Home / General / Uber’d

Uber’d

/
/
/
3 Views

Uber’s postmature attempts to change its toxic culture are off to a great start:

Uber Technologies Inc [UBER.UL] board member David Bonderman apologized on Tuesday for a comment he made about women at an all-staff meeting at the ride-hailing firm during a discussion of how it would transform itself after a probe into sexual harassment at the company.

The ill-timed remark came as Uber board member Arianna Huffington was informing employees of the importance of increasing the diversity of the board.

“There’s a lot of data that shows when there’s one woman on the board, it’s much more likely that there will be a second woman on the board,” Huffington said, according to a recording of the staff meeting published online by Yahoo Finance.

In response, Bonderman said: “Actually, what it shows is that it’s much more likely to be more talking.” Huffington responded with a laugh.

Hahahaha, just look at that mouthy broad Kamala Harris amirite?

While we’re here, let’s return to the “no path to making a profit” problem that Uber has in addition to its massive sexism problem. A few commenters expressed variations of this view:

I always thought it was openly acknowledged that Uber’s intended business model was to subsidize fares (via burning through VC money) until they achieve monopoly status, then jack up rates back to what taxis cost now. Is that even a controversial statement?

[…]

Future entrants will face a rather significant barrier to entry: access to billions in VC cash to underwrite your subsidies.

So from that standpoint, Uber’s business model makes more sense. While the first-in-the-door examples of this business model are able to drum up funding for this scheme, it’s going to be harder for future competitors to do the same when Uber already exists.

But it should be obvious that the Standard Oil model won’t work. There are two fundamental problems facing Uber’s potential profitability:

  • The inherent costs of entry are low
  • Demand for cab service is highly elastic

The circle just can’t be squared. The reason it takes a lot of venture capital to compete with Uber is because it’s massively subsidizing riders and drivers. But if you assume that Uber can charge market rates and still make a profit, then it would be easy as pie for a competitor to enter the market. To assume that market rates are profitable and that it would be extremely expensive to enter the field is a Mnuchinesque mistake. If you share my assumption (and, apparently, the assumption of the companies themselves) that they would hemorrhage riders if they charged market rates, then it doesn’t matter if Uber achieves quasi-monopoly status — it’s still losing money.

And the problem is even more acute in smaller, less dense markets than NYC and SF. Some of the problems I identified — cars in poor condition, opaque pricing, forced ridesharing — are regulatory failures and/or cases of companies being incompetent. But there’s a reason why outside of the biggest cities cab service tends to be unreliable if it’s available at all outside of transportation hubs and major hotels. Basically, in cities where people don’t take cabs for most trips you face the choice of making it worth their while for drivers to stay on the road when they don’t have passengers, or you’re going to have cases where people want cabs and can’t get them. Given that demand is particularly elastic in places where people generally have cars and rarely use cabs, cab companies are probably going to choose the latter. But this creates a downward spiral — if you need a cab and can’t get one, you’re even less likely to use a cab going forward. I don’t see anything about Uber’s technology that solves this fundamental problem.

FacebookTwitterGoogle+Share
  • Facebook
  • Twitter
  • Google+
  • Linkedin
  • Pinterest
  • Dan Mulligan

    Uber has said that self-driving cars are an existential issue for it. Unless you think that is happening soon, this may be one of the greatest scams of all time.

    • Murc

      The guys running Uber seem to be the kind of techbros who aren’t smart enough to run a scam, because they genuinely believe they’re gonna be the masters of their domain once the technology gets there.

      I mean. There’s a non-trivial chance they might be right. There’s a ton of money and expertise being thrown at self-driving cars, and that’s not gonna stop anytime soon, and it seems pretty likely that any regulatory hurdles thrown in their way once the tech gets there will be leapt because we do love our tech in this country.

      And if they are, they’ll look really smart.

      If they aren’t, they’ll look really dumb.

      • Given that the intelligence behind self-driving cars will be repurposed to many other jobs, I don’t think there will be all that many people left with the money to spend on an Uber cab.

        • Exactly.

          Even your job can’t be automated it can still be “collateral damage” when the economy collapses from lack of demand.

      • addicted44

        Even if self driving cars happen, it’s unlikely to be Uber who benefits from it.

        In terms of coverage, and partnership, Lyft has much better partnerships with companies like GM and Alphabet that would give them a huge head start on capitalizing on it.

