Comment by throwaway6977

I think that completely discounting the potential of new emergent capabilities at scale undermines this thesis significantly. We don't know until someone tries, and there is compelling evidence that there's still plenty of juice to squeeze out of both scale and engineering.