Uber Will Begin Picking Up Pittsburgh Users in Self-Driving Cars This Month

[Little Endian]

Driverless cars will be reality over time. This will take decades to happen though. My question is will uber be able to make it work? And if so will it result in better prices for the consumer? Ubers current buisness model is as a software technology company. They make an app and are responsible for marketing thats it. Uber does not own fleets of cars, and or the support staff like engineers and mechanics to maintain them. Uber does not need to worry about fuel/electricity or employee drivers. Uber also has a much smaller insurance liability since they only need to cover their passengers. They do not have to worry about vehicle or the drivers insurance liabilities. The driver is responsible for providing the primary physical and labor capital for this whole enterprise to work not to mention the risks. Ubers surge model of pricing also helps create equilibrium in the market making sure there is adequate supply to meet demand.

In a world of driverless cars Uber would need to pony up the capital for millions of driverless cars each probably costing 100k or so. Uber would be 100% responsible for the mainatanenece, fueling, repairs, and insurance. Thousands of engineers, mechanics and other service personnel would need to be hired to maintain these company owned vehicles. Uber would also be exposing itself to much greater risks in terms of liabilities and market fluctuations. Ubers buisness model works so well because it is their driver partners who assume capital costs, risks and liabilities. Uber as of now makes money regardless of what happens just as long as someone uses their service.

 

[dwaltwhit]

So why Pittsburgh?

 

[Intell]

So why Pittsburgh?

Because that's where CMU, that partnered with Uber to make this until Uber poached it all from CMU, is located. Plus, Pittsburgh is a nice little tech center.

 

[dwaltwhit]

Because that's where CMU, that partnered with Uber to make this until Uber poached it all from CMU, is located. Plus, Pittsburgh is a nice little tech center.

I figured it was business related, but I am not very familiar with cmu.

 

[bozzykid]

How did Uber pull this off so quickly?

By limiting this to only one city and likely only certain potential pickups and destinations.
[doublepost=1471577220][/doublepost]

I really can't believe that Uber is able to do this. I can't imagine that they have done more research and development then Google or Tesla.

Of course Uber hasn't done more research than Google. But I'm sure Google could run a taxi service now if they limited it to only a few roads like Uber is doing. The two companies are just taking different approaches.

 

[macduke]

I just bought a Subaru Outback. If nothing bad happens like an accident, this will very likely be the last car that I ever buy that is not electric or autonomous. It was weird thinking about that. A lot is about to change over the coming years. I figure now is a good time to buy because when I'm ready to buy again, the technology should be a bit more mature.

 

[mw360]

Driverless cars will be reality over time. This will take decades to happen though. My question is will uber be able to make it work? And if so will it result in better prices for the consumer? Ubers current buisness model is as a software technology company. They make an app and are responsible for marketing thats it. Uber does not own fleets of cars, and or the support staff like engineers and mechanics to maintain them. Uber does not need to worry about fuel/electricity or employee drivers. Uber also has a much smaller insurance liability since they only need to cover their passengers. They do not have to worry about vehicle or the drivers insurance liabilities. The driver is responsible for providing the primary physical and labor capital for this whole enterprise to work not to mention the risks. Ubers surge model of pricing also helps create equilibrium in the market making sure there is adequate supply to meet demand.

In a world of driverless cars Uber would need to pony up the capital for millions of driverless cars each probably costing 100k or so. Uber would be 100% responsible for the mainatanenece, fueling, repairs, and insurance. Thousands of engineers, mechanics and other service personnel would need to be hired to maintain these company owned vehicles. Uber would also be exposing itself to much greater risks in terms of liabilities and market fluctuations. Ubers buisness model works so well because it is their driver partners who assume capital costs, risks and liabilities. Uber as of now makes money regardless of what happens just as long as someone uses their service.

Yes, you're nearly there. Change one assumption about who will own the cars, and you find that Uber doesn't need to change their business model at all. They don't need to lay off all those million 'partners' and they don't need to take on all that extra infrastructure and liability.

 

[macs4nw]

So Uber's goal is to put their drivers out of business???

Pretty much..... eventually, once they have recouped all the upfront costs, all the profits will be theirs.

