I stopped reading at "soon to become the practice of writing software".
That belief has no basis at this point and it's been demonstrated not only that AI doesn't improve coding but also that the costs associated are not sustainable.
Because typing in text and syntax is now becoming irrelevant and mostly taken care of by language models. Computational thinking and sematics on the other hand will remain essential in the craft and always have been.
Care to link your sources? At least one of the studies that got attention here was basically done with a bunch of programmers who had no prior experience with the tools.
I continued reading, but you're right. Why did the author feel that it was necessary to include that?