I would like to thank Steve Watson and Erik Brezovec for their thoughtful response to my article Generative AI—The Transgression of Technology, and particularly for their faithful representation of my theoretical arguments and their well-informed acceptance of its central premises.
Read their response: https://onlinelibrary.wiley.com/doi/10.1002/sres.70170
My article you finde under publication on this page.
For me, conceptual development is essential if we are to understand the transformations associated with the emerging media environment. New media introduce possibilities and conditions that our existing conceptual apparatus may not yet adequately capture. Developing and refining our concepts is therefore necessary to reduce, in semantic terms, the state of latency that characterises the initial phases of media transformations, when new structures and conditions emerge before adequate semantics have developed to describe and understand them. I appreciate Watson and Brezovec’s ambition to take the discussion further by shifting attention from the technological character of generative AI to what they call an ecology of persistence. As I understand their argument, they wish to examine how technological infrastructures, organisational decisions, institutional arrangements and communicative expectations become mutually implicated in maintaining the conditions under which different systems can continue their operations. Their concepts of hyper-knotting and selection-fields are intended to capture these distributed and recursively maintained conditions. However, I find their proposed conceptual development somewhat difficult to grasp, particularly regarding what it adds to the existing systems-theoretical framework. Much of what they describe seems to me already accessible through Luhmann’s concepts of structural coupling, operational closure, expectations, decision premises and the temporal reproduction of social structures.
More fundamentally, I think we need to be careful not to confuse the environmental conditions of a system’s continued existence with the operational reproduction of the system itself. Even when systems become increasingly dependent on technological infrastructures and on one another, their operations remain system-specific and operationally closed. Systems are always already structurally adapted to their environments, and these adaptations may change as environmental conditions change. However, autopoiesis is not a matter of degree: a system either reproduces its own operations or it does not. The ecological interconnectedness of the conditions of its continued existence does not in itself explain its autopoietic reproduction. This does not mean that I reject their ambition. On the contrary, I find the question of how generative AI transforms the conditions for the continued reproduction of social systems highly relevant. My reservation concerns whether the proposed concepts provide additional explanatory power or primarily redescribe relations that can already be analysed within Luhmann’s theoretical architecture.
I remain convinced that Luhmann’s systems theory, including further conceptual developments that remain consistent with its fundamental distinctions, offers the most adequate foundation for analysing the emerging media ecology. In my view, the challenge is not to move beyond these distinctions, but to develop them sufficiently to account for the transformations we are witnessing.
I am grateful to Watson and Brezovec for engaging so seriously with my work and for opening this discussion. Their response demonstrates precisely why continued theoretical and conceptual development is necessary.
