Lyft Says Robots Will Drive Most Of Its Cars in Five Years (recode.net)

A week after its rival Uber began rolling out self-driving cars in Pittsburgh, Lyft has said it also expects to roll out its self-driving by next year. Its president John Zimmer outlined a "three-phase" plan for the company, noting that self-driving cars will be made available to Lyft users in the first phase. But in this phase, it only plans to roll out self-driving cars that can "drive along fixed routes" and that the "technology is guaranteed to be able to navigate." Recode adds: In the second phase, the self-driving cars in the fleet will navigate more than just the fixed routes, but will only drive up to 25 miles per hour. As the technology matures and the software encounters more complex environments, Zimmer wrote, cars will get faster. The third phase, expected to happen sometime in 2021 or 2022, will be when all Lyft rides will be completed by a fully autonomous car. Shortly after that phase begins, car ownership will see a steep drop-off, according to Zimmer. Zimmer, who has long been a vocal proponent of ending car ownership, set a date for the death of the personally owned car in major U.S. cities: 2025.

Comments

So a guy that runs a ride sharing company.

[sims 2]

Says that self driving cars will put an end to car ownership.

It could put a dent in it but unless this makes people so broke that they can't own their own car I think personal space will still win out.

Re:So a guy that runs a ride sharing company.

[PIBM]

Reason #1 it won't happen; families.

Here I have 4 kids, each requiring different car seats size / adjustments. We are bringing and keeping different stuff for the kids (Stroller, diapers, their favorite movies) which stay with us without needing to grab it at our Xth destination. Keeping our previous purchases safe while we go for our next stop, and the items we don't need at that stop (no stroller at the grocery store) is a major win that lyft rides won't provide.

It might definitely help reduce the percentage of ownership, but it certainly can't sign the death certificate.

Re:So a guy that runs a ride sharing company.

[drinkypoo]

Says that self driving cars will put an end to car ownership.

Even the automakers are expecting ride sharing to put a major dent in auto sales. Their analysts have been talking about this for years and most of them are invested in ride sharing. They are definitely talking about self-driving cars being part of this equation, and again, have been for years.

People who can't really afford to own a car responsibly often own one anyway, because they can't function without one in a world so dependent on private transportation. Most nations fit that description; people using public transportation suffer badly compared to vehicle owners because of public transport's many deficiencies. Self-driving cars have the potential to eliminate virtually all of those deficiencies. You can get a car when you want one, you don't have to worry about whether the driver is fit to do his job because there isn't a driver, and the vehicles don't inherently cause traffic flow problems with other vehicles, decreasing the overall efficiency of the system.

Since the economy isn't exactly improving, you can reasonably expect vehicle ownership to continue to decrease. The age of the U.S. fleet in particular continues to increase to ever-higher record levels. People are buying less cars, that's a fact. They're buying less cars both because they can not afford as many cars because the economy is still in the toilet, and because they can better function on public transportation than in the past because some new options have opened up under the name of "ride sharing". Whether they are ride-sharing or not isn't really the point here (though they aren't) but that they now exist when they didn't before. Yes, those are private transportation systems, but anyone who does not abuse them can use them.

Re:So a guy that runs a ride sharing company.

[tomhath]

If taxis are so much better than car ownership, why haven't they taken over already? I can't see how the self-driving aspect will improve the convenience or lower the price of a ride enough to make a revolutionary difference.

Re:So a guy that runs a ride sharing company.

[DirkDaring]

And can you imagine haling a 'cab' for every stop in your running around at around $10 a pop?

Re:So a guy that runs a ride sharing company.

[DirkDaring]

If you can't afford a used car, you definitely can't afford to be spending way more in 'cab' fares at around $10 a 1-way trip.

Re:So a guy that runs a ride sharing company.

[flyingfsck]

Cheap taxis are dirty. Clean limo taxis are expensive. Lots of people want their own clean cars with some personal stuff permanently lying inside - sports gear, hats, gloves, umbrellas and more. Shared cars can't offer the same convenience.

Re:So a guy that runs a ride sharing company.

[Hadlock]

Because taxis are awful and primarily exist to shuttle people between hotels and airports, and convention centers. Being able to rate your driver as "poor/good/excellent" and then cull out the bottom third (and force people to act civil to push them in to the top third) makes the experience much like being driven to school by your mother. I lived in Dallas and it was cost effective (I actually saved about $3,000 a year after insurance, maintenance, wear and tear items, not to mention speeding tickets and parking in downtown)... now here in SF where parking is $280/mo in a private garage, it makes perfect sense. My lady friend owns a car but we only use it for trips out of town. If we're going somewhere to dinner or a show we always take an uber -- parking is insane and effectively impossible. All the street parking is taken up 100% of the time by residents who don't want to pay for private parking. As more people move in to cities, private car ownership can't go up. Private ride share makes perfect sense.

Re:So a guy that runs a ride sharing company.

[drinkypoo]

If you can't afford a used car, you definitely can't afford to be spending way more in 'cab' fares at around $10 a 1-way trip.

You are forgetting TCO, and that the 'cab' fares are going to decrease because no one is going to have to pay a driver, and the bar to entry into the market is diminishing as Uber and Lyft push legal changes that reduce it — the same regulations designed to keep them from entering the market are the only thing that keep potential competitors out. If these companies actually succeed in changing the legal landscape sufficiently for their business model to be seen as generally viable, they'll be drowning in competitors quicker than you can say "automakers".

Used cars are cheap, but keeping them on the road isn't. You should see the A8 I just picked up for $200, it looks better than the one I've got already and it's got half the miles. But it needs a new slush box and best case that's around $2500. And... what will fail next?

Re:So a guy that runs a ride sharing company.

[ShanghaiBill]

And can you imagine haling a 'cab' for every stop in your running around at around $10 a pop?

What you are missing is that it won't be $10 a pop. Today, the biggest expense is the driver. Once the driver is gone, the price will fall dramatically. Also, a typical Uber/Lyft driver today drives for about 4 hours a day. A self-driving-car can operate 24/7. So the increased supply will drive down the price, while the lower amortization cost per ride will drive down the cost even more.

Uber and Lyft will also face pricing pressure from improvements in mass transit. Even for a bus, the biggest cost is the driver. As big human-driven buses driving fixed routes, are replaced with small self-driving vans driving flex-routes, prices will go down and ridership will go up.

Re:So a guy that runs a ride sharing company.

[sl3xd]

Agreed; unless the service costs about the same as energy (fuel/electicity) + the standard vehicle deprecation per mile (looks to be ~$0.54/mile), it'll be a very hard sell.

Even then, it would be a hard sell for my own daily commute: that works out to $7k/year for commute costs, which is 3-4x more than what I pay for over the life of a personally owned vehicle (including buying new cars, fuel, maintenance, insurance, registration, etc.)

Re:So a guy that runs a ride sharing company.

[zlives]

all of this points to a better public transport system even if it is automated cars.

Re: So a guy that runs a ride sharing company.

[zmooc]

All these kid requirements can easily be fixed in a self driving car. The only reason we don't have automatic seas that can scale down to the size of a child is that they're not economic. They will be in a self driving car that's used all day.

Re:So a guy that runs a ride sharing company.

[Thud457]

Get back to me when Roomba can avoid smearing dogshit all over the house.

Then maybe I'll consider the possibility that a robot car can safely discern a child running into the street from a abandoned shopping bag blown in the wind.

Re:So a guy that runs a ride sharing company.

[BlueMonk]

The bus service in my area has a monthly pass option. I suspect Lyft would offer multiple or alternate pricing plans with a similar idea that can make the service more cost effective for people depending how they use the service.

Re:So a guy that runs a ride sharing company.

[Etcetera]

... now here in SF where parking is $280/mo in a private garage, it makes perfect sense. My lady friend owns a car but we only use it for trips out of town. If we're going somewhere to dinner or a show we always take an uber -- parking is insane and effectively impossible. All the street parking is taken up 100% of the time by residents who don't want to pay for private parking. As more people move in to cities, private car ownership can't go up. Private ride share makes perfect sense.

The problem here is San Francisco. Massive development, poor freeways, high-density urban living in glorified dorms, zero parking, and eco-nuts who are trying to make a point.

Perhaps Uber/Lyft makes great sense there, like taxis in New York and the massive, mass public transit investment. In most of the rest of the country, the economics don't make sense unless you're going to try to force people to use them by making parking very inconvenient, and/or taxing the hell out of everything involved with having a car. That might be one way to achieve it, but that's hardly a fair win.

Hell I recently *moved* to a downtown urban area and parking's a pain for visitors, but not insurmountable. And every condo or rental unit comes with at least one parking space for it (usually 1 per BR). Why? Because San Diego developers are realistic about car needs and aren't trying to force people into their preferred civic paradigm. Bay Area problems come from a manifest unwillingness to face inconvenient truths.

Re:So a guy that runs a ride sharing company.

[swb]

Totally autonomous cars will have to be part of this for it to succeed, otherwise there won't be enough car owners and drivers to drive around the people who don't own a car.

I feel like automakers have been talking about this like Elon Musk is talking about going to Mars. I don't doubt that self-driving cars will be a reality, but I'm really wondering if they will be a reality in my lifetime (I'm 50).

Re: So a guy that runs a ride sharing company.

[ShanghaiBill]

You realize that the biggest cost is the driver because the driver is paying for the car right?

No. Do the math. A typical Uber driver grosses about $18/hour. If he buys a $30k car, keeps it for 10 years, and drives 4 hours per day, 5 days a week, then that is about $3 / hour. Much more is paying for the driver than the cost of the car.

For an SDC that drives 20 hr/day, 365 days per year, the cost per hour is about 50 cents.

Re:So a guy that runs a ride sharing company.

[CWCheese]

Once the driver is gone, the price will fall dramatically.

Yeah, that'll be the day. No, the price won't fall dramatically, indeed it will like increase so as to expand the revenue stream for Lyft/Uber/XYZ.

Also, think of all those drivers who will now be out of work, the shared ride service drivers, regular taxi drivers, bus drivers, delivery drivers, over the road truckers. How are they gonna afford the nominal $10 bucks a ride, presuming the price stays the same?

Re:So a guy that runs a ride sharing company.

[ShanghaiBill]

all of this points to a better public transport system even if it is automated cars.

If Uber/Lyft/etc can provide self-driving cars and vans that are faster, more convenient, and cost competitive with existing public buses, then maybe it is time to get the "public" out of transportation. At least in America, the government has done a horrible job of running public transit. Maybe the government should focus on other priorities, and leave transit to private companies that have an incentive to do a better job.

Re:So a guy that runs a ride sharing company.

[ShanghaiBill]

Cheap taxis are dirty.

Uber and Lyft rides are cheaper than taxis, and every time I have used either the cars have been very clean.

The taxis are dirty because the owners aren't concerned about bad reviews. The Uber/Lyft drivers are.

Re:So a guy that runs a ride sharing company.

[uberdilligaff]

You've hit the nail on the head. This craze will end soon after one of these driverless wonders runs over a 5-yr old kid and the operating company is sued into oblivion or legislated out of existence.

Re:So a guy that runs a ride sharing company.

[ShanghaiBill]

No, the price won't fall dramatically, indeed it will like increase so as to expand the revenue stream for Lyft/Uber/XYZ.

Nonsense. Companies can't just arbitrarily set prices. If they could make more money with higher prices THEY WOULD ALREADY BE DOING IT. If Uber and Lyft raise prices (or fail to lower them as their costs fall) then other companies will jump into the market. This is exactly what happened when they pulled out of Austin: Other ride sharing apps were up and running in a few weeks.

Also, think of all those drivers who will now be out of work

They could be paid to throw rocks at windows to create jobs for glaziers.

Look, pointless make-work activities are not "good for the economy". As people spend less money on rides, they will have more money to spend on other things, meaning that people will have an improved standard of living, and new jobs will be created to provide those new goods and services. Increased productivity is the reason, and the ONLY reason that you don't live in a shack with a dirt floor like your ancestors.

Re:So a guy that runs a ride sharing company.

[tnk1]

Except people today run over 5 yr olds too, and drivers get sued and even jailed all the time for negligence and it hasn't changed anything.

These rideshare/automated cabs will have ensured that the legislative environment for their vehicles is able to properly protect them from every little lawsuit due to a glitch.

Their first defense will likely be to trundle out the statistics showing that they have actually decreased their company's rate of accidents by moving to driverless vehicles. Which will most likely be the case unless they completely fail to test their vehicles and the software properly. And at that point, they would deserve to be sued into oblivion for complete failure to use common sense.

