What's the computer science term for increase in timeouts?

Is there a feedback loop between complexity and abstraction?

• As systems become more complex, levels of abstraction are introduced. We can see this in computer science (libraries, frameworks, functions) and we can see this in society by looking at the average agent who does not know where all of their goods and services come from or how they are produced. My question is whether these abstractions that result from increasing complexity coupled with ignorance caused by an agent's limit of resources (such as time) to explore all alternatives result in sub optimal actions which in term increase complexity even more. An example: I am writing a computer program and I am ignorant about the intricate details of the communication between software and hardware. Because of  this I produce a product that is not optimally efficient and uses workarounds and other complex methods where they are not necessarily needed. As other people build upon my product it appears that this complexity becomes more widespread. Since agents are not able to grasp the entire (chaotic) system in which they operate (aka "the world") and thus engage in suboptimal decision making which leads to inefficiencies, do these inefficiencies then lead to more complexity (more chaos), abstraction (We try to make sense of a system that is even more chaotic and negate even more of the details) and thus to less understanding of "the world". Is this a loop?

• Answer:

  It seems you assume efficiency is attainable/desirable. I believe that, as systems adapt to their very specific environments, many compromises are made and take the form of couplings and useful abstractions becoming ever more specialized/concrete, increasing the overall rigidity of the system. There will always be room for inefficiency, as adaptation is a kind of dialogue/exchange between the System and its Environment, and there are no guiding principles beyond preservation of function/identity. I believe sub-optimal operation is inherent to complex systems since there are simply too many causal chains at work (most of them usually unidentified/misunderstood by whoever is observing/describing the System) sub-systems usually engage in local interactions/exchanges under many circumstantial constraints (circumstances that came to be as part of the overall Adaptation/Growth history of the system, and whose function is not necessarily still useful/required for the preservation of the system) There is no 'meaning'/Big Picture at the local level, allowing for a local exchange to actual hinder the overall preservation of the System (but in a well adapted system, sub-systems usually act as regulators amongst themselves in a loosely coupled way, allowing the System to reach/remain in homeostasis/equilibrium within acceptable parameter thresholds). In short, I agree with you: Complexity tied to Adaptation explodes the number of Abstractions/Concretions/Specializations, but efficiency becomes less important as the system reaches a critical mass; beyond that point, sub-systems most likely become Systems in their own right, and the amount of overall abstractions diminishes.