Is there a rigorous definition of “machine behavior” as a research field? I can see how a good-enough marriage of ML/NN to classic symbolic AI could yield a sy

Is there a rigorous definition of “machine behavior” as a research field? I can see how a good-enough marriage of ML/NN to classic symbolic AI could yield a system capable of higher-order, intention-based “behavior” in complex circumstances. But we’re still a long way from achieving that, as far as I know.