Discussion of driverless cars seems to mirror some of the discussions of the original automobiles. And honestly, autos built back then were significantly more dangerous than vehicles driven by software are likely to be today and they still took right off as soon as everyone understood how useful they were.

Which is not to say that the fears were entirely unfounded. Auto accidents are a big killer of people, certainly much more so than terrorism could even dream of. And that being the case, cars have only been increasing, not decreasing, in utilization since they have been invented. This is not likely to be any different for self-driving cars, if they can even improve on the safety record of human drivers even just a little bit.

Re:So a guy that runs a ride sharing company.

[tnk1]

It might be more efficient, that doesn't always equate to better service.

Of course, replacing inefficient, but shitty government buses with efficient, but still shitty private buses might make sense in areas who already have poorly managed transit. It will be interesting to see if they can overcome public busing with a private busing service, but due to government backing and the politics of mass transit, that's going to be harder to do.

Re:So a guy that runs a ride sharing company.

[PopeRatzo]

Here I have 4 kids, each requiring different car seats size / adjustments. We are bringing and keeping different stuff for the kids (Stroller, diapers, their favorite movies) which stay with us without needing to grab it at our Xth destination.

By the time there are ubiquitous self-driving cars, your youngest will be asking if he can have the keys to that old 2019 Pinto in the garage, because he knows it's too expensive to take his girlfriend out to the ridge to watch the submarine races in an self-driving Lyft.

Cars are about privacy. I'm shocked that the Elon Galt types don't get that.

Re:So a guy that runs a ride sharing company.

[SethJohnson]

Taxi availability thus far has been constrained by city regulations designed to protect the commons from congestion. The philosophy behind limiting the number of taxis on the road at any given time is also intended to motivate people to use mass transit instead of inefficient taxi services.

Ride hailing companies have successfully flaunted these regulations due to widespread political support among people who are only thinking of their own convenience over the bigger picture. Ride hailing puts more cars on the road performing empty trips creating congestion and pollution. Instead of voting in support of bond elections to implement planned mass transit systems, people are placated by "I want it now. Me!" provided by ride hailing.

Re:So a guy that runs a ride sharing company.

[luis_a_espinal]

Reason #1 it won't happen; families.

Here I have 4 kids, each requiring different car seats size / adjustments. We are bringing and keeping different stuff for the kids (Stroller, diapers, their favorite movies) which stay with us without needing to grab it at our Xth destination. Keeping our previous purchases safe while we go for our next stop, and the items we don't need at that stop (no stroller at the grocery store) is a major win that lyft rides won't provide.

It might definitely help reduce the percentage of ownership, but it certainly can't sign the death certificate.

That's a very particular way of viewing things, very particular of this country. In cities with excellent public infrastructure (say Tokyo or the Kanto Region as a whole), you truly do not need a car, even with kids.

So the fact that it cannot happen here shows how poorly designed our cities are in general, and how costly that is. We own cars not because we are (or like to be) independent, but because we are crucially and almost fatally dependent on them for basic transportation in a way almost unimaginable in a 1st world country.

Re:So a guy that runs a ride sharing company.

[SethJohnson]

....maybe it is time to get the "public" out of transportation.

Cost competitive with public buses? This weekend I dropped my truck off at a stereo shop in the south end of town (Austin, TX). Paid $2.50 for an all-day bus pass and came home. Did chores all day and rode the bus back down to the stereo shop to pick up my truck. That would've cost me over $20 with a ride hailing company.

Looking around at the other passengers on the bus with me, I got a sense many weren't really positioned financially for becoming ride hailing customers. At least one was in a wheelchair that required a lift that wouldn't be equipped on many privately-owned vehicles. Advocating for ride hailing to replace mass transit does not consider the needs of those for whom the bus is the only transportation 'option'.

Additionally, if a ride hailing car travelled to my house (empty) took me to the stereo shop, then returned to the driver's house (empty), that would add two unoccupied trips on the public street that wouldn't have occurred had I ridden the bus or driven myself. If you multiply those empty trips times the number of people on the bus, suddenly there would be a lot more cars on the street not carrying passengers and slowing everyone else down with greater congestion.

Re:So a guy that runs a ride sharing company.

[PIBM]

"An urban area is the region surrounding a city. Most inhabitants of urban areas have nonagricultural jobs. Urban areas are very developed, meaning there is a density of human structures such as houses, commercial buildings, roads, bridges, and railways. "Urban area" can refer to towns, cities, and suburbs."

I am living in what you would call 'suburb' area. While I'm no longer in a major US city, I lived in San Francisco for a while, and there's no way we would have done what we did if we had relied only on public transportation / lyft / uber or such. I do rent cars sometimes but that do cost much more to provide the benefits that owning one does.

I was mostly pointing out that even in major us cities his point was invalid since his proposal would not solve major use cases of cars. From the summary : " Zimmer, who has long been a vocal proponent of ending car ownership, set a date for the death of the personally owned car in major U.S. cities: 2025. "

Re:So a guy that runs a ride sharing company.

[CWCheese]

You're making pointless presumptions about people actually having money "to spend less" of when they have no work opportunity

Re:So a guy that runs a ride sharing company.

[OrangeTide]

For people living in a city, it may be cheaper to use a taxi service such as Uber or Lyft than it is to own a car.

Parking space can be expensive in some cities. Then there are state registration fees which are going up, and bound to skyrocket as the gas tax becomes less effective with electric vehicles. Currently the interest rates for car loans are low (at least for some), but if the FED adjusts the interest rates that can trickle down to the banks that specialize in low cost auto loans. (my first car loan was very high, nearly the same interest as my credit card, but I know now that I was severely ripped off, and managed to get much better deals since then)

PS - I hope you enjoy the "taxi" jab at the ride sharing industry. If I hire a car and drive for a single trip, that's a taxi. It may or may not fall outside of a municipality's monopoly, but that's for lawyers to figure out. (I don't care about the legal aspects)

Re: So a guy that runs a ride sharing company.

[ShanghaiBill]

I've seen reliable numbers that the TCO for a vehicle is around $0.50/mile.

That is for a normal person, not for someone driving for a living. Most of that cost is the depreciation of the vehicle. If you are commuting 20 minutes each way, that depreciation is spread over far fewer miles than if you are driving 4 hours per day.

Re:So a guy that runs a ride sharing company.

[OrangeTide]

I don't know, seems like older kids (tweens) that have to go between various after school activities might be good use for self driving cars. It would really ease the burden off parents, who exhaust themselves every day lugging their various children around to multiple locations.
As for safety, you know self driving cars will have tons of internal cameras as a vandalism deterrent and possibly using a human to do facial identification at a central office before accepting the fare.

Re:So a guy that runs a ride sharing company.

[OrangeTide]

When I got to SF to meet up with friends, I usually park at a BART station outside of the city and ride in. Parking can be kind of unreliable to find depending on when and where, and driving around is circles until you find something sucks because driving in the city is a big pain.

Re: So a guy that runs a ride sharing company.

[ArmoredDragon]

I've always paid cash for my cars, so my monthly (and overall) costs are lower since I don't pay any finance fees/charges. My insurance is about $700 per year, repair and maintenance costs rarely exceed $200 per year, registration is $250 per year. I paid $8k for my current car, (2013 Camry SE) and usually own my cars for about 8 years, expected to sell for about $3000 at that time, so this time around roughly $625 per year just to own the car.

So my car costs me roughly $148 per month to own and use, not including fuel, which depending on usage and pricing is probably another $100. Using a taxi service of any kind instead would easily exceed $248 a month.

Re:So a guy that runs a ride sharing company.

[nukenerd]

They will not lower prices, despite reduced costs, because Joe Public is conditioned to the prices as they are at present. Lots of things have become automated without the price being reduced. The company keeps the savings themselves.

Re:So a guy that runs a ride sharing company.

[ShanghaiBill]

They will not lower prices, despite reduced costs, because Joe Public is conditioned to the prices as they are at present.

Horse pucky. They are offering a commodity service, with no "lock-in" for either drivers or riders, and people will switch instantly if they can get a better deal. I have used both Uber and Lyft, and they are basically indistinguishable. Most Lyft drivers are also Uber drivers, so it is even the same car and the same driver. The only thing different is the icon.

Lots of things have become automated without the price being reduced.

That is only true if they have either a monopoly, or a strong brand that is identified with a big difference in quality. Apple can get away with charging more than Android, because their customers do not perceive them as being interchangeable. There is no "brand premium" in ride sharing.

Re:So a guy that runs a ride sharing company.

[gregersonke]

It can't happen. 1. it's not handicap compatible nor does it offer wheelchair assistance that a driver would. 2. while driver expense will go down the maintenance expense will rise on account of shorter maintenance cycles to check damages from a bump in the road etc. 3. Rider safety is decreased on the van level as you can't protect passengers from each other. What can happen, primary is that Drivers can be paid less because they now do less. But questions will remain 1. to sub out liability can they make it so when passengers are present one of them will be behind the wheel and driving with aid of computer unless passenger is intoxicated 2. what happens when passenger falls asleep and doesn't leave the car 3. what happens if passenger dies during travel. 4. what happens when passenger needs emergency assistance

Re:So a guy that runs a ride sharing company.

[ShanghaiBill]

It can't happen. 1. it's not handicap compatible nor does it offer wheelchair assistance that a driver would.

The ADA requires "reasonable" accommodation for disability. It doesn't say you have to shutdown your business if you can't do it. I rent out a spare bedroom on Airbnb, despite the only access being up a flight of stairs.

3. Rider safety is decreased on the van level as you can't protect passengers from each other.

Is that why we only let one person at a time into elevators?

Re: So a guy that runs a ride sharing company.

[FunkSoulBrother]

Just a thought but you could... walk somewhere.

Re:So a guy that runs a ride sharing company.

[MBGMorden]

My cars been paid for for ~6 years now. Insurance is $37 per month. Fuel is about $80 per month. Maintenance comes in waves but averages around $50 per month. I don't know of anywhere within 60 miles that doesn't have parking included as part of a lease (and I personally own a home anyways so my driveway doesn't cost extra).

It seems like people try to nickel and dime THEMSELVES when it comes to justify a purchase. Car ownership isn't that expensive.

Re: So a guy that runs a ride sharing company.

[pgnas]

You can't see it, it's 2-4 generations away, out of your sight, You have to think long term. Cars will eventually be too costly to own and operate, you can be sure of it. The business model of ride sharing sucks now, due to the cost of the idiot behind the wheel, why would they (those ride sharing companies) be so quick to lose money and stay in the red? The owners and every other company know that jockeying for position now will set them up for life. Your kids grandchildren will not know what it means to pick up and go on a road trip across the country or otherwise. They won't own a car, their grandparents might have one in the garage (hehe "Red Barchetta") , but drive it? No. Everyday there is an article out there about self driving cars, autonomous vehicles, robot this and robot that, it's hype, it's the greatest thing, right? Not.

Re:So a guy that runs a ride sharing company.

[AmiMoJo]

In other words the litigious nature of US society will prevent it taking a lead in robotics and self driving vehicles.

Re: So a guy that runs a ride sharing company.

[Gussington]

For an SDC that drives 20 hr/day, 365 days per year, the cost per hour is about 50 cents.

A car driving 20 hours a day will last a couple of years at best. ( 2years @ 20 hours/day =14600 hours averaging say 30km/h = 438000kms, longer than the life of an average car) So you need 5 times as many $30k cars to get 10 years of driving out of it.
So $150k worth of cars works out to $2/hour before running cost (fuel, rubber etc)

Re:So a guy that runs a ride sharing company.

[Gussington]

Yeah, that'll be the day. No, the price won't fall dramatically, indeed it will like increase so as to expand the revenue stream for Lyft/Uber/XYZ. In a free market all this means is the RideSharev2/RobotCarv1 will come along and eat their cake.

Also, think of all those drivers who will now be out of work, the shared ride service drivers, regular taxi drivers, bus drivers, delivery drivers, over the road truckers. How are they gonna afford the nominal $10 bucks a ride, presuming the price stays the same?

Because progress has never made jobs redundant. Maybe they could get jobs fixing all the robot cars that will soon exist?

Re:So a guy that runs a ride sharing company.

[Gussington]

Except people today run over 5 yr olds too, and drivers get sued and even jailed all the time for negligence and it hasn't changed anything.

