Arrival times are more of a bell curve than evenly distributed, but the point stands with parking scaling down in some measurable way.

Naturally, yes. Inject a little surge pricing and you will find an impressively efficient load-balancing algorithm. That's a lot of the allure of an autonomous network vs. individual driver initiative - more top-down control means more network-positive, individual-negative decisions.

A 30-minute drive non-surge in LA is about $30 (and loses money, BTW). A 30-minute drive surge can be $75. Let's presume for the sake of capitalism that you're taking my vanpool. My system runs better if your ass gets up 15 minutes earlier. Will you do it if I pay you $10? How about $15?

Again, you're presuming there's no impetus to ditch parking lots. There's a big impetus to ditch parking lots. I've got a 20,000 sqft building. It's got 60 spots. It used to be a warehouse and only 20 people worked there; the other 40 spots are actually truck staging but when we converted it to office we restriped those. According to my municipal code I used to need a parking spot for every thousand square feet. If I wanna convert it to office I need one for every 300 square feet but I've got 100 square feet per employee, remember? Two out of every three cars is on the street already because I don't have any limitations on employees per square foot. But if I don't have to worry about those parking spots, I can build more employee space. I also don't have to worry about everyone else in the office park fighting over the one spot per three employees we're all running in the name of "efficiency."

I don't think everyone's going to bike 15 miles to work. But I also think that there are solutions between "this will never work" and "everything mentioned in the article." And that's because there's powerful incentives for the change to happen.

Much like my club on Pioneer Square.