Cybernetics

AI readiness

Cybernetics could also be seen as an evolutionary theory for machines. There is no need for machines to represent the world in order to achieve an output or result; instead, the focus is on the adaptive nature of machines and systems. One has to set the rules and relationships between systems, and according to those rules, systems evolve in their own way, which of course, relates back to the initial point.

As Kuhn remarked, the breakthrough of evolutionary theory was the revelation that there is no divine plan—there is no inherent goal in the world. This is also something that emerges from cybernetics theory, albeit in a different way: systems tend to find an equilibrium and evolve in non-predetermined and often unpredictable ways.

Additionally, the introduction of a new part or agent alters the stability of the system as a whole. And a key element of cybernetics is the human-machine interplay, with each giving feedback and shaping the other and the system as a whole.

Based on this, cybernetics seems to be a coherent theory for AI. We are dealing with non-deterministic systems and algorithms, and their introduction brings a new (even if not always beneficial) and stronger relationship between humans and machines.

#journal

Read Next

Through the window

I used to view my apartment as a boundary, a place that defines the inside—a small, or not-so-small, internal space that separates me from the outside world. It was the place I loved to hate during the COVID period, yet also a space we reshaped in many ways to support a different way of living.

Are we responsible for our choices?

On Project Syndicate, Peter Singer discusses choice and responsibility, exploring the extent to which we can be held accountable for our choices and how truly ours those choices are. It's a great read.