I believe Apple is heading in the same direction with their project 'Titan'. Forget about complicated distribution models, scattered service centres, and warranty headaches. Commercial for-hire transportation with centralized maintenance facilities is a market segment that will explode in the next twenty years. With plentiful clean, modern and safe self-driving transportation options available, the need for outright car ownership, at least for urbanites, will become increasingly unattractive, not even mentioning insurance, fuel, maintenance & upkeep as well as parking space headaches and cost.

Those who get this right (and there are currently an increasing number of players working furiously behind the scenes) will be sitting on a gold mine by the middle of the next decade.

Additionally, these radically new 21st century transportation options will I believe, have the potential to all but eliminate the un-appealing public transportation systems we're currently propping up at great cost to the taxpayer.

 

[617arg]

Yes, you're nearly there. Change one assumption about who will own the cars, and you find that Uber doesn't need to change their business model at all. They don't need to lay off all those million 'partners' and they don't need to take on all that extra infrastructure and liability.

Exactly. Uber doesn't have to change its business model at all. Uber drivers will be able to send out their cars while they are at work or relaxing on the beach.

 

[theluggage]

Imagine you're in the back of a self-driving car going very fast, and suddenly a six year old kid chasing a ball darts out in front of the car. Okay, not too hard. The car must steer itself into another car or telephone pole, possibly sacrificing you, the adult rider, in favor of the child. That's the moral thing to do.

Imagine you have invented a self-driving car that proves to be vastly safer than the typical human driver by dumbly obeying all the rules about speed, maintaining safe distance, not getting distracted or drunk, and using optimal braking strategies. Should you suppress this invention, and fail to save thousands of lives a year, because occasionally, under very rare and contrived no-win conditions, it will kill a pedestrian instead of the driver?


Meanwhile, the correct response to your dilemma - which should be programmed into cars - is: do your best to stop in a controlled and predictable manner, because a car spinning, rolling or mounting the sidewalk - most likely after failing to avoid the child - is likely to cause all kinds of unknown mayhem. In reality you're not going to have the luxury of more than a split second to weigh up the morality of the outcomes, nor have enough data to truly understand the alternatives (you are sure that the girl's 5-year old brother isn't sitting behind the telephone pole, that the other car isn't rushing a pregnant woman to hospital... and you did renew your life insurance so your 4 kids aren't going to be left destitute when you heroically sacrifice your life to save a jaywalker...?)

Someone's been reading too many Isaac Asimov books and/or thinking too hard about silly "moral" dilemmas about runaway trains & pushing fat men off bridges... You know, the ones where everybody argues about the relative values of human life and rationality vs. morality but never for one minute stop to consider how the heck you would know, in advance, all those precise details of the consequences of each option.

I think some states have already determined that the car is the legal driver. Therefore it will be responsible for any decisions it makes.

Google, Uber, Tesla are building self-driving cars. They're not building HAL*, GlaDOS, Skynet, R. Giskard or Multivac (the last two were the Asimov characters closer to the theme of that Will Smith trainer commercial than the short story collection iRobot).

The car is a lump of metal and circuitry. It can't be "responsible" for decisions - the liability will probably be with the human operator (unless it can be shown to be the victims fault) or, in the US, whoever looks like being most profitable to sue. These things shouldn't be made generally available until the law has been revised/clarified and suitable insurance schemes made available: I'm not setting foot in an autonomous car if I'm personally liable for the consequences any bugs in the software - the makers will have to provide some sort of indemnity.

*"Open the nearside door please Car"
"I'm sorry Dave, I can't do that. This journey is too important to allow your presence to jeopardise it."

 

[_Refurbished_]

You're not getting the picture. Much like folks who thought Amazon.com was "a little bookstore on the web".

The idea with Uber is that these folks aim take over transportation. "Taxi" as it exists today is peanuts and is not their target.

What they aim to lobby their way into is that: you wont have a municipal transportation run by your local govt plus private taxis and private cars. They, alongside the likes of Google - and Apple too - are angling basically for transportation systems in which not only will vehicles have no drivers, but personal cars will be priced or regulated out of the hands of the average person on the grounds of "optimizing system capacity". You must take a common public transport vehicle, driverless even if its a "single person vehicle that you rented". You'll have such services contracted out, such as bus and personal transport, to....them.

