It could work, if we let it.
I do think that driverless cars are going to make things like high-speed rail (particularly California’s plan) look even more monumentally stupid in retrospect than it does now (and that’s a high bar). But history teaches us that it’s very difficult to predict the societal side effects of technological advances.
They are making the same points about self-driving vehicles that I’ve been making for years: if every car on a given road is a SDV, and the road is limited access, then you can dramatically increase traffic density while increasing speed, so the passenger flux goes way, way up. This could be ready to go in 15 years, several years sooner if there were incentives for commuters to buy new compliant cars (I think the cars could be ready in 8-10 years). The technology is at the TRL 5 stage, but needs to be at TRL 9 before deploying on the roads in the manner I described; hard to see that happening in less than 8 years.
Note this is different from the non-cooperative SDV that Google is working on; you can’t get many of the benefits until you put in the vehicle to vehicle data link, and the predictability of response to anomalies (think TCAS).
The vehicle fleet turns over very slowly: the median American car has been on the road for 11 years. And even when half of cars have vehicle-to-vehicle data links (i.e. no sooner than 2029 or so, if they become mandatory in 2018), 75% of 2-car interactions won’t be able to take advantage of them because one or both cars won’t have the feature. For v2v to really help you need 80-90% penetration, which will only happen after a long time, and then only with government mandates.
By contrast, the Google approach offers immediate benefits to buyers, even if they never encounter another self-driving car. They can be adopted in an incremental, bottom-up way, like Internet services and consumer electronics, and manufacturers can rapidly iterate to meet consumer desires rather than having to target a fixed government-mandated standard. Eventually, once there are a lot of self-driving cars on the road it will make sense to take advantage of v2v, but I’d rather have the government clear obstacles to Google-style cars than force adoption of v2v.
“rapidly iterate to meet consumer desires rather than having to target a fixed government-mandated standard.”
Hmmmmm . . .
Yet another example as to why Central Planning is a failure – it cannot predict future inventions.
Driverless cars have one huge synergy with any long-distance mass transit: they’re likely to lead to much cheaper and more convenient “taxi” services, which will make it much less inconvenient to travel to a sprawling city without having your own car or your own rental car at the destination.
This doesn’t justify the weird modern left-wing train envy (for some reason America’s preferred high speed mass transit isn’t respectable because it can fly??) but it doesn’t make it even worse either.
Yes, we have that form of “high speed mass transit”, and I hate it. Flying these days is just another kind of bus, but what I don’t get is why a bus on rails is that much better than a bus with wings.
It’s funny how at the end they think it wouldn’t be a good future if we decided to take advantage of self-driving cars by abandoning mass transit and traveling by car, or if modular construction let us buy bigger houses. Sounds good to me! Typical urban elitist attitude.
.. or we could work on privatizing more roads.
You’ve all missed the obvious… SLOT CARS!
Divide the the intelligence between the road and the cars and upgrade incrementally at whatever pace works. If either is missing, it’s just a car on a road. When both exist, the driver no longer is. Cars in a slot would pack themselves together like a train but leave extra room for cars not automated.
BTW, planes should be easier to automate than cars but it hasn’t happened yet, even though planes have been capable of taking off and landing without pilot assistance since the 70s (747s.)