When outcomes are all that matters, then there is no point in learning anything.
When AI is "good enough" there is no point in excelling.
I don't think that follows. The relative contribution of the human might change significantly. If a product is worth 100 value units, and AI goes from contributing 0 to 95 of the final value units that's a fundamental shift economic shift even if the final product still ends up better with human input. The implications on wages alone are enormous. If a cheaper worker can produce 99 units of value instead of 100, why pay the substantial premium for that last 1%?
"But the top human + AI product will simply be better" you might say. Businesses don't always live or die based on whether they provide the best product. Optimizing for profitability by accepting a marginally worse product and redirecting the labor savings towards marketing and distribution is a very valid business decision.
Humans remaining necessary doesn't mean their economic value remains anything close to what it was.
It's not just magic to the unversed witness, but to the practitioners who no longer try to understand anything. They just become walking cheat-sheets full of incantations to implore some other entity to cause results?