Private monopolies on your mobility. And there wont even be jobs for drivers either.

That is how you get to their equity valuations....

Wake up....

Just woke up, it's 6:26 here.

Amazon is the Walmart / Target of the Internet. Someone had to become the dominant force and it happened to be them. They became what they are today because they expanded past books. What you're talking about with Uber requires massive support from the government, plus support from the masses....which they won't get. People want the freedom to own a vehicle with their equipment and belongings already in place. Imagine having to put two child seats into an Uber everyday. Yea...

If anything companies will need to be the ones being regulated, so they don't overcrowd the streets with their robots. Families will always be allowed to have at least two vehicles.

Private Message me in thirty years with your heartfelt apology.

 

[kdarling]

Imagine you have invented a self-driving car that proves to be vastly safer than the typical human driver by dumbly obeying all the rules about speed, maintaining safe distance, not getting distracted or drunk, and using optimal braking strategies. Should you suppress this invention, and fail to save thousands of lives a year, because occasionally, under very rare and contrived no-win conditions, it will kill a pedestrian instead of the driver?

Yes, that's the question. But the basic condition is not contrived. Every year in the US, four thousand pedestrians are killed and 59,000 injured.

The question of ultimate responsibility WILL come up. E.g. in California, it was proposed that the passenger would be ultimately responsible:

"The proposed rules hold motorists responsible for obeying traffic laws, regardless of whether they are at the wheel.

Deciding if the car or its occupant is responsible for accidents and other mishaps has been at the center of debates over how to regulate driverless cars."

- WSJ

As I said, I think that one state put the responsibility on the car; I'll try to find that reference.

Meanwhile, the correct response to your dilemma - which should be programmed into cars - is: do your best to stop in a controlled and predictable manner, because a car spinning, rolling or mounting the sidewalk - most likely after failing to avoid the child - is likely to cause all kinds of unknown mayhem.

I question if such a simple response will be seen as enough. The first time a mother pushing a baby carriage is hit by an autonomous car, there will be an outcry for better software.

For example, Tesla is very lucky that that owner who recently died while using their auto-drive mode, hit a cargo truck and not a school bus. If that had happened, I suspect that auto-drive mode would've been banned in all fifty states, until it was programmed to avoid hitting school buses at all costs.

In reality you're not going to have the luxury of more than a split second to weigh up the morality of the outcomes,

In reality, humans do this kind of split second decision all the time.

For instance, it's common for people to swerve and hit a pole or parked car, instead of hitting another car coming at them, because the driver instantly decides it will be cheaper and less likely to hurt the other car's occupants.

It's also really common for people to swerve into something else, to avoid a child (or heck, even a squirrel or pet!).

So right off the bat, your idea of simply applying brakes is not going to be seen as the best option.

Someone's been reading too many Isaac Asimov books and/or thinking too hard about silly "moral" dilemmas ...

As I said, it was something I read in an article. But I agree with the author that it's important to think ahead about the ramifications. The first time a bad choice is made and it gets national attention, it will become a bigger issue.

I'm not setting foot in an autonomous car if I'm personally liable for the consequences any bugs in the software - the makers will have to provide some sort of indemnity.

That's a very valid comment.

 

[Osamede]

Just woke up, it's 6:26 here.

Amazon is the Walmart / Target of the Internet. Someone had to become the dominant force and it happened to be them. They became what they are today because they expanded past books. What you're talking about with Uber requires massive support from the government, plus support from the masses....which they won't get. People want the freedom to own a vehicle with their equipment and belongings already in place. Imagine having to put two child seats into an Uber everyday. Yea...

If anything companies will need to be the ones being regulated, so they don't overcrowd the streets with their robots. Families will always be allowed to have at least two vehicles.

Private Message me in thirty years with your heartfelt apology.

No really you don't get it. Amazon was not founded for the sake of selling books. Books was the first business they chose to enter as they built a platform.

