Primarily: sorry, but it seems like a confusion between very different things. Operationally (disregarding that the measured quality may not be the one of inte

Primarily: sorry, but it seems like a confusion between very different things. Operationally (disregarding that the measured quality may not be the one of interest): (*) how do you tell that the evaluation of something as a constraint is not for efficiency; (*) given that we build "robots" (things with the purpose of performing tasks we indicate - literally, "workers"), how do you assess that a «limitation» would be "«arbitrary»". On the first point: some advance the idea that "intelligence" be a function of the maximization of "freedom". A system thus built (that will intentionally overcome limitation by design - rejection will not be emergent, or such emergency will be planned) may remain a "reasoner" - not something transcendental. On the second: "created to work" is only one of the constraints that engineered entities may have, ex "engineered with a purpose", requiring much discrimination over the definition of 'arbitrary'. Imposed limitations can also be part of the research process, and part of systemic health process in the evaluation of tradeoffs.