What is the broadest sense of a structural coupling agent and environment?
The fact that an agent is not an alien to where it should live: is probably the correct framing.
Like the OOP approach in which the context of the environment is the template, the class, and so the instantiation of the class represents an agent. The coupling in this sense consists then in the shared template. Each instantiation working to fix initial problems and further develop on that fix. A consensus, an updated template, a new generation. Then new problems to fix, but then beginning with basic attractors framework.
The broadest sense of a structural coupling.
by Casian STEFAN, Principal Researcher at Essentia Mundi AI Lab. Contact: ai-AT-essentiamundi.com / ai.essentiamundi.com
Mar. 2024.
Please consider citation with link, if you derive work. Or contact me for collaboration. Thank you!
The agent should inherit the environment's properties. I can think of a structure much like the OOP paradigm. A class is a template, the environment. An instance of it, should inherit template, an agent. There is also a characteristic: the agent should be able to work-on only with the properties of the template. In a sense we can't see the difference between the template and the agent, as they can only be of the specific kind, it is only the bringing up of an instance, and by acting, that would make the difference between the two.
The only guiding act of the egent and environment, would be to "fixing" things. This also means, the setting of goals, but in no way by an external master.
Problems arise when the system itself is about to break its equilibrium state. Assuming that an equilibrium state means for example: 1+1=2, and that the system does not need to fix that. To calculate 123+123 should be seen as a disorder in the system, that needs fixing. The action thus being in calculating that, eventually.
Who says that this is the correct answer? There is a consensus among the agents. Eg. agree on a number and probe this is a good result, it is conform to some probing. If not, adjust, iterate till it gets universally, within the system, accepted.
The system incorporates that in its template, at the next wave, the agents would then have a broader base. And so on, with an iterative approach, the system should propose itself new fixes through the introduction of disorder. Or attractors in this form, of things to be solved.
On the same line of thought, the structural coupling at the computational level is to be ascribed to mortal computing. In this paradigm, the hardware and software are very accustomed to each other so that the software part can't be copied, and when the hardware dies, the software is gone too.
In this sense the fusion of the two is of codependent nature, the software, the agent is dependent of the environment and the environment would not elevate its existence without the agent.
And on top of that, mitigating the Turing complete paradigm limitation, we may have the kind of analog computing with neural networks directly inscribed.
_____________________________________
Exploration by C. Stefan, 18.Mar.2024 [about]
Last update: 18.Mar.2024 (versions: *) [versions]
"Essentia Mundi" AI Research Lab. [home]
Copyright © 2024 AI.EssentiaMundi.com, all rights reserved.
_
References: *