Uber is spending mega money on lobbying. They are not doing it for the taxi business. They are doing it to set the rules of the game for them to own it all.

What Joe Blow wants is irrelevant in this country, once the powerful decide they want something else, in case you didn't know:
http://www.bbc.com/news/blogs-echochambers-27074746

That research is not abstract- it's the underlying process behind which the rules of the game are made for technology innovations, which is the goldfish right now.

It's not about taxis. Never was. Same as Amazon was about far more than books don't be naive think they Uber will tell you nicely and honestly that they want to own and tax your every move that doesn't involve your feet, whether you like it or not.

 

[0007776]

I think they are getting into the game too early. Let the car manufacturers bear the brunt of the cost of developing and testing the cars. Let them get it to the point where there is no safety driver and passenger. Those two are not going to take the ride for free, and are going to cost Uber money. A lot more money than they are currently paying drivers. I can only assume they are working some angle where they see a quick direct line to the same profit levels as they currently enjoy.

Is uber actually paying for the engineers to be there or is it a partnership with Volvo where Volvo gets the benefit of real world testing and uber gets to provide the customers while Volvo provides the engineers?

 

[_Refurbished_]

No really you don't get it. Amazon was not founded for the sake of selling books. Books was the first business they chose to enter as they built a platform.

Uber is spending mega money on lobbying. They are not doing it for the taxi business. They are doing it to set the rules of the game for them to own it all.

What Joe Blow wants is irrelevant in this country, once the powerful decide they want something else, in case you didn't know:
http://www.bbc.com/news/blogs-echochambers-27074746

That research is not abstract- it's the underlying process behind which the rules of the game are made for technology innovations, which is the goldfish right now.

It's not about taxis. Never was. Same as Amazon was about far more than books don't be naive think they Uber will tell you nicely and honestly that they want to own and tax your every move that doesn't involve your feet, whether you like it or not.

You're the only person using the word taxi.

As pointed out in my OP, there are "endless" possibilities for robotic cars. Not sure why you're placing me into the narrow minded group of people that think Uber is going to be a taxi company in 20 years....they'll have a much stronger hold on the market with autonomous vehicles.

The rest is fud. Uber doesn't have that kind of power, nor does our government. Maybe in China.

Your trajectory is quite off on this one and I'm sure, one day, you'll be able to admit that. I'm betting on me, talk to you in twenty. No point discussing this further.

 

[Osamede]

"....Uber doesn't have that kind of power, nor does our government. Maybe in China..."

You seem oblivious to the manner in which the tech companies have broken the existing rules of the game.

As for "China" your govt has more power over its citizens than China, improsons more if it's citizens than China and spies on them to a level than China can't match- all enabled by the tech industry.

"Uber" is a facade. Ditto for its counterparts. Together they will carve up this pie. Look and grasp that VC tech and financial sector giants are the drivers here. They rule. That's what that study tells you. Read it and stop taking about China. This is going on here at home.

 

[theluggage]

But the basic condition is not contrived. Every year in the US, four thousand pedestrians are killed and 59,000 injured.

...but not in the circumstances you describe, where the driver has the opportunity to sacrifice themselves rather than hit the pedestrian. Odds are that the vast majority of those deaths were due to basic driver inattention, driving at unsafe speeds or losing control of the vehicle, and that even today's self driving cars would have avoided them.

The first time a mother pushing a baby carriage is hit by an autonomous car, there will be an outcry for better software.

If that had happened, I suspect that auto-drive mode would've been banned in all fifty states, until it was programmed to avoid hitting school buses at all costs.

That's likely to happen whatever. If the authorities start to apply higher safety standards to AVs than they do to humans, it will kill the concept because they will never be good enough and end up driving everywhere at 20mph and stopping every time a leaf blows across the road.

In reality, humans do this kind of split second decision all the time.

I'd debate that. I suspect most people would either swerve without looking (and possibly cause even more damage) or just stand on the brakes or, worse, try to do both and flip the car.

For instance, it's common for people to swerve and hit a pole or parked car, instead of hitting another car coming at them, because the driver instantly decides it will be cheaper and less likely to hurt the other car's occupants.

