what do you think the end result is for the amount of machine-learning in use once the resource barriers go away?

like first the power availability limit, next cpu+memory availability?

result: every human job goes away that can be turned into a survey of questions

like all callcenters