Self-driving cars are modern public transit. "Train brain" is old thinking that public transit should take billions of dollars and decades to make fixed-route trains that only get travelers part of the way to where they actually want to go.
Trains and busses are far denser than cars, and that's part of the appeal. You can move more people around with less space needed.
"Train brain" is working just fine for China and Japan, I think it would work great in the US. Why can't I get on a train from Columbus to go to Philly? I should be able to.
> Trains and busses are far denser than cars, and that's part of the appeal. You can move more people around with less space needed.
This is true only for contrived examples comparing the vehicles directly. Comparing the real-world throughput of a highway vs. a rail line (in the US) and the density advantage goes away.
> Why can't I get on a train from Columbus to go to Philly?
The US has a different approach to property rights than China, so constructing a new right of way through thousands of private properties is much more expensive. Building new rail lines has massive environmental costs (it's a lot of steel and concrete!). Even once built, it's very hard to compete with a $100 flight that gets you between Columbus and Philly faster.
> This is true only for contrived examples comparing the vehicles directly. Comparing the real-world throughput of a highway vs. a rail line (in the US) and the density advantage goes away.
Does it? Can you post some information to back that up?
I'm not a transportation analyst, but this might be a reasonable comparison.
In Chicago there's a commuter rail-line that carries 6,171,000 passengers per year (~17,000/day).
There's a tollway that runs roughly the same route that carries 240,000 vehicles per day. Even if you divide that across 8 lanes you get 30,000/day. That number would go up if since there's often more than one person per vehicle.
Trains and rails are amazing at moving people between fixed points during rush hour, but are vastly underutilized the rest of the time.
In the future, it would be great to see rail lines replaced by dedicated self-driving lanes where cars could safely go 100+ mph.
> And building 20 lane highways doesn't take billions of dollars?
Building highways is very expensive, but much much less expensive than new rail. The most cost effective and environmentally sound thing to do is to make the most possible use out of existing infrastructure. As safe and predictable drivers (no fast stops and starts, swerving, or accidents), Waymo smooths traffic for everyone.
Depends on how you define public. The important characteristics to me are that there is affordable transport available to everyone. Government ownership and operation of the vehicles is not important to me.
Government ownership makes the most sense when there is a natural monopoly. There is not a need for a monopoly on urban transportation.