        And in terms of technology, Uber’s sleazy tactics have stalled a LOT of the progress they’ve made. Openly stealing Google’s IP, destroying their partnership with CMU, and putting self driven cars on the road well before they were ready were all sleazy Uber like tactics which are now catching up with them and have probably put them way behind their competitors.

        So yeah, Uber is unlikely to be the beneficiary of self driven cars either.

    • Peterr

      To a certain degree, when I hear Uber saying this, I hear them saying to auto companies “if you want a big pilot project and an opportunity to get your cars some really great publicity, give us your cars for free and we’ll put them on the road.”

      Think of it as asking for an in-kind venture capital contribution, where Uber doesn’t even have to give back stock.

      • keta

        That’s what I hear, too. I’ve heard this dynamic called sinner…sinur…synergy!

      • A lot is being made of Uber’s trying to be first to market with robot cars. This seems silly if the game is robot cars. But the game is not robot cars. The game is owning, and I mean literally owning, the A.I. behind robot cars. Whoever patents that will own the world. This is why billions are being poured into the research.

        • Whirrlaway

          Could be you! Make A.I. from home! Get started now! https://opensource.com/article/17/5/python-machine-learning-introduction

          • Who knows! Truth is stranger than fiction :-)

        • Just as well that the IP model of ARM, which designs the processors that go into cars – the latest top model, the A75, is designed for machine learning – is open licensing. This isn’t altruistic but a clever survival strategy. It keeps out rivals, even Intel, and the Chinese will cooperate as the fees are low and very good value compared to DIY piracy.
          So your AI monopoly would have to be all in the software copyright. That’s pretty weak.

          • And yet how much money have Uber and Alphabet (Google) spent in fighting over who owns the IP? The hardware is just the hardware. The real money will be in owning the software platform that ties the hardware together.

            • addicted44

              They haven’t fought over the IP.

              Kalanick was an idiot who openly conspired with Google’s lead autonomous car guy while he was working for Google (and Google were stupid enough to pay him millions while he literally worked for their competitor), and then he moved wholesale over to Uber setting back Google’s efforts big time.

              The IP fight is probably little more than payback (and a threat to future employees that they can’t obnoxiously abuse the freedom Google gives them to basically work for a competitor on Google’s dime).

        • ggrzw72

          Whatever the game is, I struggle to understand how having the largest market share among “ridesharing” apps the day before self-driving cars go into production helps to win it, let alone how it’s so essential that’s it’s worth spending $2 billion dollars on.

    • Richard Hershberger

      I don’t believe self-driving cars are coming any time soon, if by “self-driving” we mean no controls at all. If we mean “self-driving, except when it isn’t, at which point the driver takes over” then we already have that. But that doesn’t help Uber.

      The next problem is that even if we stipulate to true driverless cars in five years, why should we believe it will be Uber–an ap company with pretensions–will be the company that gets us there? This seems wildly implausible on its face, even apart from Uber’s dysfunctionality.

      Finally, even if we have true driverless cars, if Uber doesn’t have a monopoly on the technology, what is to prevent Lyft and everyone else from entering the market? If Uber still has bucketloads of cash at that point, this would give it an advantage. But even so, I don’t see the barrier to entry that would let it lock up markets.

      I don’t think this started out as a scam. I think it started out as yet another poorly thought out gee-whiz! tech startup. Then it got hot with the funding crowd, followed by the realization that it had to figure out what it was going to do now. That being said, when the whole thing crashes and burns, I expect that the top players will come out with healthy bank accounts.

      • Peterr

        The next problem is that even if we stipulate to true driverless cars in five years, why should we believe it will be Uber–an ap company with pretensions–will be the company that gets us there? This seems wildly implausible on its face, even apart from Uber’s dysfunctionality.

        This is where my comment above comes in. What Uber can offer is not the cars themselves, but an opportunity for a builder of driverless cars to get ahead of the pack and put a bunch of their cars out for everyone to see, all at once. “Your competitor will be selling them one at a time, but you’re gonna get a bunch of press when we announce that we’re flooding five mid-sized cities with them. You’ll get more press when the cars appear and every local tv station does a ‘live from a driverless Uber’ report. All the while, your competitors will be eating your dust. You want to be first? Sign up with us.”

