Nice Myth
This dream has already fallen flat with Cargolifter. They had the same vision of transporting cargo to hard to reach places (like Alaskan oil rigs) and found out the hard way that demand was not sufficient to pay for the staggering cost of building large craft for that purpose.
When you're talking about tons of cargo, the additional design cost to design in a pilot's seat is small, and the additional operating cost for the human pilot becomes negligible. (That's one of the reasons why some pilots can command insane salaries -- it doesn't factor in next to the capital costs and fuel costs!)
Sure, airborne vehicles' advantage is that you don't need infrastructure between points A and B. But there's still some overhead needed at points A and B, like fuel delivery and storage, equipment for repairs, etc. And even without a pilot, you don't want your million dollar drones to crash, so marginalizing safety is not an option.
Eventually, except in a few rare cases, it will be cheaper to build a land-based network. And there's more innovative solutions beyond asphalt to choose from, such as aearial trams or riverboats.
Biting the hand that feeds IT
