Extending Artificial Agency to Incorporate Ethics-oriented Social Interaction


Argyris Arnellos, Thomas Spyrou and John Darzentas


This paper suggests how information systems may make use of Systems Thinking to incorporate ethics. In Systems Thinking, it is expected that a first action towards which a group of parts is directed is to attempt to distinguish between themselves and their environments. This is a prerequisite for these parts to be called a ‘system’. Although the degree of consciousness in such primitive steps cannot be measured or evaluated, one cannot assert the same thing in cases concerning further states of a system’s evolution.

Contemporary information systems have evolved a long way from such primitives. They are developed systems, mostly technologically implemented, for purposeful communication and most often with a complex structure involving different ‘parts’, including people, data and technology. These parts, in their turn, can be seen as subsystems, which, in a continuously evolving information system, produce numerous interactions each with differing types and characteristics.

One of the most interesting subsystems is the one responsible for the application of knowledge-based decisions in an information system. Such subsystems are usually called ‘intelligent’ and the support that they provide has its sources in the domain of artificial intelligence, knowledge-based systems, fuzzy systems, neural nets and other theories and techniques, which are applied to the problem domain.

These subsystems have their own sets of observations and distinctions which are carried within the information system, in order to play a leading role in almost all of the operations of the information system. Since a primary condition for the preservation of an information system is its interaction with other systems, it is essential that the observations and distinctions underlying the system serve the ethics of the system itself as well as the ones of the systems with which it interacts.

This paper proposes that, in the realm of information systems, such ethical aspects are closely connected to the preservation of diversity, intersubjectivity and evolutionary complexity of all participating agents, either natural or artificial. In regard to the latter, so far, all ethical aspects of the artificial are indirectly fulfilled by the externally and a-priori given observation and categorization of the intelligent subsystem’s designers. However, considering an information system as a socio-technical system acting in a society consisting of webs of meaning, its artificial part ideally should have the complexity and the ability to decide and act in terms of its own ethical considerations. With this perspective, and making use at the same time of a systemic context for the analysis of the interactions of an information system, ethics can be framed in terms of the meanings and intentions which play a major role in the purposeful interaction of a system.

In this paper an analysis of the influence of systems theory in the design of artificial information systems is given, concentrating on the terms of the concept of information, as this was adopted in each stage of the theory. It is argued that in the context of an autopoietic socio-communicative framework, where information is not taken in the ordinary sense, ethics are reflected in the quality of information processing by the system. This processing is made through the creation of new meaning structures, based on the intentions and goals of the system. As complexity increases, in order for the artificial agencies in the Information System to be able to perceive the ethics-oriented characteristics of their interaction, a parallel increase in the system’s autonomy and adaptivity is expected. Given the system’s evolutionary adaptation process, this means that it should have the potential to produce and categorize new meanings, thus increasing its range and achieving the ability to incorporate ethics-oriented constraints in its decisions.

This paper presents the argument that for artificial agencies in information systems to be able to confront the ethical aspects of their interaction, they should be designed within a systemic and continuously evolving interactive framework, such as that encountered in information driven socio-communicative systems. The main purpose is the creation and communication of new meanings where ethics are related to the intentions of certain agents in such systems, as well as to the goals, scopes and constraints of the interacting environment. The role of information in such systems is examined and its ethical implications are analysed in terms of their influence on Information Systems as a whole. The shift of meaning inside the artificial system in relation to the sense of information is presented. It is shown that this is a highly likely move towards the designof systems capable of ethics-oriented social interaction.