        • Richard Hershberger

          Then why is Uber dumping buckets of money into driverless car research? If the business model is for it to buy driverless cars from someone else, it should be partnering with some company positioned to be the first to put these cars on the road. The course it is following is to own the tech itself.

          • Lost Left Coaster

            Or (and obviously this is just speculation) Uber might be using all this “look at all the research we’re doing!” just for the sake of publicity, and to mollify investors. Buying time, in other words, until they can get their hands on these cars one way or another.

            • ggrzw72

              But if that’s the case, then doesn’t Uber’s entire budget consist of (1) overhead and (2) setting piles of money on fire to defraud investors? That’s the kind of thing that the investors will probably notice eventually.

      • Scott Lemieux

        The next problem is that even if we stipulate to true driverless cars in five years, why should we believe it will be Uber–an ap company with pretensions–will be the company that gets us there? This seems wildly implausible on its face, even apart from Uber’s dysfunctionality.

        Exactly. There’s no reason to think Uber can beat Google or actual car companies to the technology, and if they don’t whoever invents it can just use their own app.

        • Captain Splendid

          Uber doesn’t need to beat Google or the car companies, it just needs the driverless cars so it can dispense with the drivers.

          Speaking of which, we’re going to get those cars sooner rather than later. Once a couple of large companies decide to bite the bullet and transform their fleet, it’s over.

          • lunaticllama

            What is Uber’s competitive advantage then if it doesn’t own the car tech?

            The application itself has been replicated by competitors and would be trivially easy for any company to stand up at this point with its own fleet of Google cars in this driverless car future.

            • Captain Splendid

              Theoretically, not much. But plenty of tech companies have made hay on such slim disparities. And, as mentioned elsewhere in this thread, a good chunk of revenue will come from selling customer data.

              • Captain Splendid

                (stupid edit window)

                Also worth mentioning that the future envisioned by these companies is one where individual car ownership no longer exists, replaced by Uber and Lyft and so on. That’s a pretty huge market.

                • lunaticllama

                  That seems profoundly unlikely. I know many people who couldn’t give up their car/truck, because they use them for accessing outdoor activities and I live in an urban area that should be ground zero for such a business idea.

                • Scott Lemieux

                  Also worth mentioning that the future envisioned by these companies is one where individual car ownership no longer exists, replaced by Uber and Lyft and so on.

                  Given how most American cities have developed, this is insane.

                • Ithaqua

                  Yeah, like I’m going to rely on Uber to get my kids to school on time *every* morning, leaving between 7:55 and 8:00 so they can sleep in as late as possible.

                • Richard Hershberger

                  “Yeah, like I’m going to rely on Uber to get my kids to school on time *every* morning, leaving between 7:55 and 8:00 so they can sleep in as late as possible.”

                  You are my twin, except we leave half an hour earlier due to an insane start time mandated by the district’s cost cutting on buses.

              • BiloSagdiyev

                And don’t think there won’t be loud, inescapable TV ads inside the cars. They can sell eyeballs

                • Captain Splendid

                  That part’s easy. Going camping for the weekend? Specify vehicle type when you place your order. Want to keep it around in order to go mudding/exploring during that weekend? Pay extra.

                • Richard Hershberger

                  “Going camping for the weekend? Specify vehicle type when you place your order”

                  How does a driverless vehicle with no controls handle dirt roads through the woods?

                • liberalrob

                  How does a driverless vehicle with no controls handle dirt roads through the woods?

                  Well, that is a detail, and I am not a detail man. –Will Rogers

                • weirdnoise

                  How does a driverless vehicle with no controls handle dirt roads through the woods?

                  That’s been worked on by a number of groups for some time now.

          • A significant part of Uber’s business model is that it pushes most of the operational costs off to its drivers. Not sure how getting rid of the drivers, but then incurring all the operation costs is going to suddenly make the company super-profitable.

            • Captain Splendid

              The idea is that an on-demand driverless car fleet will be far cheaper and more efficient than the alternatives. Whether that’s true or not will be tested soon enough.
              And as I said upthread, the Ubers of the world are essentially trying to replace the consumer car market which is a volume business where the profit per ride matters less than there actually being a profit.

              • Upthread, too, I comment that the real game is owning and patenting the A.I. behind robot cars, which will be worth trillions. If Uber does that, it will almost certainly transform itself from a ride-share company to an A.I. company overnight.

