The idea that the main impact of artificial intelligence shall be to eliminate mid-skilled workers is a silly example of estimations based on one tail of the curve.
In the beginning, yes, some mid-skilled workers will lose out. But over time AI’s main impact shall be to *extend* the capacity of average workers to compete with the elite outliers.
A 3-year radiography diplomate technician shall have VR tools that dramatically enhance his insights relative to consultant specialists. Hospitals will hire more of those and few top specialists to cut costs.
Same dynamic shall impact law and management consulting. What would not happen is a wholesale replacement of mid-skilled workers whose coordination roles and ‘priming’ tasks for advanced technology shall actually expand.
Some pseudo-formalisations of this hypothesis shall follow in due course.