Crap. Legislation is modified all the time exactly due to such events, ie in order to improve safety and lower risk. It's precisely why the road death rate has been steady decline over the last hundred years. That isn't just pure chance.

Re:So a guy that runs a ride sharing company.

[goose-incarnated]

In other words the litigious nature of US society will prevent it taking a lead in robotics and self driving vehicles.

Self-driving vehicles aren't going to happen in your lifetime. You need strong general AI for that. If we had that we'd have reached the singularity and probably won't even need cars.

Re:So a guy that runs a ride sharing company.

[drinkypoo]

Additionally, if a ride hailing car travelled to my house (empty) took me to the stereo shop, then returned to the driver's house (empty), that would add two unoccupied trips on the public street that wouldn't have occurred had I ridden the bus or driven myself. If you multiply those empty trips times the number of people on the bus, suddenly there would be a lot more cars on the street not carrying passengers and slowing everyone else down with greater congestion.

Well, this is where the government could provide some assistance, if the people running it acted in good faith. They'd operate the ride-sharing website, you'd open it up and you'd be able to see all the options and how much they cost, from free (someone is willing to actually engage in ride sharing) up through executive hire cars. And you'd be able to see how you could get to your destination as cheaply, quickly, or comfortably as possible. Unfortunately, taxi licensing is a scam that these governmental entities don't want to give up control of, so we can't have nice things.

Re:So a guy that runs a ride sharing company.

[DirkDaring]

0 per month, have owned it (Mazda 3) for 4 years. Paid it off in 3. Gas is about $30/week. Insurance is under $50. Maintenance has been inexpensive, only replaced the brakes once and the tires twice. Have never paid for parking. About to hit 100k miles.

Re:So a guy that runs a ride sharing company.

[drinkypoo]

Continue to decrease? It isn't decreasing at all. Car ownership is increasing.

Car ownership is not increasing among the young, which is the segment of the population which is going to have to buy cars in the future if anyone does, because the old people who buy them now will be dead and buried as is the natural scheme of things. Millenials care less about automobiles than has any prior generation since their introduction. And the age of the American fleet continues to increase; sure, more cars are being sold, because the population is growing. But the automakers are still projecting continued decreases in sales year-on-year, for just these reasons which I've been discussing. I find this out by watching shows with their execs and analysts, and listening to them talk about it.

Re:So a guy that runs a ride sharing company.

[drinkypoo]

You sort of blew your point when you mentioned you bought an Audi used. Never do that unless you want to pay more to fix it than it does to actually buy one new. That's less about used, and more about Audi.

You sort of blew your point when you suggested it would cost more to fix it than to buy one... because these A8s are seventy thousand dollar cars. If I spend $2500 fixing up a $200 Audi (which is sort of what I'm doing to wind up doing, because the transmission, front suspension links, and other parts from my $2000 Audi are going to go on this $200 car, which also needs new front and rear glass) then I'm going to get more than twice as much car as buying a $2500 Honda. I've got two engines to choose from now, and even the one with 225k on it has good compression on all cylinders because it's not some piece of garbage to begin with.

I'm not entirely against Japanese cars, although I don't understand why anyone would buy anything but a Nissan or a Subaru unless they actually hated cars. For what I have into my brace of A8s, you couldn't get a Honda worth owning. The only car Honda has ever made that comes close to the A8 in terms of quality and design merit is the NSX (which admittedly exceeds it in most ways) and that costs a mint, plus I wouldn't fit in one.

I'll also go ahead and say that I wouldn't bother restoring a steel Audi. So there's merit to your argument, but it doesn't tell the entire story.

Re:So a guy that runs a ride sharing company.

[drinkypoo]

Cost competitive with public buses? This weekend I dropped my truck off at a stereo shop in the south end of town (Austin, TX). Paid $2.50 for an all-day bus pass and came home. Did chores all day and rode the bus back down to the stereo shop to pick up my truck. That would've cost me over $20 with a ride hailing company.

With a current ride-hailing company using human drivers, yes it would.

Looking around at the other passengers on the bus with me, I got a sense many weren't really positioned financially for becoming ride hailing customers. At least one was in a wheelchair that required a lift that wouldn't be equipped on many privately-owned vehicles. Advocating for ride hailing to replace mass transit does not consider the needs of those for whom the bus is the only transportation 'option'.

It's not really wise to replace it, but it will be useful to augment it. Personally, I don't care who owns the vehicles, as the GP might; I'm perfectly happy if the state owns them and operates them at cost as part of the public transportation network. In fact, I might even be happier with that than the private sector, if not for corruption. I don't mind if the state is in the business so long as they're not passing anticompetitive laws to protect their monopoly. If someone else can come up with a way to provide service safely while lowering costs, they can still compete with a government-operated at-cost transportation system. It's monopolies that grate.

The flip side is that if you provided cheap car-based public transport, some of those disabled people would be able to run a lot more errands in one day. The vehicle could be wheelchair accessible if it were more or less van-shaped.

Additionally, if a ride hailing car travelled to my house (empty) took me to the stereo shop, then returned to the driver's house (empty), that would add two unoccupied trips on the public street that wouldn't have occurred had I ridden the bus or driven myself. If you multiply those empty trips times the number of people on the bus, suddenly there would be a lot more cars on the street not carrying passengers and slowing everyone else down with greater congestion.

The time when car hailing as a substitute for vehicle ownership becomes a problem is at commute time. That's the time when it suffers the same weakness as private vehicle ownership; all those people need to go to the same place at the same time. Of course, some actual ride sharing added in to the mix would perhaps finally implement mass carpooling; you'd specify your preferences for ridemates and the system would join you up with [say] 2-4 other people (by preference) who would all split the fare (plus pay a per-passenger handling fee, no doubt, hopefully a small one) and get dropped off at their respective offices, all of which would of course be either near one another or at least lying upon the same route. Indeed, the cars could pick people up and drop them off along the route, where applicable. This provides the same sort of benefit as public transport (vehicle sharing) without all the obvious downsides which I shall not enumerate here.

Re:So a guy that runs a ride sharing company.

[PIBM]

To kill ownership would mean that 100% of owners got rid of their cars. It only takes one to prevent that. While you think that there aren't many families in the city, you are very wrong. A lot of my coworkers had 2, 3 and even some others had 4 kids too. I was living near the bay bridge (Baele st) in SF, was going to work with my bicycle. My wife and kids would go to the beach (Pacifica was one of their favorite destination), to happy hollow park in San Jose and multiple others, to the Zoo, aquarium, in some nice montains nearby (Mount diablo trails were a win) or on the North side of the golden gate, then once I would know at which time I'd be done working we'd set ourself a rendez-vous point and I'd cycle there when she was driving back. Spending the evening in some nice place, getting the bike back on the rack, and driving home.

On the weekends we would just use the car and drive away to Santa Cruz, or Monterey and the nice aquarium, or Big Sur for some camping, or Lake Tahoe for a long weekend, or Point Reyes, or spend some time in Napa / Sonoma. And much more! I don't think anyone would say that SF is not a major US city, and this lifestyle of mine was also relatively common with my coworkers. I discovered many nice places with them & other friends on the weekends, and even if I'm no longer living there, they do and keep sharing their experiences.

So, I am sorry that you can't envision this lifestyle, but I quite enjoy it.

Re:So a guy that runs a ride sharing company.

[PIBM]

Edge case for slashdot, I guess you are right. So many other nice examples keeps coming up, and a lot are also among my friends & family (ie, not just me ;) ). A single outlier is enough to prevent the envisionned death. I have a hard time trying to evaluate the current car ownership rate for people who really need it. Perhaps 5% ?

Re: So a guy that runs a ride sharing company.

[Extreme_biker0]

Yeah it's also true that people here in the UK seem to be buying cheap cars too because we can't afford good ones generally. You see tons (literally) of cheap dacia and Kia on the road, with cheap as can be dirty diesel engines. If ride sharing in autonomous cars becomes popular it will be because the cost is equivalent to owning only a small portion of a car. This smaller burden would allow regulators to tighten emissions targets while still allowing people to get around.

No, they won't.

[Anonymous Coward]

Anyone who believes this is a fool. So-called 'self driving cars' are NOT going to be capable or safe to be unattended anytime in the next 5 years, or 10 years, or 20 years, because we do not have true AI and will not have true AI until we understand how our own brain works -- and we're not anywhere near to understanding that yet. Until then all we have are these cheesy 'learning algorithms' that are pale in comparison to an actual living brain and just plain can't be safe enough to have human lives depending on them. All this 'autonomous car' stuff is just hype and nonsense and requires at least one human being hovering over it because they KNOW it can't be trusted, and this crap is all being rushed to market as fast as they can so the investors and stockholders don't revolt and start demanding high-ranking people be fired over it. Meanwhile if it's allowed on streets unattended people will DIE and there will be no justice for the survivors of the victims. It's an un-ready, un-safe technlogy that will not deliver what it promises, will actually make the roads LESS safe, and it should NOT be allowed.

Re:No, they won't.

[kheldan]

Like too many people you can't seem to think your way through the problem, and furthermore you can't recognize the FACT that you're taking something for granted that you've been doing for years and years. Driving is NOT a trivial task, and it is NOT acceptable for some piece of software to come to the end of a decision-tree and just pull over and come to a stop, because it can't cope with it's input, and that is what will happen -- until they can emulate the human mind, and create something that is REAL AI and not the ersatz they keep trotting out to impress the investors, so they can keep their jobs. You don't understand the complexity of the issue and I do, and that's why you and people like you keep insisting that it's 'here, now' or 'only 5 years away' when that's clearly and simply NOT TRUE.

Re:No, they won't.

[DirkDaring]

Well I do agree with the 20 years part, but not so much with the AI part. Google has a VERY impressive system, I highly recommend watching this video (very cool stuff too): https://www.youtube.com/watch?v=Uj-rK8V-rik

But I still think it's a long way off even with how good it is.

Re:No, they won't.

[fluffernutter]

They need to be able to determine the difference between a rabbit hopping across the street and a bouncing ball.. because a child may be running after the ball. A human would stop and so should AI.

Re:No, they won't.

[drinkypoo]

They need to be able to determine the difference between a rabbit hopping across the street and a bouncing ball.. because a child may be running after the ball. A human would stop and so should AI.

They don't need to do that because they can just slow down anyplace the view is obscured, so that if someone comes out from between parked cars they don't hit them. They can also simply assume that any object crossing the road rapidly might be followed by another; that's true whether we're discussing a child following a ball or one in which there's two rabbits... or more likely, multiple deer. A small fawn is about the same size as a large ball...

Re:No, they won't.

[HornWumpus]

Any object? A blowing leaf? A plastic bag?

Kids will throw balls into highway traffic just to watch the fun.

Re:No, they won't.

[drinkypoo]

Kids will throw balls into highway traffic just to watch the fun.

There is already a mechanism in place to handle that problem, which will no doubt be aided by the addition of a bunch of high-resolution imagery of the event; even moreso if the number of vehicles capturing such data is increased. Taxi V2V could well include a request to other vehicles to help capture photographic evidence.

Re:No, they won't.

[BlueMonk]

I suspect that, as soon as the accident rate for automated drivers is lower than that of human drivers, insurance companies will not only be open to insuring automated vehicles, but promoting it. Furthermore, I think we've already passed that point.

Re:No, they won't.

[AvitarX]

Kids already do this, and throw rocks from overpasses.

Fortunately, most humans, even children, aren't horrible people, and it's rare.

Do you really think a kickball thrown onto a busy highway won't cause an accident already?

Re:No, they won't.

[Gornkleschnitzer]

"Anything that is supposed to do the job of driving a car" is a set that includes humans. Do you therefore believe that any human (of which there are thousands per year) who makes a mistake resulting in a traffic fatality should be destroyed?

I sense a problem with this viewpoint.

Re:No, they won't.

[HornWumpus]

No, it likely won't. How terrible are the drivers where you live?

It won't necessarily cause an accident with automated cars either, just a huge traffic jam.

Automatic elevators were first

[mi]

How sad to see the nice, well-groomed and jovial men operating elevators replaced with the soulless automation.

We are going to miss the nice, well-groomed and jovial cab-drivers too...

Re:Automatic elevators were first

[kheldan]

Oh look, it's this tired-out old argument, which identifies the poster as someone who apparently can't think critically or logically. Do I really have to go through this again, just to show how dumb you're being? An elevator goes up and down in a finite space and that's ALL it does, it's not in any way shape or form comparable to an automobile. Neither is a train, and neither is a plane, and neither is a boat or ship, so stop trying to compare anything to an automobile.