But is that assumption true? You can't see if anybody is in the parked cars, you can't see if anybody is standing behind or between the parked cars, you might rebound back into the road and hit the other car anyway, the other car might swerve in the same direction. Unless there's a very clear "escape route" I'm still gonna vote for stand on the brakes, brace, and trust your anti-lock brakes and airbags. Again, though, your AV is less likely to be driving at 50 down a single-track road lined with parked cars without looking where it is going.

It's also really common for people to swerve into something else, to avoid a child (or heck, even a squirrel or pet!).

...again, not necessarily rational unless you are really sure that (a) it really is a child and not a squirrel and (b) there's nobody else standing where you're about to swerve. Also, any idea which is worst for the most people? Hitting a pedestrian at 10mph (after attempting to brake), hitting an oncoming car at a relative 60mph because you swerved instead, or landing upside-down on top of the pedestrian because you tried to swerve and brake?

As I said, it was something I read in an article. But I agree with the author that it's important to think ahead about the ramifications. The first time a bad choice is made and it gets national attention, it will become a bigger issue.

...but the way to think ahead is to program simple risk-minimisation rules into the AV - which I strongly suspect will boil down to "maintain control of the vehicle at all cost". AVs will be much better at that decision than humans because they have all the data about road conditions, vehicle capabilities and accident survivability and can run all the models in a split second. Your/the article's problem is to overthink it and start dragging in morality issues from SF robot stories that are as far beyond the realms of current AV tech as quantum mechanics to a bee. (Actually, I wouldn't bet against the bee, with all those spooky emergent properties of swarms...)

The last thing you want to do is start raising any expectation that AVs are going to make moral decisions about who to save. They should slash accident rates by simply following the rules that we know human drivers ignore.

Otherwise, you're just as likely to be dealing with the situation where an AV swerves onto the sidewalk and runs down the child who didn't dash out after their ball, because it thought the ball was the child, or drowns the celebrity patron of the society for cute kids and fluffy animals by driving her into the river when the infamous kitten strangler Rotter "Lofty" McRotterface jumps in front of her car.

 

[_Refurbished_]

"....Uber doesn't have that kind of power, nor does our government. Maybe in China..."

You seem oblivious to the manner in which the tech companies have broken the existing rules of the game.

As for "China" your govt has more power over its citizens than China, improsons more if it's citizens than China and spies on them to a level than China can't match- all enabled by the tech industry.

"Uber" is a facade. Ditto for its counterparts. Together they will carve up this pie. Look and grasp that VC tech and financial sector giants are the drivers here. They rule. That's what that study tells you. Read it and stop taking about China. This is going on here at home.

I hear Amazon is selling tinfoil hats w/ free two day shipping!

 

[kdarling]

That's likely to happen whatever. If the authorities start to apply higher safety standards to AVs than they do to humans, it will kill the concept because they will never be good enough and end up driving everywhere at 20mph and stopping every time a leaf blows across the road.

People already expect more from AVs.

The first reaction that most people have to AVs, is that they'll be safer because a computer can make faster and better decisions.

This hope, is of course, dependent on whether other vehicles do what the AV expects. Look at that recent Google car accident where its computer knew it had right-of-way, so it continued to pull out into a lane even though a huge city bus ignored the rules and bullied its way onward in that lane. Result: bus runs into car. At a higher speed, the "safer" AV might well have killed its occupant.

So I think AVs will do best if/when they're the only kind of vehicle on a particular road... and all share the same set of rules. (Imagine if one AV type was more of a bully.)

I'd debate that. I suspect most people would either swerve without looking (and possibly cause even more damage) or just stand on the brakes or, worse, try to do both and flip the car.

I think it depends on how much warning you have. I had a deer dart in front of me really close once. All I could do was slam on the brakes and barely miss him. That corresponds with your proposed simplistic AV response.

OTOH, my father had another car run a stop sign at high speed, but he could almost tell that the guy was going to, so he had just enough time to swerve into a pole instead of getting hit. It was a instant choice of which seemed the better accident. I'd say that it'd be nice if an AV was able to notice if other drivers have run a light, and take evasive action.

