A FAANG employee's labor is higher than a Uber driver's, driven by both supply (most people can drive an Uber) and demand (FAANG companies make a lot of money, something like >$1mm per engineer).
But getting back to the article, it only looks at the productivity of industries (which speaks to demand for certain types of labor), but the cost of training, say, a string quartet musician hasn't markedly decreased since the cost of education also hasn't fallen much.
If we talk about a highly skilled programmer requiring a lot of labor to produce, then we talk about something like "amortized labor" having uniform cost (the article mentions that it takes a lot of labor produce an orchestra player, for example). Even if some people are more skilled than other, if it takes a lot of labor to figure out who's really talented, you can price things this way.
As a human, working for a for-profit tech company for 8 hours coats approximately the same amount of energy / effort as volunteering for a non-profit tech company, but it costs the organization vastly different amounts of money.
As we improve technology fewer and fewer hours are required to create an item. But a violin concerto takes the same amount of human time.
Which then leads to the conclusion that we really exchange time when trading. Each of us only has a finite amount of time after all.