Cybernetic Thinking
The human body, with its internal functions and organs and systems of organs, has long served as a strong analogy to better grasp and explain concepts and mechanisms in natural sciences and technology but also in social and political spheres.
A virus is both something that science through the microscope lens has been able to research and at the same time, a way to describe malicious code in software. In a social context, a virus and its spreading serve to describe, e.g., the dissemination of political ideas through the social body.
It’s this analogy that cybernetics has exemplified via the concepts of feedback and control right from its inception as a theoretical construction. The strength that analogy brings stems, obviously, from the fact that we still haven’t lost our bodies as humans, and this still remains our main vehicle to interact with the environment. We understand by sensing and constructing ideas to explain and interact with the world. Things get a little more interesting, though, as this continues recursively—our sensing and understanding continuously evolve.
There are two concepts in cybernetic thinking—as I understand them—that are more than relevant today. The one is the performative nature of intelligence—Andrew Pickering describes vividly “performative brain” as one of the major traits of cybernetic thinking in his book “The Cybernetic Brain”. Contemporary AI systems and agentic AI can be better explained and examined via this cybernetic lens. Emergent intelligence from probabilistic systems, complexity from simple processes and improvement through feedback loops and reinforcement learning iterations.
The second has to do with the emphasis on nondeterministic processes. In the deterministic world, we used to describe all the steps towards a predefined goal. But maybe what holds greater value today is a set of problems and an attack method that does not need a strictly predefined goal with a dynamic process towards that goal. Cybernetics seems to be more adaptive to these kinds of problem sets and technologies.
Relevant reads: