I question this. The difference in reliability between a new car and a car that’s one-year old is minimal, but the price difference is considerable. Despite this, people buy or lease new cars all the time. My experience is that it’s largely vanity-driven; reliability doesn’t really come into it near as much as people like to think - it may be one of the excuses they give, but again, if cost-effective reliability was a real reason, they wouldn’t buy new, they’d buy a 1-2 yr old car. I would imagine that fuel efficiency and safety would be more important than reliability, and at least there you could make an argument that new technology in newer cars could be a differentiater.
I’m sure there will be a minority of the population that will continue to self-own cars, either due to specific circumstances that make it cost-effective, vanity or ‘get off my lawn’ old men shouting at the clouds. The rest of us welcome never having to waste money on a piece of metal that’s idle over 90% of the time in the driveway.
I’m really going to enjoy coming back to this thread in 20 years and laughing at people.
But we aren’t really talking about “new vs used.” We are talking about viable, convenient, and less expensive alternatives to new cars. I’m not even sure what you are arguing with me about. Used cars are a viable, convenient, and less expensive alternative to buying a new car. Sure there are pros and cons of buying used. But millions of Americans have weighed up those pros and cons and decided that the pros of buying used outweigh the cons of buying used.
Automated ubers are not currently a viable, convenient, and less expensive alternative to new cars. Some time in the future that may or may not change. But it hasn’t happened yet. When Uber or some other company does enter the marketplace people the choices that people make will be entirely dependent on how successful the rollout will be. It won’t just be a choice between “a new car” or “automated ubers.” It will also be a choice between used cars and traditional taxis and traditional ride sharing and walking and public transport and who-knows-what-else. Uber (or whoever it is that "wins the race to be first) will have to fight to get market share. I’m not convinced that Uber has what it takes. Somebody will get there. But when they do: they will always be just another option. And whether or not people adopt those services will depend entirely on the very same set of calculations that people are doing right now.
In a mature business model, that 25% should be plenty to manage dispatch and billing. Let’s imagine Uber operating an AV fleet. They have a fleet of self driving cars that they pay to own and operate, replacing the current share model, let’s call that division UberFleet they also have a division to do what Uber does now, let’s call that UberDispatch. The two divisions share revenue from the rides.
If UberDispatch needs significantly more than a 25% cut of the revenue to be profitable, their model is beyond broken. If that’s the reality of the business, I’m with you, it’s a non-starter. The cost of telling the car “go to this address, then that address” can’t be 25+% of the cost of the car itself. When I drive, that shit’s free, the CAR is the expensive part.
I think this is extremely unlikely, but the idea of a fleet model having extremely high barriers to entry is not. I’d think there is enough driving to be done to support multiple fleets, but the market dynamic could push toward a single fleet model.
Having bought a used car recently, what I can say is that people usually get rid of their new car when it starts needing serious maintenance. There are very few used cars on the lots that are not due for transmission overhauls, timing chain replacement, catalytic converter replacements, and probably the suspension is starting to get iffy.
There are not all that many 1-2 year old cars out there on the market, they are almost entirely off lease “certified pre-owned” type cars. They are few in number, and much higher in cost.
Buying a used car does mean that you are buying someone else’s problems, someone else’s service bills that they put off until they bought a new car, and have put them onto you.
Those who are admanst that people will own their own cars will probably continue to own their own cars. As someone pointed out upthread, there are people who own their own airplanes, even, and that’s much harder to justify as being practical.
The economics of it are no question at all. It costs me about $15 a day just to own my car, before ever even putting the keys in the ignition, just for loan payments and insurance. Maintenance is expensive, too, and is often times more time dependant than wear dependant, adding another several dollars a day to the cost. Gas is a tiny, tiny part of car ownership, but it is the part that many tend to focus on.
I don’t see Uber as being the end all be all in the market either. There are quite a number of rental car agencies that are already better positioned in the market to provide livery service. As is, they will come pick you up, and that’s with a human driver. I think that most of the AV cars on the road would belong to one of several competing company fleets of vehicles.
Calling a car for whatever need you happen to have at that time would be simple, and would be scheduled as far in advance as you would like (and pre scheduling should save you money). If you want to take a car to the other side of the country, that’s not that big a deal, as you can currently rent a car from a car rental company, and drive it to the other side of the country, and they don’t seem to have any problem with that right now (so long as you tell them that’s what you are doing), so why would that be a problem with driverless cars?
I don’t think that many people would be renting their cars out for public use very often. People that keep their cars would be doing so because they are worried about other people sitting in their cars. It’s not entirely rational, but if they have the money, to make that decision, then that’s fine. I know people that own far more cars than they are able to drive, and they don’t rent them out either. Just because you have a useful asset doesn’t mean that you will use it efficiently.
But, I would ask about that, even. If we have a more or less driverless AV model with most cars on the road belonging to company fleets, if there were to be a huge event or function in your home town, and you weren’t going to it, if the surge pricing for trips became high enough, would you consider sending your car out?