I've also known quite a few people who swerved to avoid a squirrel or cat, and ended up against a guard rail or in a ditch. Heck, my mom drove us into a creek one time to avoid a dog. And we kids thought was okay of her

As for "flipping over", modern cars rarely do that, because of computer anti-skid side control.

...but the way to think ahead is to program simple risk-minimisation rules into the AV - which I strongly suspect will boil down to "maintain control of the vehicle at all cost".

Interesting thought. I think that's not enough. I think taking (possibly risky) evasive action will often be necessary when other drivers (and yes, pedestrians) disobey the rules.

E.g. as with my father, hitting a pole was a better choice than getting T-boned with a high likelihood of major injuries.

The last thing you want to do is start raising any expectation that AVs are going to make moral decisions about who to save. They should slash accident rates by simply following the rules that we know human drivers ignore.

All I'm saying is that people will expect more than just braking. I agree that some of my scenarios are no-win and cannot be programmed for.

Nevertheless, the first time a child is killed when all the car had to do was swerve into something else, I think there's going to be a public backlash. Do you not think so?

Interesting debate.

 

[617arg]

...

The last thing you want to do is start raising any expectation that AVs are going to make moral decisions about who to save. They should slash accident rates by simply following the rules that we know human drivers ignore.

Otherwise, you're just as likely to be dealing with the situation where an AV swerves onto the sidewalk and runs down the child who didn't dash out after their ball, because it thought the ball was the child, or drowns the celebrity patron of the society for cute kids and fluffy animals by driving her into the river when the infamous kitten strangler Rotter "Lofty" McRotterface jumps in front of her car.

The thing is that this is exactly the conversation that is happening, and needs to continue. The tech is here and it's only going to get better. Full autonomy is within reach (probably less than 20 years), and programmers and law makers will have to determine how to weigh "decisions" the car makes. This is different than human error precisely because the actions must be programmed.

Here's a link discussing this matter: https://www.technologyreview.com/s/542626/why-self-driving-cars-must-be-programmed-to-kill/

These are real ethical dilemma's. At what point would you feel comfortable that a self driving car decided to sacrifice the occupants? When it senses a 70% chance that more people will be killed if it hit the primary obstacle? An 80% chance? 90%?

 

[theluggage]

Look at that recent Google car accident where its computer knew it had right-of-way, so it continued to pull out into a lane even though a huge city bus ignored the rules and bullied its way onward in that lane. Result: bus runs into car.

Well, yes, that's why Google hasn't even announced a product yet, why Uber is only letting customers "ride along" with a human driver and technician and why Telsa are getting flak for prematurely releasing autopilot. Sounds like they need to increase a risk factor parameter somewhere in their code...

OTOH, my father had another car run a stop sign at high speed, but he could almost tell that the guy was going to, so he had just enough time to swerve into a pole instead of getting hit. It was a instant choice of which seemed the better accident. I'd say that it'd be nice if an AV was able to notice if other drivers have run a light, and take evasive action.

That's quite a different scenario c.f. someone running in front of you. So its "risk of high-speed T-bone" vs "risk of hitting pole at 30mph (assuming no risk of hitting any pedestrians near the pole)". I suspect that, in that scenario, even the cold equations would have chosen the pole.

Now, in a moral conundrum, there'd be a pregnant woman standing behind the pole and the driver of the oncoming car would be a convicted child molester... but in the real world neither you or the AV are likely to have that that level of information.

Of course, the real trick there is to replicate your father's intuition that the car was going to run the stop sign.

Nevertheless, the first time a child is killed when all the car had to do was swerve into something else, I think there's going to be a public backlash. Do you not think so?

Yes. There's going to be a public backlash the first time an AV kills an "innocent", even if the accident was totally unavoidable (the press won't initially mention that the kid vaulted over a 4' fence to get on to the highway). There will be a public backlash the first time an AV drives its owner into a tree to avoid a dog. There will be a public backlash the first time an AV drives over a dog and traumatises the 6-year old watching from the sidewalk (and if the alternative was to swerve and hit the kid instead the press won't let that stand in the way of a good story). Oh and by "public" I mean "astroturfing by every big vested interest without a horse in the AV/EV race". Fortunately, Google, Telsa, Ford, Uber are big enough and ugly enough to fight that battle for themselves. The honest defence will be to point to overall figures and potential lives saved (although a few contrived "my AV saved my life" stories will help).