Re:Automatic elevators were first

[OzPeter]

An elevator goes up and down in a finite space and that's ALL it does,

Not any more they don't: Next-gen elevator goes sideways as well as up and down

Re:Automatic elevators were first

[mi]

An elevator goes up and down in a finite space and that's ALL it does

Which is why were able to automate them first — decades before coming for automobile drivers.

Some trains were also automated decades ago, though wider adoption is still met with fierce opposition from organized labor and their idiotic sympathizers. Even replacing the "conductors" with video-cameras was deemed to violate labor-agreements in NYC, getting rid of the nice, well-groomed and jovial motormen would've been a non-starter.

it's not in any way shape or form comparable to an automobile

Automobile is comparable in that it is increasingly possible to automate its functioning.

Trains and ships could've been automated even before automobiles, but the cost of crew in those is a relatively minor share of the overall cost of operation, so there was not as much incentive to fight Labor Unions on that. Humans are still involved, though, perhaps, not for much longer.

Re:Automatic elevators were first

[lgw]

stop trying to compare anything to an automobile.

Wait, so you're demanding Slashdot without car metaphors? Get out! We don't serve your kind here.

Welcome to Johnny Cab!

[the_skywise]

https://www.youtube.com/watch?...

Picard meme "Not this shit again"

[OzPeter]

The third phase, which according to the graph the company expects to be sometime in 2021 or 2022, will be when all Lyft rides will be completed by a fully autonomous car. Shortly after that phase begins, car ownership will see a steep drop-off, according to Zimmer. Zimmer, who has long been a vocal proponent of ending car ownership, set a date for the death of the personally owned car in major U.S. cities: 2025.

So he wants to totally rewrite the entire legal basis for cars, road rules and insurance in the US within the next 9 years? And expect that in a country where people almost idolize their cars that they are suddenly going to say "hell yes .. I'm selling my car tomorrow!"

Where can I get some of what he's smoking?

Alternatively its purely a BS reaction to Uber to try and remind people that Lyft exists and is relevant.

Re:Picard meme "Not this shit again"

[Br00se]

Hmm, no market for used cars since everyone is just giving them up. No market for new cars either, except the self driving ones that few individuals will own.

I don't see automakers adopting this business model any time soon. Same for the finance industry that exists to loan people money for cars.

I'd be careful to avoid what he is smoking, you don't want your mind altered that much.

Re:Picard meme "Not this shit again"

[drinkypoo]

Let me handle these in reverse order.

Alternatively its purely a BS reaction to Uber to try and remind people that Lyft exists and is relevant.

This is almost certainly the truth.

Where can I get some of what he's smoking?

Probably pretty much anywhere.

So he wants to totally rewrite the entire legal basis for cars, road rules and insurance in the US within the next 9 years?

I don't think it's going to change changes that massive.

And expect that in a country where people almost idolize their cars that they are suddenly going to say "hell yes .. I'm selling my car tomorrow!"

The US fleet continues to age to record levels, year on year. People are buying less cars. Unfortunately, sales of hybrids are flat and sales of EVs are way down. These vehicles have a higher energy cost of production, so it's beneficial for the environment when people keep them longer. But when people keep ordinary gasoline-powered vehicles longer, they just produce worse emissions...

Re:Picard meme "Not this shit again"

[bluefoxlucid]

The major defect in his argument is generational: we're all used to driving, to owning, and otherwise to just having our own private car. There's a familiarity in this which is in many ways romantic, filling us with a sense of security as we remain surrounded in our own territory and firmly in control during every commute.

A new generation will have to introduce itself to that familiarity, and to the classical concept of car ownership and of driving. Those of this new generation will spend their lives observing an exciting transition to self-driving cars and chartered transportation without the restrictions of buses and trains. Their decision won't be between a new form of transportation and an old; it will be between two available options, neither of which is a staple of their lives and identities.

That generation will likely abandon the costs, risks, and inconvenience of driving in favor of chartering self-driving cars. Why spend $28,000 every 5 years, plus $3,000 in fuel, plus $7,500 in maintenance, plus $5,000 in insurance? $725/month, or you can spend $10/day and end up with $300/month. The difference comes from the vehicles turning over a large profit per their usual lifetime, being better-utilized between maintenance cycles, having more-efficient insurance (self-driving, large fleet, self-insured), better maintenance (access to mechanics as a factored-in cost), and better lifetimes (due to better maintenance), as well as a tax deduction for their depreciation (eventually cuts 40% of the cost from the purchase price) and all other costs (that base $725/month becomes $435/month to start with).

There's an interesting economic advantage here, too: those enormous tax deductions ($3,480 per car per year) are eliminated from consumer expense. If the taxi service has a higher profit margin, then part of that is captured in taxes (i.e. if they save $3,480 and charge you $3,000 less, that extra $480 is taxed at business 40% rate). Whatever savings go to the consumer are spent on other goods, which eventually get taxed--less taxes on what goes to wages of those other businesses supplying other goods, and the same taxes on those business's profit margins. The enormous tax savings from deductions and depreciation is just captured somewhere else, creating more jobs and more consumer wealth in the process.

Of course, those newly-created jobs aren't an expansion of employment; they just replace the taxi driver and maintenance worker jobs lost in this transition. They also take time to materialize, meaning sudden economic shock (changing to driverless taxis overnight) creates a spike in unemployment, while slowly converting the market (changing to driverless taxis over half a generation) creates visible strain on the taxi driver market (unemployment in that sector) while some other market grows to compensate (e.g. IT jobs, doctors), with a much smaller peak unemployment increase.

Re:Picard meme "Not this shit again"

[bluefoxlucid]

Automakers and finance can campaign against it, to try to control the public mind and maybe get laws passed to impede the change (rent-seeking behavior). They can't stop it forever. The public mind won't hold up as you change over through a new generation; and, two generations later, congress starts filling with people who won't buy your bullshit. Somewhere along the way, the public mind starts pressing old politicians, so you don't necessarily need to wait 40 years for everything to kick in; the media can also start talking about exciting new opportunities and cause existing public sentiment to shift, enough to create political favor for whomever works to deliver on these ideals, and damn what the banks think about it.

Re:Picard meme "Not this shit again"

[JaredOfEuropa]

A lot of all that can be worked out as we go along. Rules for motor cars were not all hammered out before the first one was allowed on the roads.

Legal issues: this is mostly about liability. Has to be worked out but this does not have to take absurdly long.
Road rules: no need to rewrite those, autopilots are built to work within the existing rules. At first there may be some additional rules (and traffic signs) about where you may and may not engage the autopilot. Again, very little keeping us from allowing autonomous vehicles on the road today.
Insurance: this is potentially harder: autonomous cars will likely require special insurance, and at first the risk of this insurance will be very hard to gauge. But insurers have deep pockets and may be willing to take on that risk in order to be the first with "autopilot insurance". If that insurance turns out to be too expensive for consumers, auto manufacturers or even governments may decide to sponsor this type of insurance or assume part of the risk in order to speed adoption. A temporary arrangement, as insurance premiums for autonomous vehicles are likely to drop below levels of insurance for regular cars.

With that said, you are right about one thing: there is no way in hell privately owned cars will disappear from major cities anywhere near 2025. There are still loads of technical problems to overcome, and a lot of them are solved in an iterative fashion, by building on real life experience with the previous release. Nothing that gets solved overnight. And even if we see car hire services for fully autonomous cars (and I mean real cars, not golf carts limited to 25 MPH) emerge by 2020, which I doubt, there'll still be many people who are not going to give up the convenience (and status) of having their own car. The first thing we may see disappearing by 2025, if technical development milestones are achieved as predicted here, is ownership of second cars for occasional use.

Re:Picard meme "Not this shit again"

[HornWumpus]

It's been 20 years since OBD2. Any reasonably new and maintained car puts out about the same emissions.

Classics are driven so few miles they aren't worth counting.

Average age going up 2 years is just a rational (albeit slow) response to improved metallurgy. Mostly hard chromed piston rings.

Re:Picard meme "Not this shit again"

[drinkypoo]

The major defect in his argument is generational: we're all used to driving, to owning, and otherwise to just having our own private car.

Perhaps that explains my perspective, since I grew up raised by a single parent who never owned a car. I've been driving since I was sixteen, though.

That generation will likely abandon the costs, risks, and inconvenience of driving in favor of chartering self-driving cars.

I agree. However...

Why spend $28,000 every 5 years

Every 11 years.

Re:Picard meme "Not this shit again"

[bluefoxlucid]

The used car market is fed by people who bought new cars and are buying a new car. Those 11-year-old trucks were once new trucks, then 5-year-old trucks that got sold, then bought for cheap by people who might keep them 3, 5, or 8 years.

People aren't buying brand new cars and keeping them on the road for an average of 11 years; and the people buying used cars *definitely* aren't averaging 11 years holding them, because otherwise the average age of the car would be higher.

Re:Picard meme "Not this shit again"

[drinkypoo]

It's been 20 years since OBD2. Any reasonably new and maintained car puts out about the same emissions.

O2 sensors and catalysts both wear out. If a vehicle is driven regularly you should replace the O2 sensors at about five years. If you don't do that, you substantially increase the risk of catalyst damage, and then your emissions most definitely increase. People don't replace things like that until lights go on.

Average age going up 2 years is just a rational (albeit slow) response to improved metallurgy. Mostly hard chromed piston rings.

The rest of the vehicle still degrades at about the same rate as ever, especially here in America where most interior is still hot garbage. People want to replace these vehicles sooner, they just can't afford to.

Re: Picard meme "Not this shit again"

[bluefoxlucid]

Looks like it's 35% now. I thought it'd been raised to 40% recently.

Re:Picard meme "Not this shit again"

[Hadlock]

GM invested $2 billion in Lyft; car ownership is down, drivers licences are down, not as a percentage but as a whole. The CEO of Ford said as much last week in an interview with the WSJ. Wages are down and savings are down. People can't afford to drive and maintain cars, and people are moving towards cities. Ford, Volvo and others are already designing and marketing their next generation of driverless cars and vans. Finally, cities are full. There's no more room to build freeways in urban cores, and even if there were, where would you park 12 million cars in downtown manhattan? San Francisco? Chicago? Because those are the population centers and they already have low car ownership rates, and that is where people are moving to. Vacation and buisness hotspots rely heavily on Uber, which is going to put pressure on Austin to open back up the market to ride sharing again.

Re:Picard meme "Not this shit again"

[HornWumpus]

You don't need to do preventive maintenance on O2 sensors or cats. When the check engine light goes on, you replace the broken part. OBD2 is a rolling smog check, if the light is off, the part is doing it's job. (Don't even suggest a preventative O2 sensor change to V8 Volvo drivers...those pieces of shit have 16 O2 sensors and 8 cats, all overpriced.)

New car people do replace their cars before the first bit of plastic breaks, but that's not what we are talking about. Average age is fleet average, not age at first resale.

With the onset of throttle by wire, old cars are looking better and better. The last rental car I had overruled my throttle stomp as it thought the steering was too far turned for any power.

Re:Picard meme "Not this shit again"

[drinkypoo]

You don't need to do preventive maintenance on O2 sensors or cats. When the check engine light goes on, you replace the broken part. OBD2 is a rolling smog check,

...which lights the MIL when the sensors detect that the vehicle is producing 2.5 times the federally allowable standard.

(Don't even suggest a preventative O2 sensor change to V8 Volvo drivers...those pieces of shit have 16 O2 sensors and 8 cats, all overpriced.)

My A8s have four sensors and two cats, but the downstream sensors only monitor the cats so you can safely ignore them until you get a MIL (as you say) so long as you don't ignore the upstream ones on the down pipes.

New car people do replace their cars before the first bit of plastic breaks, but that's not what we are talking about. Average age is fleet average, not age at first resale.

You're right, but most people don't want a beat down used car, either. They want it to still have all the buttons and knobs and for everything to work.

With the onset of throttle by wire, old cars are looking better and better. The last rental car I had overruled my throttle stomp as it thought the steering was too far turned for any power.

If I can just figure out how to disable traction control on my A8s (easy on FWD models, hard on AWD ones) then I'll be in there. I've got two now and I think I'm going to end up swapping my working transmission into what was supposed to be a parts car... anyway, mine are too early to have ESP, so they don't think for you much. The throttle is cabled. There is a slush box, but it's got tiptronic and I've re-coded it into euro mode which is much nicer than USA mode in auto mode, and also disables the learning bullcrap so it always behaves the same way.

