[seanw265]

This is not what the article says at all.

The article is about the constraints of computation, scaling of current inference architecture, and economics.

It is completely unrelated to your claim that cognition is entirely separate from computation.