The topic at hand seems to shift the quality of the discussion greatly these days. Many people have thoughts on coding agents because they are aimed at the lower quartiles of coders. Far less have detailed opinions on other ways they could wield a Markov model.