Re:Picard meme "Not this shit again"

[HornWumpus]

O2 sensors get old and respond slowly (engines are usually bouncing between slightly rich and slightly lean). The ECU sees that and throws a code. The back O2 sensor also throws a code when the cat is no longer working (as you say, it runs dirty).

Re:Picard meme "Not this shit again"

[lgw]

People are buying less cars.

fewer cars

Re:Picard meme "Not this shit again"

[drinkypoo]

O2 sensors get old and respond slowly (engines are usually bouncing between slightly rich and slightly lean).

Yes, but they also read incorrectly when they age. Reading cross counts will tell you when the O2 sensor is old, but it won't tell you by how much it is incorrect. The comprehensive monitor only trips (and thus illuminates the MIL) when the cross counts are VERY low, or when the O2 sensor sends an implausible signal. The O2 sensor is sending an incorrect signal for years before complete failure in most cases.

The back O2 sensor also throws a code when the cat is no longer working (as you say, it runs dirty).

I'm only going to repeat this once: The code is only thrown when emissions of 2.5 times the FTP are detected. That means that your car can fail a smog test badly and still not fail the emissions monitor, and thus light the MIL.

It's time for you to read up on the OBD-II monitors.

Laws and other stuff will take longer then 5 years

[Joe_Dragon]

Laws and other stuff will take longer then 5 years
http://worksnewage.blogspot.co...

When was the last time you were in a cab?

[mykepredko]

I don't think I've ever had a cab driver that could be described as "nice, well-groomed and jovial".

Maybe Angry, slovenly (with the overwhelming scent of B.O.) and Anti-social would be a better description of the cab drivers I've had throughout North America.

Personally, I welcome our Johnny Cab overlords: https://www.youtube.com/watch?...

Re:When was the last time you were in a cab?

[nicolaiplum]

In the last 2 weeks I have taken licensed taxis in the Netherlands, UK, and USA. In all cases the drivers were reasonably competent, at least as well groomed as me, and either quiet and polite or jovial and talkative.

In the USA, the main problem with taxis is in the places with a limited number of licenses for taxis, possessed by rentiers who have no incentive to improve service and whose employee drivers have no autonomy and earn very little. Un-limit the taxi number and give them autonomy and quality would greatly improve.

us secret service with not use auto drive car any

[Joe_Dragon]

us secret service with not use auto drive car any time soon and if they do that may need a rule over ride mode that may need to go as far as running people over to get away.

technical judgment

[ooloorie]

Well, with this kind of stellar background, how could anybody doubt his judgment on the capabilities of technology, or even his honesty when making pronouncements about what his company will or will not accomplish?

Zimmer grew up in Greenwich, Connecticut. In 2006, Zimmer graduated first in his class at Cornell University School of Hotel Administration and was a member of the Quill and Dagger society, as well as a member of Sigma Pi Fraternity.[1] After graduation, Zimmer worked as an analyst in real estate finance at Lehman Brothers in New York City. While at Lehman Brothers, Zimmer spent most of his time creating real estate models in Microsoft Excel. Zimmer left Lehman Brothers three months before it declared bankruptcy. In 2007, alongside Logan Green, Zimmer founded Zimride, a ridesharing platform that coordinated carpools, especially across college campuses.

Re:technical judgment

[RespekMyAthorati]

Good research!
I'd mod you up if I knew how.

HOW are they NOT a TAXI service at this point?

[tekrat]

Lyft and Uber can no longer maintain that they are a "ride sharing service" where independent contractors own and operate their own vehicles.

In the autonomous cars case, Lyft and Uber CLEARLY own the vehicles!!!! They are a TAXI service and need to be regulated as such.

How the hell have state and local prosecutors allowed them to get away with this??

Re:HOW are they NOT a TAXI service at this point?

[nicolaiplum]

The way Uber gets away with this is by being faster and more widespread than regulators. This doesn't happen in countries where the regulator covers the whole economy, like in most European countries. US-wide companies against individual State regulators means the regulators are rarely effective and can rarely keep up.

Yet Americans widely oppose "The Feds" regulating anything - so they get the companies they want, because they oppose regulation of them. They get the country they want.

Re:HOW are they NOT a TAXI service at this point?

[drinkypoo]

In the autonomous cars case, Lyft and Uber CLEARLY own the vehicles!!!! They are a TAXI service and need to be regulated as such.

Obviously the regulation will have to be different, because there are no drivers.

How the hell have state and local prosecutors allowed them to get away with this??

It hasn't happened yet. Calm down.

Re:HOW are they NOT a TAXI service at this point?

[lgw]

hey are a TAXI service and need to be regulated as such.

No, they are a hired car service: the broader category that includes taxis, limo services, and so on. As long as you can't hail them from the street, they aren't taxis.

Re:HOW are they NOT a TAXI service at this point?

[Gussington]

In the autonomous cars case, Lyft and Uber CLEARLY own the vehicles!!!! They are a TAXI service and need to be regulated as such.

They aren't cars, they are sideways elevators. And last time I checked we don't apply taxi regulations to elevators.

How the hell have state and local prosecutors allowed them to get away with this??

Get way with what, it hasn't happened yet. It is only a clickbait headline, relax...

Try 20 years and no it won't.

[DirkDaring]

They don't even have have the laws changed yet so driverless cars can even operate. Good luck on that one within 5 years. And once it happens and these things start having accidents (which will happen, they will be on the road with human drivers who will always blame the machine which we see already in Tesla accidents) and lawyers get involved it will stretch the time-frame out even further. But on to when they eventually do...

So everyone is going to hail a 'cab' whenever they want to go somewhere? Is that really their plan? Like a trip to Starbucks? To the grocery store? To a friends house? To soccer/band/swim/etc practice? To a school football game? If to a big event and 300 people need a ride home how many hours of a wait will it be? Do they plan on having a massive amount of cars sit idle for big weekend events? And how much will all these trips be? Sounds way more expensive than owning a car and about a thousand times more annoying. Try going to Disney World during the summer and wait for a bus back to your resort after Magic Kingdom closes. They have the best transportation system I've ever seen and you can still have an hour wait.

But he did get some free advertisement. OMG driverless carz!

Re:Try 20 years and no it won't.

[mbone]

They'll just 3-D print these driverless cars when there is demand.

Get out of your city more often

[Etcetera]

The tech industry would be much less hilarious if so-called visionaries left the Bay Area or the East Coast Boston/Phil/NY/DC metropolitan areas more often.

If you get out into the other 98% of America, you'd realize that a) we like cars and b) we like ownership of property and things like cars.

Stop trying to solve problems that only exist in San Francisco. Thanks.

Re:Get out of your city more often

[JustNiz]

This. If only I had mod points you'd get them all.

Re:Get out of your city more often

[psmoot]

Stop trying to solve problems that only exist in San Francisco. Thanks.

(Full disclosure: I've lived in the Bay Area for 25+ years and work in the tech industry.)

I've never really wanted a car when I lived in Boston. I don't want one when I visit NYC (especially Manhattan). Didn't particularly want one on London, Paris, or Rome either. I can't imagine I'd want one in Beijing or Tokyo. I'm seeing a trend here.

I find as I grow older (I'm in my 50s), the romance of owning a car is really diminishing. I no longer identify as powerful and free because I have a car. My kids are the same way, they kinda don't care whether they own a car or even have a driver's license. Maybe the Bay Area is weird but I think this is far from unique.

Re:Get out of your city more often

[Tom]

Here in Europe, private ownership of cars could go the way of the Dodo bird and many people would welcome it.

The main differences:

One - our cities are older and streets smaller, the insanity that is hundreds of thousands of people each driving in a huge metal box that is mostly empty becomes visible very fast under such conditions. Parking in most European cities is a nightmare.

Two - we actually have working public transport.

I would be more than happy to use self-driving taxis in the city, and keep my car only for long-distance trips where train is not a good choice for some reason (remote village, castle, animal park, whatever in the countryside, etc.)

Maybe americans love their cars so much - but half of them also liked slavery and anyway it was abolished.

Re:Get out of your city more often

[Drunkulus]

Mod this up, I would if I could.

Re:Get out of your city more often

[khoonirobo]

I do sometimes find it cute how parochial the crowd on slashdot gets. I understand your point but do you realise how small a component of the market (defined in 10 - 50 years) of transportation you are? So, you stay outside a major population center in the US, population under a million?, now let's go and look at :

https://en.wikipedia.org/wiki/...
or
https://en.wikipedia.org/wiki/...

and see the areas where people would prefer good public transport infrastructure rather than private ownership of cars. These would be densely populated metropolitan areas, which represent high economic activity and demand. At a rough glance, that's 500 million to a billion people who depend upon public transport in their daily economic activity (even if they own a personal vehicle). That's the market that exists *today* which is only going to increase in the future.

Outside the US, Canada, Europe and Australia, personal vehicles are not common in non metropolitan areas. I have no figures but I think that'd be another 4 - 5 billion people. I'm pretty sure they'd welcome such services if the price drops enough and the technological problems are solved (this will take much longer but will be solved eventually).

This is the market at stake being addressed, it's by any accounts sizeable enough. The problems exist.

Re:Get out of your city more often

[lgw]

would be more than happy to use self-driving taxis in the city, and keep my car only for long-distance trips where train is not a good choice for some reason (remote village, castle, animal park, whatever in the countryside, etc.)

So, what you're saying is, you're not actually planning to stop owning a car?

I suspect you'll find that rural Europe isn't going to stop owning private vehicles any sooner than rural America.

Re:Get out of your city more often

[bluegutang]

Self driving cars will allow travels at 4x current speeds. You may love driving and owning cars but going 200 miles per hour trumps all that.

Not gonna happen. Roads are built with a certain turning radius. Go much faster than that, and the ride becomes uncomfortable like a roller coaster, and the car's tires will not be able to effectively grip the road. So speeds will go up marginally rather than by a factor of 4.

Re:Get out of your city more often

[ThatsMyNick]

Metropolitan/urban areas house about 80% of america's population. So even if only metros adopt, I am sure he will be extremely happy. Whats wrong with solving problems that only affect 80% of Americans? Why do you want him to stop?

The other 20% may like car ownership, he doesnt care of course. I know I wouldnt.

Re:Get out of your city more often

[JustNiz]

> Here in Europe, private ownership of cars could go the way of the Dodo bird and many people would welcome it.

Bullcrap. As an someone born and bred in rural England I can tell you that the half hour ride on the one bus a day between the village and the nearest town/shopping isn't ever going to replace the privately owned vehicle, nor would people thank you for taking their cars from them.

Re:Get out of your city more often

[Gussington]

If you get out into the other 98% of America...

Stop trying to solve problems that only exist in San Francisco. Thanks.

Did you ever think that maybe there are places that exist outside of SF that aren't part of America? I know you're probably American so your Geography education at an infant-like level, but for the 99% of the car using universe that isn't the US, this solves a lot of issues.

Re:Get out of your city more often

[Gussington]

Metropolitan/urban areas house about 80% of america's population. So even if only metros adopt, I am sure he will be extremely happy. Whats wrong with solving problems that only affect 80% of Americans? Why do you want him to stop?

The other 20% may like car ownership, he doesnt care of course. I know I wouldnt.

And the elephant in the room which the Americans in here seem blissfully unaware, is that there are people outside of America too. And they like to move around. So for the other 97% of the planet, it's a winner.

Well that's wrong

[SuperKendall]

many things are a 'judgement call' and that requires a conscious, sentient, self-aware brain to decide what to do.

That's not at all true. Anything directing the car, no matter how dumb, can make a choice.

In fact on average a computer will have much better visibility to the full surroundings than a human driver could, so even if "dumber" the computer is better informed and so statistically will probably make better choices.

I also disagree with your assumption that all drivers can be classed as "conscious" when there are a lot of bad drivers who are zoned out while driving.

Re:Well that's wrong

[fluffernutter]

Can AI differ a hopping rabbit and a ball bouncing in the street and therefore watch for a child? Oh you programmed for that? What about a cat and a ball? Bird and a frisbee? Shopping bag and a frisbee? Ball and a tumbleweed? It needs to be smart enough to accurately categorize between *something* that *no one* will be running after and *something* that *someone* may be running after, which means it needs to know what everything is smaller than a breadbox. Unless it can sense off road movement through infrared or something, but I don't know if any AI cars have that yet. Even then it will see a shape and need to determine what it is and what it is likely to do. That's just a small example. No one can think of every possibility and program every shape in the world, much less an understanding of it; so understanding will need to be inherent. Not just decision making, but understanding.