Keep in mind that most people do not ruin everything that they touch. Most people don’t vomit in car, or smear peanut butter around. There is absolutely nothing wrong with putting a camera in your vehicle. If someone makes a mess, they can pay to clean it up. If they make a mess, then they would also have a lower rider rating, meaning that they would have to pay more for rides, and you may not allow your car to be used by anyone with less than a stellar record.
After you made a few bucks for doing nothing but hitting “I agree” on your app, you may realize that you don’t work Mondays, or you work from home on that day, and that surge pricing from 8-9 is really high, and you think to yourself, “Self, I could make a few extra bucks here, with no risk to myself, just push a little button.” And you ensure that no one that doesn’t have 5 stars over 1000 ratings can be in your car, and you make a few bucks for no work at all.
I would find it tempting to generate revenue by doing absolutely nothing, no matter how attached I am to my car.
In a sense, people today are buying used cars from themselves, since the average lifetime of a car is 79 months. My 8 year-old Prius has never been to the shop due to a part wearing out. We sold our 12 year ols CRV and bought a new one for safety features, not because the old one was unreliable.
Back 30 years ago the argument that five year old cars were going to be in the shop all the time was a good one. Not today.
I think the optimal strategy is buying a two-year old car, which takes advantage of much of the depreciation, but that still is going to be relatively expensive.
One thing I don’t get about the Uber AV model.
Rush hour.
In the Bay Area, roads in the commute direction are jammed from say 6:30 am to 10:00 am. How is a ride service going to afford all the cars to meet this demand. Now, theoretically they can get a fare to go back to the source of the traffic, but given the low traffic volume in the counter-commute direction most cars are not going to get fares. So they drive empty for a good distance and try to get another fare. Some commutes, like along 580, are very long. Getting two a morning would be quite a feat.
Ditto for evening rush hour.
And what are these cars going to do during the day when there are far more cars than rides needed? Uber drivers can go off duty and do another gig - AVs can’t.
They’ll make sense in cities, and for short trips to and from mass transit, but most cars commuting are not going to be ride shares, especially not AV ones. What is going to be the average wait time at 5 pm? Someone stuck at work an extra half hour because no ride is available is going to ditch the idea real fast.
Self-owned AVs are another matter, since they convert useless time commuting to valuable time, and thus will have a high perceived value for the owners, while still being available.
This past weekend, we sold the family Toyota - 19 years old and still going strong, almost no time in the shop except for the required motor vehicle inspections - but it was too small for our family of four + dog + in-laws. We purchased a one-year old Sienta in essentially mint condition for 40% less than the price of a new car; easily seats seven, came with collision avoidance auto-braking, built-in sat nav, TV in the back etc.
Two car-sharing spots have opened up in our area in the past year. It would have been a serious consideration over buying a used car if one of them had been a bit closer.
One of the things is that with AVs, you are going to have far less congestion in the first place. We humans are pretty bad drivers, and so we need lots of space in between cars, both in the front and back, and on the sides. We need to keep longer following distances, and we bottleneck when we need to cooperate on lane mergers or intersections.
Getting rid of the human drivers and having AVs do the work is going to completely change those traffic patterns. One of the reasons for bottlenecking is specifically due to the failure of human reaction time and impatience. This is a major reason for slowdowns, and is something that could be completely eliminated by AVs that communicate and cooperate.
That compounded with the fact that a ride sharing system like this is going to be much more amenable to car or even van pooling will decrease the cars on the road rather dramatically as well.
So, car brings 4-5 people into the city, and takes 1-2 out in the morning, and reverses that in the evening.
Commute times are shorter, there is no stress of actually having to navigate the traffic and negotiate with other cars, and you are dropped off at the front door of your workplace.
Lots of space between cars? Clearly you haven’t driven in the Bay Area or especially LA. When I commuted, there was lots of cooperation in mergers.
We’re talking the economics of Uber-owned AVs here, not AVs by themselves. This benefit will still accrue when we switch to mostly AVs, which I think will happen.
In the Bay Area car pools get special lanes and cheaper tolls. Hasn’t helped much. In Silicon Valley you would be limited in working in a van pool by the fact that your competition might be in the pool with you. Plus you can’t make calls. Not true for private AVs. Google and Facebook buses might have the call problem, but not the proprietary information problem.
You’d have to charge each of those five - if you could get them - less than you’d charge a single rider. Even getting 1-2 back would be problematic if there are enough cars. And you still have the unprofitable downtime between rush hours.
Anyone who has taken something like SuperShuttle knows that having many people on the van increases your commute time a lot. Not that big a deal going to the airport once in a while, but every day?
I’m not saying van pools won’t exist, but don’t expect anywhere near 100% usage.
