The notion of feedback represents one of the most important contribution of the cybernetics to the science of (any) complex systems design. In the cybernetic tradition, machines capable of adapting themselves actively to the environments via trial-and-error process based on negative feedback (used for autocorrection) were called “servomechanisms” or “negative feedback automata”. A recent example of machines using these type of mechanisms comes from the Artificial Intelligence, since current artificial neural networks (at least the once that are more successful nowadays) make use, in their supervised learning phase, of a well knoan feedback mechanism called "backpropagation".
The book "Feedback, how to destroy or save the worlds" (Springer, 2024) by Péter Érdi is an intellectual journey in the science of feedback and shows, with clear examples coming from different fields, how either the design or the discovery of feedback mechanisms is crucial for the advancement of science, technology, economics and society since feedback is a crucial component of any complex system.
The book focuses not only on the well known importance of negative feedback (e.g. for learning) but also, and more importantly, on the importance of understanding what are feedback mechanisms governing the behavior of a certain system and upon which it is possibile to intervene (using both positive and negative corrective signals) in order to lead to (biological/computational/mechanical/physical/chemical...) states that can lead to homeostatsis, market stability, environmental sustainability and so on.
If there is any chance of saving this World, the realization of powerful feedback mechanisms able to avoid catastrophic outcomes represents one of the few elements to put in place. And, in this state of affairs, the science of feedback (Cybernetics) strikes back and seems to be even more relevant today!