Re:Well that's wrong

[ranton]

Our computers fuck up all the time because of bad sofrware. So do your phones. So will so-called 'autonomous cars', and the more complex the software, the harder it is to find the bugs in it, and in this case the bugs WILL GET PEOPLE KILLED

Bugs in the human brain killed 38,300 people in 2015 and injured another 4.4 million. And these bugs are far harder to find and fix than autonomous driving cars would be. Yes, it's very likely each year self-driving cars will kill thousands of people and injure hundreds of thousands more. But even if this is true they would still be an order of magnitude safer than human drivers.

I'm not saying it is an absolute that self-driving cars will be safer than humans in the next 10 years, but it is certainly not nonsense.

Re:Well that's wrong

[Gilgaron]

You're ascribing human limits (finite field of view, need to shift attention) to the machine that may well have a 360 degree field of view and faster reaction time.

Re:Well that's wrong

[fluffernutter]

How will that field of view work for it if the ball is visible but the child is in the process of running out from between two parked cars?

Re:Well that's wrong

[Anonymous Coward]

You can't program every possible scenario, but you can program enough scenarios to have less a lot less deaths than human drivers, which is the whole point.

Re: Well that's wrong

[fluffernutter]

So tell me the number of sufficient cases you need to determine if an object is any kind of child's toy or not any kind of childs toy? If you don't know the amount of trials you need to do to cover any object in the world, then how do you know if you're going to be killing a child or not the first time it happens?

Re:Well that's wrong

[fluffernutter]

Another thing, the car's reaction time doesn't mean anything if the road is icy or slippery.. So will an automated car drive 10 mph everywhere if the road is slippery and a child may run out into the road and block traffic behind it? Or will it actually be able to understand the object and anticipate the child to give more stopping time? These are very common situations in the real world; I'm sure it happens somewhere a hundred times a day.

Re:Well that's wrong

[parkinglot777]

In fact on average a computer will have much better visibility to the full surroundings than a human driver could, ...

I disagree with this part. Why? It has nothing to do with a computer, but it is depended on Sensors & how the information is converted in a form that a computer can understand. Computer vision is NOT the same as human vision by the way. And you over value on "on average" cases. Who cares for average cases because there is no significant damages and/or life hazard issues (trivia). The focus is all about extreme cases which lead to dangers and damages. So I'm not sure how you can compare a computer with human in this case..

Re: Well that's wrong

[zlives]

third world market beta testing

Re:Well that's wrong

[drinkypoo]

How will that field of view work for it if the ball is visible but the child is in the process of running out from between two parked cars?

We've covered this argument extensively. The answer is the same to the question "will my car run over a pair of nuns to save my life", and it's that the car will slow down and not get into the situation in the first place. If the sensors can't see what's by the side of the roadway, the vehicle is going to detect that and slow down. The vehicles will learn routes which tend to have less obstructions, so that this is minimally inconvenient.

Re:Well that's wrong

[Kjella]

many things are a 'judgement call' and that requires a conscious, sentient, self-aware brain to decide what to do.

That's not at all true. Anything directing the car, no matter how dumb, can make a choice.

There's also a huge gap between the smartest thing to do and the minimum required by law. If I see a crazy drunk driver that's about to ram me in an intersection I might want to slam the brakes, but it's not like I'll get charged for not avoiding it. If I come across a hill and get blinded by the low sun I could brake hard. I might also be rear ended by the next guy coming over the hill, but technically that's his fault. Even if a semi has lost its brakes and is barrelling down the hill towards me I legally don't need to get out of the way. Now I, with my sense of self-preservation would. But if an autonomous car doesn't it won't be the charged with reckless driving or manslaughter.

And that really does solve most of the imaginary "hard" problems, it doesn't have to do better than a human all the time. It just needs to sufficiently obey the rules of the road that it can't be held culpable for causing the accident. Even if that means running over two people who jumped out from cover instead of one on the sidewalk. Obey all signs and limits, stay in your lane, signal as appropriate and brake hard for any obstacle and you'll be hard pressed to be found reckless. Accidents can still happen, but a tire can blow out today just like a sensor can malfunction. And I imagine it'll have more redundant systems than a human...

Re:Well that's wrong

[kheldan]

Not just decision making, but understanding.

Thank you! That's a way of putting things that hadn't occurred to me. There are more dimensions to the issue of operating a motor vehicle on a public road than just sensory_input->decision_output.

Re:Well that's wrong

[fluffernutter]

You haven't answered how this would work in terms of the scenario I presented. Road is slippery, ball bounces out from between two cars, child runs out after it. We can make the assumption there will be no 'street level' scanner that will see the child's feet under the cars so it doesn't see the child. How will it know how to stop? Will it stop and wait 15 seconds for every single thing on the road? If you don't explain the mechanics then no point in arguing that it can happen.

Re:Well that's wrong

[drinkypoo]

You haven't answered how this would work in terms of the scenario I presented.

That's because you presented a stupid scenario.

Road is slippery,

Car detects the slip from its ESP system (mandatory in the US since 2010) and slows down.

ball bounces out from between two cars,

Ball bounces out, car slows down just in case, just as a human driver should. Cars are parked along the roadway such that you can't see what's behind them, car slows down just in case, just as a human driver should. So yes, I very much did answer this question, in two ways, which I will recap for you now. The car will prepare for the situation by being cautious, and it will also respond to the situation by being cautious.

Will it stop and wait 15 seconds for every single thing on the road?

Like I said, it will slow down for every single thing on the road that it cannot positively identify as a non-issue.

Re:Well that's wrong

[fluffernutter]

Is it a stupid scenario? You propose that children never run out from between parked cars? That's funny because I was told not to do it at least fifty times as a child.

The question is, how MUCH does the car slow down as a result of detecting the road is slippery... let's look at it this way, what stopping distance is the car to target? It won't be a stopping distance of two feet, because then it will be driving at walking speed everywhere. Stopping distance for an average vehicle on ice at 40mph is 138 feet. So lets say AI is better than a human and call it 120 feet. A person seeing a ball can easily see it 2 seconds before the kid runs out, and in two seconds at 40mph you travel that stopping distance. So yes it will be very important that the car determines that a child may be running out.

There are always things on the road.. Cars will have to avoid some and slow down for others in order to be workable. You don't want a car to slow down for a shopping bag or a leaf blowing across the road, yet if it is a small gosling which can be the same color and size as a leaf, you do want the car to slow down.

You just aren't opening your mind to all the subtitles in the situations that occur in the real world, if it can physically happen then it will so I am really not grasping at straws here even though it may seem to be a tailored scenerio. A person would be stupid to bet their life that current AI won't cause a death or injury at some point so why bet the lives of others?

Re:Well that's wrong

[drinkypoo]

Is it a stupid scenario? You propose that children never run out from between parked cars?

No, and this is the last time I'm going to repeat myself: the car is going to slow down before it gets to the parked cars because it's going to detect that there is a hazard there, in the form of poor visibility. The average human driver will just blow that hazard off, which is why they're at risk of hitting the child in the first place.

Re:Well that's wrong

[fluffernutter]

Lol.. So you think people will use self driving cars if they slow down every time they pass a parked car? and I repeat... What stopping distance is adequate when passing each and every parked car? Lets say the car picks a 100% safe stopping distance.. two feet. That's 7 miles per hour on ice. So now you are saying that self driving cars will be driving 7 miles per hour past each and every parked vehicle in case a child runs out. That's going to be interesting, especially when the manual cars lining up behind the automated car don't expect it.

Re:Well that's wrong

[tnk1]

Except that's a silly case. A human has slower reaction time and lacks always attentive sensors and is often distracted. The humans can identify the ball, sure, but that's useful because they need the cue.

A self driving car will detect the ball *and* the child and react faster than any human driver to the actual child without having to have a dawning realization about the ball. They don't need to know that it's a child's ball because the car will have detected the actual issue and reacted to it directly.

And yes, in some cases the child will be fast enough or erratic enough to end up in the path of the car close enough that the car can't do anything to avoid it. And that would be the exact same problem for a human driver, only the self-driving car would break fast enough to prevent harm to the child in that case appreciably more often than a human would be able to. That's all down to reaction times and dedicated decision making about actual data. In that case, I'll statistically take the car over any human.

Re:Well that's wrong

[tnk1]

I won't call it stupid, but it's not a very useful example because you've just defined a scenario that a human would do just as bad at, or worse, than the onboard computer.

A human:

Is less likely to reduce speed in environments like that.
Is less likely to even see cues in the first place
Might be on a cell phone or otherwise distracted.

The computer still might fail in that situation, but I don't see why you think an average human would do any better in this convoluted example. An individual human might outperform the average computer, if they were using a considerable amount of intellect to use all cues to their advantage, but even then, humans lack the task oriented senses and reaction times of these systems.

So, even if the human had full situational awareness, were of a professional quality of driver, and took proper precautions, they might still fail in a situation that the computer would succeed at simply due to reaction time and sensor capabilities. And the average human is nowhere near that level of capability and focus on driving.

Re:Well that's wrong

[lgw]

Unlikely. In too many cities, almost all the roads are lined with parked cars.

Re:Well that's wrong

[fluffernutter]

I disagree. A human is going to see a ball in the road and look for a child chasing the ball. It will be instinctual. Some other things a human might notice would be:
- A child in a yard beside the road with a baseball mitt staring at the road (at another child going for a ball)
- A yard with toys strewn about and no child (where is the child?)
- A group of children calling to (a child between the cars)
- Passing a school with children playing

These are the kinds of cues that a human picks up on very easily. I live on a block where there are a lot of kids that play, I always look for their parents or signs that they are playing such as open garage doors before I even get in my vehicle. These are cues that AI will have to pick up on in order to be at least as good as a human.

The thing that I agree with you on is that a human is less likely to slow down, but not for the reason you are suggesting. A human will drive faster in such situations because they can intuit the danger, and thus drive a more reasonable speed. They may drive slower past a row of parked cars near a park or a school, slower still if there are other children playing in that park or school yard. AI will have to have this intuition to be safe. If automation is to coexist with humans it will need to exhibit this same behavior or it will cause nothing but traffic problems.

I'm not saying that AI will never be better than a human in some situations, but that isn't the goal. The goal is for AI not to kill or injure people in every situation possible where a human would not. This means that it must have no weaknesses or scenarios where it does worse than a human, ever, and we are no where near that yet.

Re:Well that's wrong

[fluffernutter]

Please follow along. The situation is that there is a ball in the road giving a clue as to what will happen.

Re:Well that's wrong

[SuperKendall]

You haven't answered how this would work in terms of the scenario I presented. Road is slippery, ball bounces out from between two cars, child runs out after it.

I can say with absolute certainly the car would react to that vastly faster than a human. It wouldn't know if the ball was solid or a child itself; it would take action to halt or swerve the moment the ball appeared, long before the human even considered what the ball meant...

Next.

We can make the assumption there will be no 'street level' scanner that will see the child's feet under the cars

That is a REALLY STUPID assumption. Sensors will usually be at low level, why would they not specifically be looking for human legs? I mean that is like sensor analysis 101 DUH.

Re:Well that's wrong

[fluffernutter]

it would take action to halt or swerve the moment the ball appeared

But what if that is the wrong thing to do? Currently AI cannot tell the difference between a puffed out plastic shopping bag and a ball. You don't want the car to swerve or stop for a shopping bag, since humans wouldn't.

To see to the other side of a parked car, you need to practically lie on the street. In order for the sensor to see along the road under a car, it would have to be placed below the car that is driving. So unless some sensors are going to extend downward once the car start driving, somehow able to survive bumps as the shocks of the vehicle compress, and look directly across, they will not be able to see a child coming through a row of parked cars. Unless of course they have x-ray vision that can see through metal. Any other placement will be looking at the parked car at a downward angle and will therefore be obscured. Also, it might not be a car. It might be a fence, or a community mailbox, or some other fixed object.

Re:Well that's wrong

[SuperKendall]

I disagree with this part. Why?

Because you have no idea what you are talking about?

It's irrefutable a computer has better vision, I'm not just talking cameras but LIDAR (which will be used in combination to see things a human cannot). BTW LIDAR is not RADAR is case you were going to raise some point about THAT....

Computer vision is NOT the same as human vision by the way.

Yes, I work with computer vision stuff in OpenCV, which is why I know it is better in this application (if for NO OTHER REASON than it is not blocked by various pillars, that are absurdly thick in modern cars and greatly limit visibility). I have also been studying sensor tech for self driving cars, have you?

