This seems to assume that those scaling laws are a hard bound, rather than just a contingent fact about a particular machine learning method.
This seems to assume that those scaling laws are a hard bound, rather than just a contingent fact about a particular machine learning method.