatically provide a perfect understanding of the whole structure. The study of these mechanisms, which are a kind of subset of nonlinear dynamical systems, is highly interdisciplinary and combines knowledge of the natural and social sciences to develop models and representations of the highest level that take into account heterogeneous factors, phase transition, and other nuances.
They are complex in that they are dynamic networks of interactions, and their relationships are not sets of separate static objects, that is, the behavior of the ensemble is not predicted by the behavior of the components. They are adaptive in that individual and collective behavior mutate and self-organize in accordance with the initiating change of a micro-event or set of events. They are a complex macroscopic collection of relatively similar and partially connected microstructures formed to adapt to a changing environment and increase their survival as a macrostructure.
Application
The term “complex adaptive systems” (CAS) or the science of complexity is often used to describe the poorly organized academic field that has grown around the study of such systems. The science of complexity is not a single theory - it covers more than one theoretical basis and is highly interdisciplinary, looking for answers to some fundamental questions about living, adaptable, variable systems. CAS research focuses on the complex, emergent, and macroscopic properties of a system. John H. Holland said that CASs are systems that have a large number of components, often called agents, that interact, adapt, or train.
Examples
Typical examples of adaptive systems include:
- climate;
- cities;
- firms;
- Markets
- Governments
- industry;
- ecosystems;
- social networks;
- Electricity of the net;
- flocks of animals;
- traffic flows;
- colonies of social insects (e.g. ants);
- brain and immune system;
- cells and a developing embryo.
But that is not all. Also in this list you can include adaptive systems in cybernetics, which are gaining more and more popularity. Organizations based on social groups of people such as political parties, communities, geopolitical communities, wars, and terrorist networks are also considered CAS. The Internet and cyberspace, consisting, collaborating and managed by a complex set of human-computer interactions, are also seen as a complex adaptive system. CAS may be hierarchical, but at the same time, it will always more often display aspects of self-organization. Thus, some modern technologies (for example, neural networks) can be called self-learning and self-tuning information systems.
Differences
What distinguishes CAS from a pure multi-agent system (MAS) is its attention to the properties and functions of a higher level such as self-similarity, complexity of structure, and self-organization. MAS is defined as a system consisting of several interacting agents, while in CAS the agents and system are adaptive, and the system itself is self-similar.
CAS is a complex collection of interacting adaptive agents. Such systems are characterized by a high degree of adaptation, which gives them extraordinary stability in the face of changes, crises and disasters. This should be considered when developing an adaptive system.
Other important properties are: adaptation (or homeostasis), communication, collaboration, specialization, spatial and temporal organization and reproduction. They can be found at all levels: cells specialize, adapt and multiply, as larger organisms do. Communication and collaboration takes place at all levels, from agent to system level. The forces driving cooperation between agents in such a system can in some cases be analyzed using game theory.
Modeling
CAS are adaptive systems. Sometimes they are modeled using agent and complex network models. Agent-based ones are developed using various methods and tools, primarily by first identifying the various agents within the model. Another method for developing models for CAS involves developing complex network models by using the interaction data of various CAS components, such as an adaptive communications system.
In 2013, SpringerOpen / BioMed Central launched an open-access online magazine on the topic of Complex System Modeling (CASM).
Living organisms are complex adaptive systems. Although the difficulty of quantifying biology is difficult, evolution has given rise to several amazing organisms. This observation has led the widespread misconception about evolution to be progressive.
Desire for complexity
If the above was generally true, evolution would have an active tendency toward complexity. In this type of process, the value of the most common degree of complexity will increase over time. Indeed, some artificial life simulations suggest that CAS generation is an inevitable feature of evolution.
Nevertheless, the idea of a general tendency toward complexity in evolution can also be explained through a passive process. This includes an increase in variance, but the most common value, mode, does not change. Thus, the maximum level of complexity increases over time, but only as an indirect product of the total number of organisms. This type of random process is also called limited random walk.
In this hypothesis, the obvious tendency to complicate the structure of organisms is an illusion. It arises from the concentration on a small number of large, very complex organisms that inhabit the right tail of the distribution of complexity, and the neglect of simpler and much more common organisms. This passive model emphasizes that the vast majority of species are microscopic prokaryotes, which make up about half of the world's biomass and the vast majority of the Earth’s biodiversity. Therefore, simple life remains dominant on Earth, and complex life seems more diverse just because of sample bias.
If biology does not have a general tendency toward complication, this will not prevent the existence of forces that lead systems to complexity in a subset of cases. These minor trends will be balanced by other evolutionary pressures that lead systems to less complex states.
The immune system
An adaptive immune system (also known as acquired or, less commonly, specific) is a subsystem of the general immune system. It consists of highly specialized cells and processes that eliminate pathogens or prevent their growth. The acquired immune system is one of the two main immunity strategies in vertebrates (the other is the innate immune system). Acquired immunity creates an immunological memory after the initial response to a specific pathogen and leads to an enhanced response to subsequent meetings with it. This process of acquired immunity is the basis of vaccination. Like the innate system, the acquired system includes not only components of humoral immunity, but also components of cellular immunity.
Term history
The term “adaptive” was first used by Robert Hood in relation to antibody responses in frogs as a synonym for an acquired immune response in 1964. Goode acknowledged that he used these terms as synonyms, but explained only that he preferred to use the term. Perhaps he was thinking about an implausible theory of antibody formation in which they were plastic and could adapt to the molecular form of antigens, or about the concept of adaptive enzymes, the expression of which could be caused by their substrates. The phrase was used almost exclusively by Goode and his students, as well as several other immunologists working with marginal organisms until the 1990s. Then it began to be widely used in combination with the term "innate immunity", which became a popular subject after the discovery of the Toll receptor system. in Drosophila, a previously marginal organism for the study of immunology. The term “adaptive”, used in immunology, is problematic because acquired immune responses can be either adaptive or maladaptive in physiological terms. Indeed, both acquired and immune responses can be adaptive and non-adaptive in an evolutionary sense. Most textbooks today use the term “adaptive” exclusively, noting that it is synonymous with the word “acquired”.
Biological adaptation
Since the discovery, the classical meaning of acquired immunity has come to mean antigen-specific immunity mediated by rearrangements of somatic genes that create antigen receptors that define clones. In the last decade, the term “adaptive” has been increasingly applied to another class of immune response, which until now has not been associated with rearrangements of somatic genes. These include the expansion of natural killer (NK) cells with yet unexplained specificity for antigens, the expansion of NK cells expressing receptors encoded by the germ line, and the activation of other innate immune cells in an activated state that provides short-term immune memory. In this sense, adaptive immunity is closer to the concept of “activated state” or “heterostasis”, thus returning to the physiological meaning of “adaptation” to environmental changes. Simply put, today it is almost a synonym for biological adaptation.