Hacker News Reader

Comment by jseliger | Hacker News Reader

I live in Phoenix and now take Waymo regularly, and it seems like we're close to a world in which most people take self-driving cars most of the time, crash rates plummet, and these kinds of articles come to resemble articles from 1910 about horse-related problems.

Humans suck at driving: https://jakeseliger.com/2019/12/16/maybe-cars-are-just-reall...

View on HN

Define "close."

(Anecdotally, I'm outside Harrisburg, and there are no self-driving cars in... this state? People drive their cars just like cave people did, with their hands on the wheel!)

Of note, the article mentions your sentiment near the conclusion:

> Perhaps one day we’ll all whizz around in self-driving cars and accidents will be rare. But that’s still a ways off.

> I'm outside Harrisburg

Hell, out there you've still got people using horse and carriage.

Phoenix doesn't really have the related weather concerns of the midwest which allow Waymo to flourish in your state.

It’s also a grid roadway as a post-automobile city.

Granted they got SF working too, which is pretty varied terrain but it took a lot of training. I guess we can roll it out city by city but will they ever recoup the cost?

I've long thought that the bus market would be ideal - the city bus doesn't take most of the roads in your city so you don't need to program them in (or than a dead in). Just make sure the roads department works with the transit department to program in just the detours needed before starting work.

Phoenix has extremely wide and straight roads, with very few pedestrians, and very little rain or snow.

Compared to the entire world this is VERY anomalous, so I think we're still pretty far from "most people" using self-driving cars.

I welcome self driving technology when they become available, but from a system perspective, cars and car related infrastructure imposed tremendous cost on our environment and urban fabric.

I wonder what we are going to do with all of the car parking that will become as obsolete as horse stables.

This is actually the thing I am most excited about with the prospect of self-driving, the biggest way car-centric infrastructure ruins cities is by parking requirements.

Unfortunately most cities in the USA have parking minimums so even if a market change occurs and they are all empty the lots will stick around for a while until they are no longer legally required.

I think it is silly to think self driving cars gets rid of parking. If you don't use a car much it is cheaper to use a taxi service. However people who use a car often will be money ahead owning their own. Plus if you own the car you can keep your "junk" in it while doing something else. Most cars are needed during rush hour, so there isn't much need for cars the rest of the day.

All of which means we need nearly as much parking in the future. Sure we save a little because the car will leave your immediate area (read CO2 and other emissions) for free parking out of sight, but they are still parked. You will want it parked somewhat close to you because you never know when you will have a family emergency and have to leave early.

On a tangent, I wonder how many car manufacturers would actually sell a self-driving car if they were able to build one. Creating a taxi fleet and charging for usage seems much more profitable.

Or at the very least, the price would go up a lot. Who cares about the cost, when the value of the sold good is so high? Why sell it cheap? I note that Waymo has no intention of selling the things.

Most of them. Car manufacturers aren't too greatly in vertical integration. One could argue that having a car rental agency for manufacturer would be great idea. Get cars at cost, rent them short to medium term, keep maintenance inhouse sell them after some period for most of the price.

But I don't think there is any manufacturers that rent vehicles for short term and neither there are any rental firms that manufacture their cars...

In the end making things and renting them is different businesses.

The economics don't work. Some people are willing to pay extra for a car in perfect condition. Some people will happily save money by using a beat up car. By selling a car they get their money now from a high priced car. If they try to be the taxi they need to figure out who will demand (and pay for) the perfect car, vs who will demand the cheapest car. This is one more way that cars are not interchangeable.

I envision a bunch of those car carousels with charging plugs!

https://www.siemens.com/global/en/company/stories/research-t...

https://www.parkmatic.com/automated/carousel/

We’ve been “close” for 10+ years now. We are not close.

5 years ago you couldn't hop in a commercial self-driving taxi in a major US city. Now you can. The progress has been slower than everyone's predictions but it is still substantial.

In 1897, you could hop in a commercial electric taxi in London. Alas, Londoners still have many gasoline cars on the road.

Availability and mass adoption are two very entirely different things.

The time in which people take most commutes by any kind of car is: today... 51% of people commute by car currently. That took 136 years.

'most of the time' for 'most people' is quite a high bar. That's a lot of travel, for about 4 billion people. A few people in Phoenix taking one after a night at the bar doesn't even to begin to scratch that surface.

We're very much still at the 'Carl Benz demoing his horseless carriage' stage of self-driving cars.

IF, and this is a big if - self driving cars prove safer than human drivers in all conditions - governments will mandate it in all new cars and in 12 more years they will be the majority, and a few more years and governments will ban everything else (if not directly insurance via high rates).

However right now they are not ready for that. They don't work in all conditions and we don't have unbiased safety studies.

For varying definitions of “governments” and “will”. You’re looking at this from a very western perspective.

For example, airbags have proven to drastically increase safety, yet almost a half-century later they’re still not required by many countries across 2 and a half continents. Governments only require safety equipment as far as their population can afford it.

Most governments don't need to as they can leach off of it not being worth making a model without airbags.

Except that didn't happen. Leaving out parts is pretty easy to do, especially when those parts are hundreds of dollars each.

Global automakers have, and still do, make different vehicles for different markets. And further, some only make vehicles for developing economies.

For example, Pakistan required airbags in 2022, and they found necessary to do so because automakers were still making vehicles without them. https://propakistani.pk/2021/12/27/pakistan-to-ban-cars-with...

65 years ago we had never traveled outside Earth's atmosphere. 50 years ago we had traveled to the Moon. Surely by now we are visiting the nearest star, no?

It's incredible to me how many people honestly believe they can predict the future, when the future is so clearly unpredictable. Even while you acknowledge that everyone's predictions were too optimistic, in the very same sentence you insist on an optimistic prediction.

I didn't make any prediction at all.

Waymo in SF is super expensive. Like, $1/min+. Is Phoenix more financially accessible? I’d love to ditch my car for Waymo but I’d easily drop $40+/day which is a lot more than, say, a $300/month lease.

Reply on news.ycombinator.com