Hacker News new | ask | show | jobs
by sanderjd 3502 days ago
All of your hypotheticals seem to still assume individually-used cars. This is much nearer term thinking than the real interesting stuff. Cars wouldn't need to circle or park themselves. They would just go pick somebody else up. When you're done, some other car would come pick you up.
1 comments

Many people will prefer to keep individually used cars. Cars is not only transportation, but storage too. From children seats to umbrella to things you bought at previous stop.

The random self-driving car is pretty much the same as public transport. Individual cars seem to be still a thing, even though we have buses, taxis etc. Having random car pick you up is not some totally new idea that wasn't tried ever before.

Future humans will laugh at this quaint preference for one's own generic auto. Perhaps there will still be a point to owning specialized vehicles that e.g. go over difficult terrain or have extra comfort amenities. The idea that one Hyundai Sonata could be preferred to another Hyundai Sonata, however, is very odd.
People already laugh at people who love their vehicles to death. They don't drive a generic car though. Those people can usually explain why they drive that specific vehicle with those specific addons in no less than an hour.

For the record, I drive a stock car. I like few extra comfort bits I got, but many cars got that these days. I'm fine with people having fun in a way I don't get though.

It would be new: the price could be much lower, perhaps closer to public transportation, while the privacy and directness of pickup and drop-off would be like taxis.

The point about storage and kids is definitely interesting. Seems like potential for business opportunities if we ever find ourselves in this world.

They'd drive quite a lot of junk miles, which is virtually non-existant for public transit. And wouldn't have several passengers either. Taxis don't have economy of scale passengers-wise either, but they try to minimise junk miles a lot. Sometimes event rejecting travels to certain destinations or charging 2x to cover return.

I guess bottom line would end up ~ the same as current car sharing services. Cheaper if you want occasional use, but more expensive as daily driver. And availability issues at peak times.

I don't follow your logic here. Why would they need to drive more "junk miles" than taxis? It seems to me that there would be quite a bit more opportunities for optimization. How does the bottom line end up about the same as current car sharing services when you have removed one of the largest costs? Availability also seems easier to manage than with a fleet of people driving...
Over there, taxis refuse going out of city limits or even to remote parts of town where they know they won't have a passenger back. They have reserved spots for parking in many areas (just idle in grey-legal spots). There would be many autonomous cars and they'd follow rules more strictly I guess. So they'd have to drive off to find some parking. Then come back to pick up next customer. They'd drive a shitload of junk miles to pick up/drop off people to suburbs. Taxis either refuse such trips or charge $$$.

Current car sharing service (e.g. zipcar) is virtually the same as autonomous car sharing. You're not removing any large costs. Fleet balance is arguably the only cost removed. But it seems to be more or less solved problem even without autonomous cars. At least in my city, they no longer require to return (most) cars to the same lot. Wether car is autonomous or not, you still have to purchase the vehicle, pay insurance, gas, maintenance... Same price. The only change would be paying for idle car, it could drive back by itself. But user would have to pay for those junk miles. Which may be the same as paying for idling in the long run.

If car sharing service would want to have good availability at any time, they'd have a lot of overhead cars. Which would increase the cost. Otherwise they'd have to have reservations. And we get into a whole can of worms..