The term agent is used today to mean anything between a mere subroutine to a conscious entity. There are "helper" agents for web retrieval and computer maintenance, robotic agents to venture into inhospitable environments, agents in an economy, etc. Intuitively, for an object to be referred to as an agent it must possess some degree of autonomy, that is, it must be in some sense distinguishable from its environment by some kind of spatial, temporal, or functional boundary. It must possess some kind of identity to be identifiable in its environment. To make the definition of agent useful, we often further require that agents must have some autonomy of action, that they can engage in tasks in an environment, independently or without external control. This is, in effect, a definition of agency directly related to the one put forward in the XIII century by Thomas Aquinas: an entity capable of election, or choice.
This is a very important definition indeed; for an entity to be referred to as an agent, it must be able to step out of the dynamics of an environment, and make a decision about what action to take next---a decision that may even go against the natural course of its environment. By this, simplistically, I mean that an agent (say in a steep incline) can opt to go uphill rather than roll with the force of gravity. Since choice is a term loaded with many connotations from theology, philosophy, cognitive science, and so forth, I prefer to discuss instead the ability of some agents to step out of the dynamics of its interaction with an environment and explore different behavior alternatives. In physics we refer to such a process as dynamical incoherence . In computer science, Von Neumann, based on the work of Turing on universal computing devices, referred to these systems as memory-based systems. That is, systems capable of engaging with their environments beyond concurrent state-determined interaction by using memory to store descriptions and representations of their environments. Such agents are dynamically incoherent in the sense that their next state or action is not solely dependent on the previous state, but also on some (random-access) stable memory that keeps the same value until it is accessed and does not change with the dynamics of the environment-agent interaction. In contrast, state-determined systems are dynamically coherent (or coupled) to their environments because they function by reaction to present input and state using some iterative mapping in a state space.
Let us then refer to the view of agency as a dynamically incoherent system-environment engagement or coupling as the strong sense of agency, and to the view of agency as some degree of identity and autonomy in dynamically coherent system-environment coupling as the weak sense of agency. The strong sense of agency is more precise because of its explicit requirement for memory and ability to effectively explore and select alternatives. Indeed, the weak sense of agency is much more subjective because the definition of autonomy, a boundary, or identity (in a loop) are largely arbitrary in dynamically coherent couplings.
I have been working on various types of agent models that are either based on the strong sense of agency detailed above, or attempt to study the emergence of such agency from dynamically coherent environments. The strong sense of agency has been further detailed in an overview of agent models for a previous project of the modeling of socio-technical systems as well as in an overview of research on complex systems modeling I wrote. Examples of agent-based models I have worked on are: the simulations of evolving agents with different kinds of reproduction strategies using Fuzzy Development Programs, the agent-based model of genotype editing, the evolving cellular automata experiments, the soft computing agents for recommendation systems, the immune-inspired spam detection algorithm, etc.
We have developed a bio-inspired solution for binary classification of textual documents inspired by T-cell cross-regulation in the vertebrate adaptive immune system, which is a complex adaptive system of millions of cells interacting to distinguish between self and nonself substances. In analogy, automatic document classification assumes that the interaction and co-occurrence of thousands of words in text can be used to identify conceptually-related classes of documents—at a minimum, two classes with relevant and irrelevant documents for a given concept (e.g. articles with protein-protein interaction information). Our agent-based method for document classification expands the analytical model of Carneiro et al, by allowing us to deal simultaneously with many distinct populations of antigen-specific T-Cells and their collective dynamics. We have extended this model to produce a spam-detection system. We have also developed our agent-based model further to apply it to biomedical article classification, testing it on a dataset of biomedical articles provided by the BioCreative 2.5 challenge. Our results are useful for biomedical text mining, but they also help us understand T-cell cross-regulation as a potential general principle of classification available to collectives of molecules without a central controller. While there is still much to know about the specifics of T-cell cross-regulation in adaptive immunity, Artificial Life allows us to explore alternative emergent classification principles while producing useful bio-inspired tools. Recently, we started expanding this algorithm to other forms of classification such as sensor data from human-robot interactions under an IUCRG project.
Evolutionary models in theoretical biology at large, and computational biology and artificial life in particular, rarely deal with ontogenetic, non-inherited alteration of genetic information because they are based on a direct genotype-phenotype mapping. In contrast, in Nature several processes have been discovered which alter genetic information encoded in DNA before it is translated into amino-acid chains. Ontogenetically altered genetic information is not inherited but extensively used in regulation and development of phenotypes, giving organisms the ability to, in a sense, re-program their genotypes according to environmental clues. An example of post-transcriptional alteration of gene-encoding sequences is the process of RNA Editing. Our latest agent-based model of genotype editing presents a novel architecture for evolving agents in which coding and non-coding genetic components are allowed to coevolve. Our goal is twofold: (1) to study the role of RNA Editing regulation in the evolutionary process, and (2) to investigate the conditions under which genotype edition improves the optimization performance of evolutionary algorithms. We have shown that genotype edition allows evolving agents to perform better in several classes of fitness functions, both in static and dynamic environments. We are also investigating the ways in which the indirect genotype/phenotype mapping resulting from genotype editing lead to a better exploration/exploitation compromise in the search process. In the past year we developed an entirely new modeling platform in Python to run experiments to explore the evolutionary advantages of RNA editing.
This research is described in greater detail in the separate Evolutionary Models of Genotype Editing page.
Ian B Wood