I recall very similar arguments made for van pooling. Hell, non-AV Uber could do it today - find drivers with vans, add a car pool option to the app, and you are good to go. And don’t say that the reason this doesn’t work is that the roads are too crowded - it is not like we’d transition to 100% AV vanpools right away. Individual AVs have benefits even with a single rider, so we can expect the traffic benefits to kick in over time.
Yes, lots of space. You think that that half a car length is tailgating, but it is wasting a half a car length of space. Self driving can function much more effectively, in fact, they could even link up and “wolfpack” through dense traffic areas if they were designed that way.
I see some cooperation in merges, but all it takes is one asshole who doesn’t cooperate to seriously decrease the traffic flow. And there is more than one asshole on the road.
Right, but your complaint was that traffic is congested, and therefore, you wouldn’t be able to move as many cars through as quickly. My point is that with AVs, you would be able to move much more traffic much more quickly.
All things that are very solvable. As you would have a much more flexible system for routing the rides in the first place, it would be much, much easier to ensure that you are not pooling with competitors, and in fact, it would make much more sense for you to be riding with co-workers, as you are all going to the same place.
There is also the fact that we are taking for granted that cars need unobstructed views through all the windows. This is something that the driver really would like, anyway. With self driving cars, putting up partitions between passengers would be far more practical.
And that’s how all livery services work right now. Reducing the cost of the trip by no longer needing a driver and the fact that electric cars should be much lower maintenance costs will only drive down the prices and make them more popular.
Yes, they will have downtime, no one is saying that they won’t, why is that a deal breaker?
Why does having more people in a vehicle increase your commute time?
I am not saying that either. Glad we are in agreement.
I do think that the advances in dispatching, both from a technology standpoint of having better computers and algorithms, and from a usage standpoint, where the greater the number of users makes the system more flexible, changes things quite dramatically.
For the first five or ten years, they are going to leave as much space as recommended today. For normal braking, like braking assists in cars now, you can leave less, but the processor is going to be handling other stuff. In emergency situations lots of things happen at once and leaving more stopping distance will decrease the odds that the car won’t get to braking before it hits the car ahead of it. Who can’t be assumed to be an AV car which communicates braking info. Real time programming is tricky, and it gets trickier the faster guaranteed response you need.
The two major sources of slowdowns I saw were road boulders and people moving from the car pool lane (to the left of the fast lane) to an exit lane very rapidly. The Bay Area uses metering lights which reduces the merge problem quite a lot.
Many large companies have tried van pooling, and have often gotten incentives from the state/county to do so. Back 25 years ago or so it was a big thing. It never made much of a difference.
Scratch the claustrophobes!
Livery services work for the well-to-do and for people who find them cheaper than the alternative, like airport parking. Remember, when we’re comparing prices we’re not just comparing against BMWs but against 8 year old fully depreciated cars. Like I said, Uber could do this today, and not worry about down time since the driver eats that.
Because it represents a significant amount of time during which the car won’t be earning any money.
Since you need to pick up those other three passengers and drop them off. The less dense the area, and the more places people work, the longer it is. I mentioned SuperShuttle. That works because hotels are mostly close together, and because so many of the rides go to or come from the airport. But my last ride picked me up way early. I had a meeting in the outskirts of San Diego once, and the van took forever to get there from the airport, dropping people all over the place.
A similar commuting service is going to have to pick up passengers early enough so that they can get to work by the required time, which is going to be a lot earlier than you’d have to leave yourself. How much sleep time will they be willing to sacrifice for this? There are a lot of non-monetary costs involved in this thing.
I often had late meetings. I’d have to order a car when the meeting broke up, and then wait for it, instead of just walking to my car in the parking lot. How much is that worth?
Like I said, once in a while is no problem, but every day?
Uber certainly has the technical capability of implementing a driver-based commute service today. You don’t get the supposed traffic benefit, unless you get enough users, but no real technical hurdles.
The issue here isn’t algorithms - I agree that AVs will solve all those problems. It is the value people put on their time and flexibility. Lifetime city dwellers may be fine taking an Uber or taxi to and from the grocery, but probably not the rest of us.
AVs will replace cabs and driver-Ubers, no question. But we’re talking about drastically reducing private car ownership, and I doubt that will happen.
Sort of; the original taxi companies had what are called “medallions”, which are in effect taxi licenses, and have traditionally been issued by muncipalities in limited quantities specifically for the purpose of regulating the industry and ensuring they could turn a decent profit. The alternative is any joe picking someone up and driving them to their destination (sound familiar?).
Uber and Lyft have somehow managed to upend that scheme and basically allow any old joe to pick someone up and drop them off at their destination. I’m not at all sure how they skirt the taxi medallion laws, but they clearly do.
But as a result, they aren’t profitable, many(most?) of their drivers aren’t profitable, and they’re wrecking the taxi industry wholesale.
The best I can assume is that their goal is to destroy the traditional taxi industry and then are banking on the idea that their position as a sort of dispatching system will allow them to be profitable even when the industry is essentially unregulated.