The focus is all about extreme cases

In which the computer has an even greater advantage. The computer will not be speeding because it "looks safe". The computer will actually take conditions into account using data the human does not know (like actually sensing ice before the driver ever could). The computer has vastly faster reflexes than any human, even highly trained drivers - and we all know the vast majority of drivers are not highly trained, not even in Germany...

In any extreme situation the computer has already factored in potential issues long before a human driver would or can.

I'm not saying computers are perfect and no mistakes will be made. But they will make a LOT fewer mistakes than humans will in just a few more years, you have zero idea how many resources are being put towards this now. To say it will never be better is to ignore all of human and computer history.

Re:Well that's wrong

[tnk1]

I don't think anyone believes that a computer driven vehicle will ever be able perfectly avoid killing people. That's not even reasonable, let alone the goal of the program. Computer driven cars will kill people, that's simply going to happen.

The utility of such a system is that there is certainly higher (and improving) safety, but also the potential for better traffic flow, gas mileage, as well as considerable convenience involved.

For commercial applications, it removes the need for a driver which is positive for cost, as well as being able to use the truck more efficiently with more non-stop travel hours for each rig.

I enjoy driving as much as the next person, but not having to do it even when I don't feel like it, like in bumper to bumper traffic, would be a huge relief. While I have a good record of not getting in accidents, that sort of traffic makes it way too easy for fender benders and such to occur and the stress level is pretty high. A car that dealt with that for me would probably add at least a year to my life.

Re:Well that's wrong

[fluffernutter]

So then you admit putting AI on the roads today amounts to vehicular manslaughter then. Thanks for being honest. The question is whether automation will be able to survive the lawsuits, and I hope any family that loses a member will bring the fires of hell down on the company responsible for releasing that particular AI. Even more importantly, the execs of that company should have to meet with the family and explain to them why their solution was worth their son's/daughter's/father's/mother's life. Relieving people of the burden of driving doesn't quite cut it with me, because that really is a first world problem. This all just seems to be about luxury and convenience.

Re:Well that's wrong

[goose-incarnated]

In fact on average a computer will have much better visibility to the full surroundings than a human driver could, so even if "dumber" the computer is better informed and so statistically will probably make better choices.

Why would you assume that? The electronic visibility of an airplane is much better than any pilot, and yet 40 years after the invention of the "autopilot" we still do not have software that matches the pilot.

"Self-piloting" is a much much easier problem to solve than "self-driving", and yet we haven't solved that yet. I expect that 5 years after we've gotten passenger aircraft fully autonomous, we'll have self-driving cars.

But lets first see if anyone can solve the easy problem first, okay?

Re:Well that's wrong

[david_thornley]

We have reliable software in many areas where safety matters. There's still some bugs, but they rarely have serious effects. (We learned from the Therac-25 incident.) We know how to do reliable software; it's just that it costs a lot more than letting your customers beta-test it.

Re:Well that's wrong

[david_thornley]

There's no comparison here between cars and jumbo jets. Figure that the cost of operating a vehicle is something like fifty cents per mile, so at freeway speeds the car costs about $30/hour. You're not going to get many good drivers for less than $15/hour (including all the indirect expenses here), so a driver is at least one-third of the expense.

Airliners are far more expensive, so even though the flight crew is paid more the cost savings in removing them would be small.

Re:Well that's wrong

[david_thornley]

AI isn't ready to be on the roads today, so it doesn't matter whether it's vehicular homicide or not. When it is ready, it's likely to be somewhat safer than a human, so going out and driving your own car will be a little more like vehicular homicide. You seem to be assuming that humans never kill each other in accidents, which means you never looked at the statistics. If self-driving cars take over and kill ten thousand people a year in the US, that's a big improvement over the status quo.

We have laws governing legal liability for traffic accidents. Self-driving cars will fit into that legal framework. We have drunk drivers killing people and not apologizing to the survivors, even though they deliberately did something unsafe towards everyone on and near the road.

You seem to have a delusional belief that human drivers don't kill people, and therefore that an AI that kills someone will be some sort of unprecedented atrocity. You're wrong.

Re:Well that's wrong

[fluffernutter]

I'm not saying humans don't kill each other in accidents. I'm saying that I'm not ready to believe that there will be enough people with automation to make a dent in the accidents that happen. People can't afford it under capitalism. So any accident that automation causes today is just another death, not part of some grand scheme. I mean that whole thing almost sounds like a religion.. the great oncoming of our robot saviors. Automated cars will always be luxury items, or at least there is no evidence otherwise. So let's not kill more people by rushing them out. It's half-baked.

Re:Well that's wrong

[david_thornley]

We sure don't want to rush these out half-baked, and I'm happy to report that the Pittsburgh experimental vehicles have drivers ready to take over at a moment's notice. One observer reported difficulty in staying engaged enough, which was expected, and will continue to be a problem.

I don't think automated cars will always be luxury items. They'll cost more than non-automated cars, unless we get to the point where we no longer need driver controls and displays, and if that happens at all it's a long time in the future. If they show sufficient advantages in safety, they may be mandated at some point down the line, like seat belts and catalytic converters. I believe there will always be non-automated vehicles for things like race cars, but they may eventually not be considered street-legal.

For a minimum time on that, figure at least another five years to make autonomous vehicles usable, another five to notice the safety difference, and another fifteen to change over. I am not at all confident it will happen in my lifetime.

Lyft (and Uber) good for the drivers? Hah.

[nicolaiplum]

So Lyft (and Uber) are good for the drivers, eh? All the claims of offering people flexible work, improving earning chances of people without firm employment, etc, turn out to be complete rubbish.

Lyft and Uber are after the money, not helping anyone, not creating jobs, just the money. I feel sorry for anyone tricked into expecting Uber (especially) would be a long lasting way to work for money; they were just a disposable tool on the way to more money for the investors and executives of Lyft and Uber.

Re:Lyft (and Uber) good for the drivers? Hah.

[drinkypoo]

I feel sorry for anyone tricked into expecting Uber (especially) would be a long lasting way to work for money

I feel sorry for anyone tricked into believing in long-lasting ways to work for money. There are some, like booze and candy. People buy those even in recessions and they're easy enough to make that automation hasn't eliminated humans.

They should be banned.

[JustNiz]

>> but will only drive up to 25 miles per hour.

Can you imagine how pissed off people will rightfully get when stuck in a giant tailback of traffic behind one of these f***ing things.

Re:They should be banned.

[fluffernutter]

In a lot of places that would constitute blocking traffic and therefore be a moving violation.

Re:They should be banned.

[HornWumpus]

You will get an obstructing traffic ticket for going the speed limit in California.

The rule is simple, if you have 3 cars backed up behind you, pull over.

Less true in school zones than divided highways.

Re:They should be banned.

[Joe_Dragon]

driving 55 mph on I-294 can get you almost run off the road.

Re:They should be banned.

[simp7264]

There is no way this is true that you will get a ticket for going the speed limit. Do you have some source for this? Everything I see says the opposite but can't find a great source.

“A vehicle driving in the left lane of a multi-lane highway slower than other traffic, but at the speed limit, is not in violation of the law,” says Erin Komatsubara, a public information officer for the CHP. Vehicles passing on the right of a left-lane driver who is at the speed limit would be violating the speed limit, Komatsubara says.
http://www.autoinsurancecenter...

Uhh....

[easyTree]

Lyft Says Robots Will Drive Most Of Its Cars in Five Years

That's nothing. My new crowd-sourced-transport startup will have mostly robotic passengers by 2017!

Re:uh huh

[darkain]

1 word: parking.

Only applicable in urban hipster neighborhoods

[ErichTheRed]

Outside of cities, I can't see the economics of this working. Telephone service in rural areas had to be subsidized by a universal service fee. Why? Because the for-profit telephone company, even with a monopoly, didn't want to extend the network for a small number of customers unless there was an incentive. Imagine Uber/Lyft having to guarantee that one of their self-driving cars would be available to take you wherever you wanted to go, 24/7, with 30 minutes' notice regardless of where you live.

The other reason why I don't think personal cars are completely doomed is families. If you have kids, you know that the car becomes another room of the house if you live in the suburbs and have to drive everywhere. Imagine having to haul all your crap out of your self-driving Uber cab when you reach your destination, then put it back into another car when you want to go back.

I think some of this stuff is really cool, but the business model seems exactly like a myopic view of the entire world being a dense city filled with well-to-do hipster singles or married people who don't have kids. It's the same model as Blue Apron and all those other delivery services...Ironic mustache and goatee Swift developer and his marketing liaison coordinator wife arrive home from another 12-hour shift at the unicorn startups they work at. Rather than call an Uber to take them to the trendy new Ethiopian-Thai fusion place again, or hang out with the LUDDITES at the grocery store, Blue Apron has a box delivered to their front door with meals in it! It's brilliant! Everyone will love it! Give us $100 million!!!

Re:Only applicable in urban hipster neighborhoods

[Gilgaron]

Sure but it could be useful for a family who might then only need to own one car. There will likely always be a need for personal vehicles, but I'd agree that the trend is likely that there will be fewer of them.

Re:Only applicable in urban hipster neighborhoods

[psmoot]

Outside of cities, I can't see the economics of this working.

We recently crossed a threshold where more than half the world's population lives in urban areas. Just sayin'

The other reason why I don't think personal cars are completely doomed is families.

I'm with you there. My wife is the queen of having bags of stuff in her trunk, mostly purchases which need to be returned. She's never sure when she'll be flying by a store she needs to stop into.

I also pretty sure Zimmer is drinking his own Kool-Aid. I really doubt truly autonomous vehicles will be on the road that soon, let alone dominating. Autonomous driving is really, really hard and getting to the point where you're absolutely, positively sure you'll never need a driver is a really high bar. Then it takes people 10 years to trade in their cars for a new generation so don't expect our new robot overlords for at least another two decades.

Re:Only applicable in urban hipster neighborhoods

[Sir_Eptishous]

Ironic mustache and goatee Swift developer and his marketing liaison coordinator wife arrive home from another 12-hour shift at the unicorn startups they work at. Rather than call an Uber to take them to the trendy new Ethiopian-Thai fusion place again, or hang out with the LUDDITES at the grocery store, Blue Apron has a box delivered to their front door with meals in it! It's brilliant! Everyone will love it! Give us $100 million!!!

Congratulations!
You just won the Internets!

Re:Only applicable in urban hipster neighborhoods

[HideyoshiJP]

...hang out with the LUDDITES at the grocery store...

Hey, we found the app appers guy!

Re:Only applicable in urban hipster neighborhoods

[Gussington]

Outside of cities, I can't see the economics of this working.

And?
We don't build subways in rural towns either for the same reason. They still work though.

Re:Only applicable in urban hipster neighborhoods

[RespekMyAthorati]

We recently crossed a threshold where more than half the world's population lives in urban areas. Just sayin'

That still leaves 3 1/2 billion who don't. Just sayin'.

Re:uh huh

[DirkDaring]

It's not a cost saving, its far more expensive in fact. Can you imagine paying around $10 for every 1-way trip you do? I'm guessing I do 50 1-way rides a week all the running around to work, kids events, shopping, etc. $500 a week is not what I call saving money.

Re:Laws and other stuff will take longer then 5 ye

[flyingfsck]

Judges don't need new laws to make decisions. New laws may actually hamper them more than it helps. That is why they are called Judges. They judge.

What Would Brian Boitano Do?

[kackle]

I JUST got lunch two (albeit long) blocks away in our industrial park. Unlike ever before, the street was loaded with parked semis, a forklift flitting about, and a landscaper's truck with trailer. Traffic was relegated to one lane. This artificial lane was many semi trucks long. As a firmware guy with decades of experience, I see no way for contemporary technology to sort out that situation in a fraction of a second, especially as there is also an intersection ahead at the end of those semis where oncoming traffic would NOT be using the single lane. An AV would be baffled by such a random, but common, occurrence. (Can you imagine if the street was curved?)

Start paying attention and you will notice all kinds of "puzzles" in your own daily driving that would force the AV to a crawl or stop, or worse, keep going as is.

Re:And that's where you are wrong.

[HornWumpus]

Again? Really?

Once again. Tesla is full of shit. They compare 'autopilot' accident rates to all human driving. They should compare autopilot rates to human driving on divided highways. They don't because that would make them look bad.