The really tricky decision with AVs is much more mundane: how anally should they stick to the rules of the road, and how much should they prioritise safety? Riding in an AV which religiously drives around at 5mph below the posted limit, doesn't pull out of a junction until there's an absurdly large window and slams on the anchors if there's a pedestrian within a 200' radius is going to be pretty frustrating (and you're going to get rear-ended a lot). Where I live we've got a lot of pointless 20mph limits for no rational reason (the planners have encouraged builders of new supermarkets to "sponsor" 20mph zones as a political gesture to address concerns about additional traffic. Go figure). Most people ignore them (I slow down, but if you drive them at 18mph you'll have a string of cars trying to drive up your tailpipe, which is a hazard in itself). Then, I can think of plenty of blind road junctions where, if you waited until you were absolutely positive that nothing was coming, your skeleton would be found festooned with cobwebs in the rusty hulk of your car 10 years later - you have to take a small risk of T-boning every time you pull out.

Of course, AVs should religiously stick to the rules and allow a generous safety margin... that's where most of the lives are going to be saved. But, if they do that, will anybody want to ride in them?

 

[MillieWales]

This is still a testing phase. That's why it's only rolling out in one city right now too.

I know, that doesn't stop it being funny though. I'm not sure two people in the car are enough, they should have to have 5 and the passengers ride in the boot (or trunk to our American friends)

 

[iOSFangirl6001]

I really can't believe that Uber is able to do this. I can't imagine that they have done more research and development then Google or Tesla.

Also Ford saying that they will have a level 4 autonomous vehicle Shipped by 2021 is crazy. They aren't even doing real world testing yet (if they are it's minimal).

I wouldn't trust Uber or Ford with this. Tesla has the data of millions of miles of roads with mapping and real world conditions.

I really feel like Uber is going to ruin this for all of the other companies by doing this so quickly, if there is a fatal accident and it's the vehicles fault this will hold everything up for everyone else.

I just can't imagine that Uber has been able to do more R&D then google who has had their level 4 car doing testing for years.

Although I can't say I'm a fan of Uber or autonomous you'd be surprised how much testing has been done also what's NOT covered in this or some articles on MR about it

IS Uber has a few partners including 2 colleges they aren't doing it solo also they're supposedly using some notes, data and other outside resources from "experienced knowledgeable sources already having success in the field"

 

[Mildredop]

It will completely change our culture. It's going to have the biggest impact on our lives since the introduction of the cell phone. There is a crazy amount of stuff that's going to change in the next 20 years, due to self driving vehicles. Here's a few examples, but I could go on for hours...

Want a new iPhone? Send your car to the Apple Store and they will drop it off in the car. On the way home, bring back pizza for the family. Oh, we needed more toilet paper, I'll make sure to send the car to Costco on the way back to pick some up. No delivery fees, no waiting = endless possibilities.

I live in California and my best friend lives in Utah. Sleep in my car bed and wake up in Utah.

I can go on and on with this one...

But I can do all this already. Supermarket deliveries, pizza deliveries etc. As for sleeping in your car bed... that's called a taxi.

I'm afraid your examples haven't convinced me yet.

 

[_Refurbished_]

But I can do all this already. Supermarket deliveries, pizza deliveries etc. As for sleeping in your car bed... that's called a taxi.

I'm afraid your examples haven't convinced me yet.

Currently, you need to pay for all of these services. How much is it to pay for a taxi from Cali to Utah, with a recliner?

Pizza delivery isn't instant. You're at the mercy of the driver. I've waited two hours for delivery before. Self driving pickup has no time delay and ensures you're not on a long delivery route that causes your food to become cold and soggy.

Supermarket deliveries will always be expensive because it still requires a human to shop for you. Until there is a robotic supermarket, labor fees will restrict the service.

Wealthy people already have the luxury of delivery services or personal assistants. Self driving cars will bring this to the masses.