If you haven't already done so, read 'How to Lie with Statistics'.

bald tires, upatched OS, questionable brakes

[Thud457]

Eliminating the driver also eliminates the person most likely to complain loudly about life-threatening deferred maintenance.
win-win

Re:bald tires, upatched OS, questionable brakes

[ShanghaiBill]

Eliminating the driver also eliminates the person most likely to complain loudly about life-threatening deferred maintenance.

Almost all traffic accidents are because of human error or inattention. The second biggest reason is road conditions. Poor maintenance and mechanical failure account for about 3% of all accidents, but less than 1% of fatal or injury causing accidents.

An accident, or even a mechanical breakdown on the road, is much more expensive than routine maintenance. So a profit-maximizing company would ensure that routine maintenance is done. This is no different than rental car companies today.

Re:bald tires, upatched OS, questionable brakes

[OrangeTide]

A few really nasty limo accidents that made it into the news were due to maintenance issues. But part of that is a design flaw of limos that wouldn't exist in a standard car. (occupants can usually escape a car if there is a fire)

Re:And that's where you are wrong.

[drinkypoo]

Once again. Tesla is full of shit. They compare 'autopilot' accident rates to all human driving. They should compare autopilot rates to human driving on divided highways. They don't because that would make them look bad.

So, have you done that? Has anyone? And if so, where's that comparison? I've looked up some related statistics but none of them speak directly to this point. Over one-third of collisions are on divided highways, for example; that suggests that in fact the comparison would not make them look bad, but I honestly don't know.

Re:And that's where you are wrong.

[HornWumpus]

Yes I did. IIRC divided highway accident rates are about 1/4 the average for all driving. (You can Google for yourself, it's not hard to find. You want the accidents/mile not accidents/hour to compare to Tesla stats.) By far the safest mode.

Tesla's in autopilot are (by current data) much more dangerous than regular cars in similar situations.

This has been beat to death in the last three Tesla threads. But the fanbois keep bullshitting/parroting.

The future of automobiles? Govt fleets

[mveloso]

The future of automobiles will be autonomous fleets run by your government entity.

Why?

Because it's safer and more cost effective. You need a car because you need to go from point A to point B. What if you had the convenience of a cab without the cost? What if you could it whenever you wanted, and didn't have to wait for a scheduled pickup like a bus?

And what if the roadways were reserved for these cars and "manual drivers" only had one lane of road?

Why wouldn't you use the car service?

Instead of spending money on your car, you could spend it on something else.

The fleet would be all-electric, of course. If you needed longer trips you'd go to a normal car-rental place.

And of course you could always have your car - at a cost. A very high cost. Your insurance premium would be sky-high, because 90% of drivers would opt for the cheaper, more convenient automated fleet.

Your house would be bigger, because you wouldn't need a garage.

You'd have more money (maybe), because you wouldn't have to operate your car. OTOH your taxes would be higher, since the govt would be operating the cars for you.

I think it'd be great.

Re:us secret service with not use auto drive car a

[epyT-R]

Since when do these freedom restricting technologies ever apply to the ruling class?

Re:Long haul

[AvitarX]

Because the cost per passenger of a bus driver is far less than the cost per passenger in point to point transit.

Predicting the future, is rather stupid.

[geekmux]

"...Zimmer, who has long been a vocal proponent of ending car ownership, set a date for the death of the personally owned car in major U.S. cities: 2025."

Hey Zimmer, here's an eye-opening revelation for you; the overwhelming majority of car owners today have never used your product, and never intend to.

Given that fact, kindly STFU about the death of car ownership. You could at least wait until you handle your first wrongful death lawsuit due to your premature AI deployment schedule before making such asinine predictions.

The DEVIL's cybernetic jitney

[Thud457]

I still say the only way for these newfangled contraptions to be safe is to hire a man to walk ahead of it waving a flag, or at night or in inclement weather, a lantern.

Car ownership is, in general, a terrible thing

[luis_a_espinal]

Says that self driving cars will put an end to car ownership.

It could put a dent in it but unless this makes people so broke that they can't own their own car I think personal space will still win out.

We own cars because we have to. Our public transportation systems usually suck, and we cannot go anywhere (jobs, entertainment, whatever) without it. I refer to you (as I've done many times in other threads) to Tokyo. You can party, go to work and take your kids anywhere in the whole Kanto region without a car.

As our car ownership increases, the negative impact in our infrastructure, economy, environment and productivity decreases. Cities around this country are battling with how to deal with this. Out of necessity, car ownership (or at least individual driving within a city) will decrease.

Autonomous cars will be one of the many ways this will be addressed. It will happen. And car manufacturers are betting on it and trying to jump ahead of it instead of being relegated to the losing side of things.

Business-wise, it is the smart move.

Re:Car ownership is, in general, a terrible thing

[goose-incarnated]

Says that self driving cars will put an end to car ownership.

It could put a dent in it but unless this makes people so broke that they can't own their own car I think personal space will still win out.

I refer to you (as I've done many times in other threads) to Tokyo. You can party, go to work and take your kids anywhere in the whole Kanto region without a car.

That only works in very high-density areas. Those of us who don't live in battery-hen conditions won't be getting good access to public transport anytime soon. You may enjoy living in such a manner that four or five of the six sides that make up your home are shared with someone else. I do not.

I won't be moving to a high-density area anytime soon as I'm used to my 700sqm house on 4000sqm land with a 80000l pool in the yard. and (literally) park-like front garden. You probably paid the same as I did, except you got tiny cube in the city. But, hey, at least you don't need to own a car!

Re:Car ownership is, in general, a terrible thing

[luis_a_espinal]

Says that self driving cars will put an end to car ownership.

It could put a dent in it but unless this makes people so broke that they can't own their own car I think personal space will still win out.

I refer to you (as I've done many times in other threads) to Tokyo. You can party, go to work and take your kids anywhere in the whole Kanto region without a car.

That only works in very high-density areas. Those of us who don't live in battery-hen conditions won't be getting good access to public transport anytime soon. You may enjoy living in such a manner that four or five of the six sides that make up your home are shared with someone else. I do not.

The fact that you call it a 'battery-hen condition" shows you have never been in Tokyo... so it is hard to say how objective your description is. And even when you step out of Tokyo, say, go Chiba, you find sizeable homes... and yet, you still have access to excellent public transportation. You barely need a car.

I don't live in Tokyo currently, but I've stayed in Shinawaga (beautiful condos with lots of greenery) and in Yokohama where my in-laws live. Their 2-story house is not large, as it lacks a large patio, but it is very comfortable, not to mention the adjacent amenities, green areas and what not that come with a well-run alpha city.

The fact that large cities in the US do not have good public transportation is simply lack of planning. There is nothing inherently in urban sprawl or large homes that preclude us from having it.

I won't be moving to a high-density area anytime soon as I'm used to my 700sqm house on 4000sqm land with a 80000l pool in the yard.

Which is OK, but that doesn't preclude us from having good transportation. My house is 167 sqm (I wouldn't live in anything over 371 sqm, the cleaning just gets a part-time job) with a good enough patio for fruit trees and stuff. It is in a community of houses that size and big, with plenty of green, among similar communities...

... it is situated in a city ranked the 8th best in the country (Weston), 7 miles away from Southwest Ranches - a ranching community where I might buy a 2-acre/8000 sqm property, double of what you have, when my kids get older - all enclosed within the 8th largest metropolitan area in the US (South Florida) with a 300 billion dollar economy rivaling Singapore.

We have a lot of condo canyons mind you, from the claustrophobic to the incredibly spacious (with the only think lacking being a patio). But we have a large, specious homes and ranches.

There is nothing, ABSOFUCKINGLUTELY NOTHING that preclude us in South Florida (or any well-run megapolis) from having good public transportation other than myopic self-interests, politics and lack of urban planning.

When I lived in LA, I did not need a car at all. I lived in Van Nuy and worked in Burbank and traveled to Downtown, Santa Monica and Hawthorne on a regular basis. LOOK MA, NO CAR!

Buses travel every other artery at a 30-min frequency, and houses weren't small (though like in all urban areas, there are more condos than full-sized homes.) Traffic is terrible obviously, but public transportation is absolutely superior to what we have here in South Florida, and more than sufficient for people to exercise job mobility without having to own a car.

and (literally) park-like front garden.

Good for you. Since we are going to compare properties, why not? I have a very nice lake view on my backyard with plenty of fruit trees and some northfolk pines or weeping willows in the near future. Here is a picture from the front going back all the way to the lake on the backyard. I still have some fixes to do. Not a 700sqm home, but certainly not a box.

http://bit.ly/2cArnM4

And you know what's awesome? Within driving distance I can t

Re:Car ownership is, in general, a terrible thing

[luis_a_espinal]

That only works in very high-density areas.

Besides, my observation was obviously meant for high-density areas, like pretty much every important city in the US, which is where the majority of the people live, where the engines of the economy run. It is not applicable to small density areas, which as charming as they might be, are not where the bulk of the nation is, and which are more economically unsustainable as time goes by.

So I really don't get what the hell your comment was all about.

5ys plan

[umghhh]

There have been many good people that were obsessed with 5ys plans. But even diedushka Jozef S. who had that fever for 5y plans too, failed miserably or rather the plans failed him miserably...

News reports from another planet?

[mbone]

I wasn't aware that we had made contact with aliens, much less that we were receiving news reports from another planet.

Well that's even wronger.

[SuperKendall]

Lol.. So you think people will use self driving cars if they slow down every time they pass a parked car?

If I'm in a self driving car, it means I no longer have to pay attention to the road. It means that I don't even notice it slowed down a bit for parked cars, or that it's already going a bit over the speed limit instead of way over like most humans would in that situation.

So yes people will buy them in droves because they can actually focus on something else while in the car, then whatever caution the car takes simply doesn't matter to the buyer because of the massive gift of time they receive in return. Do you honestly not understand that, even a little?

Re:Well that's even wronger.

[fluffernutter]

There was already a discussion about this. Even most people on slashdot admitted that they would not be completely productive in a perfect self driving car. The goal will always be to get from point A to point B in the fastest amount of time. Also, going unnecessarily slow is frustrating to other drivers, and might amount to a violation of impeding the flow of traffic.

Re:Well that's even wronger.

[drinkypoo]

Even most people on slashdot admitted that they would not be completely productive in a perfect self driving car. The goal will always be to get from point A to point B in the fastest amount of time

No. In a commute situation I can get from point A to point B very quickly by driving a tank and simply driving over the top of traffic in between. The goal is to do so safely and legally.

Also, going unnecessarily slow is frustrating to other drivers, and might amount to a violation of impeding the flow of traffic.

It's not unnecessary. The law already demands that you not drive at speeds unsafe for conditions. That means slowing down when you pass parked cars, because someone might jump out from between them. You're just ignoring the law for your convenience at the expense of public safety. You are a menace and self-driving cars can't come fast enough to take driving out of the hands of people like you. I bet you don't even slow down when you're in the passing and/or carpool lane going past a long row of stopped (or excessively slowed) traffic that might suddenly pull out in front of you. I recently lost a Fb friend for insisting that motorcyclists who are lane splitting have responsibilities that they're ignoring, and that's why they've got a bad name. You know what that responsibility was? Passing at a safe speed. If you pass something too quickly you give up the ability to react to it, because of the way cars move. And that's what you are apparently doing ever time you drive through city streets.

Better data means better driver, period.

[SuperKendall]

Vastly better sensors and instant reflexes outweigh intuitive understanding tied to slow reflexes, every time.

I say this as a driver who loves to drive, who has trained to do autocross... I am under no illusion that even trained and always scanning for issues as I drive, I am in any way better than a few sensors all around the car constantly monitoring for everything.

Re:And that's where you are wrong.

[goose-incarnated]

Once again. Tesla is full of shit. They compare 'autopilot' accident rates to all human driving. They should compare autopilot rates to human driving on divided highways. They don't because that would make them look bad.

So, have you done that? Has anyone? And if so, where's that comparison? I've looked up some related statistics but none of them speak directly to this point. Over one-third of collisions are on divided highways, for example; that suggests that in fact the comparison would not make them look bad, but I honestly don't know.

No need to go that far - they are comparing "autopilot in perfect driving conditions with a human ready to take over when the software gets it wrong" with "All human drivers, in every road condition imaginable".

You shouldn't need a book on statistics to know that the comparison of "autopilot+human+perfect conditions" with "humans+all conditions" is not a statistical comparison but a marketing one.

In short, you shouldn't need to be told that the comparison is not legitimate; it's obvious after a moment's thought.