              • searcher

                The problem is that if you want to operate a driverless car company you need … boots on the ground.

                Sure, the car drives itself. But who cleans the vomit (and other fluids) out between fares? Who changes the tires when they start to get bald, checks the oil, and whatever else needs to be done so that your ride hailing app isn’t just sending a flaming poo-filled wreck to your current location?

                If Uber really wants to be positioning themselves for self-driving cars in the next five years, they should be buying up garages around the world from which they can service a fleet of self-driving cars.

                • Captain Splendid

                  Lot cheaper and efficient to do maintenance of hundreds of cars per depot than everybody taking their car individually to the mechanic/spending weekends cleaning the car.

                  As for the garage thing, brick and mortar space is dying. Easy to snap up useful locations on the cheap.

                • so-in-so

                  In the 1930s companies figured out that routine maintenance and replacement by time in use saved money in the long run “checking” parts like tires or windshield wipers. Now there are already cars with sensors that check the tire pressure and Presumably you would need some kind of interior sensors to determine that the car needed to serviced because of issues left by the last passenger.

                • If you have robots smart enough to drive the cars, you have robots smart enough to clean and maintain them, too.

                • SIS1

                  “If you have robots smart enough to drive the cars, you have robots smart enough to clean and maintain them, too.”

                  IN a self-driving car, the car itself is the “robot”. A machine to conduct basic maintenance would be a whole different engineering issue, which is not even about just the AI, but the mechanics of the thing and its dexterity and the nature of the sensor array.

                • IN a self-driving car, the car itself is the “robot”. A machine to conduct basic maintenance would be a whole different engineering issue, which is not even about just the AI, but the mechanics of the thing and its dexterity and the nature of the sensor array.

                  Sure, if you think that robot cars, especially cabs, will be cars of today with a robot driver. But they probably won’t. They will probably be designed to be cleaned and maintained by robots.

                • so-in-so

                  IN a self-driving car, the car itself is the “robot”. A machine to conduct basic maintenance would be a whole different engineering issue, which is not even about just the AI, but the mechanics of the thing and its dexterity and the nature of the sensor array.

                  The documentary “Cars” told me this was possible.

                  Also, a future with no people.

                • Warren Terra

                  Sure, the car drives itself. But who cleans the vomit (and other fluids) out between fares? Who changes the tires when they start to get bald, checks the oil, and whatever else needs to be done so that your ride hailing app isn’t just sending a flaming poo-filled wreck to your current location?

                  A company already solved this with its large fleet of self-driving cars more than a decade ago.

                  That company is called “Zipcar”.

                  The cars are self-driving, because you drive them yourself. OK, I’m playing word games, but all the issues you raise are exactly the same with Zipcar, only it’s more work for Zipcar because they can’t tell the car to turn itself in to the garage to get its interior shampooed and its oil changed. All that maintenance and supervision you talk about, that’s hard with no employee in the car, is more true for Zipcar than for a hypothetical robot cab company.

                • searcher

                  Zipcar is now a subsidiary of Avis Budget Group … a company which has the physical infrastructure to maintain an international fleet of vehicles. They were purchased for $500M, or 0.7% of Uber’s valuation.

                • Warren Terra

                  Interesting that Avis bought Zipcar, but that doesn’t detract from my point. Zipcar has a huge fleet of cars they let out into the world unsupervised and containing no Zipcar employees, and they’re able to handle all the logistical problems of attending to those unsupervised cars that are, from their point of view, driving themselves around the city. Getting the logistical services supplied by Avis seems like it might be efficient in a way Uber could easily replicate.

          • Uber doesn’t need to beat Google or the car companies, it just needs the driverless cars so it can dispense with the drivers.

            It specifically needs lots of driverless cars owned by other people who want to farm out their cars while they are at work or otherwise occupied, but still pay for the gas and maintenance.

        • Shantanu Saha

          It’s difficult to see how Uber, a company which prides itself on going outside the law when it comes to developing a market, can ever stand up a driverless car. Let me point you to this article:

          Ask a company like Audi why it doesn’t have a self-driving car at dealerships today, and it will tell you it’s not because it can’t build one. It’s because it can’t build one that’s certain to have overcome the thousands of tiny points of error 100% of the time, whether that means technological errors made by the car or psychological errors made by the driver in it. That’s a good mentality for a car company to have! And yet, it’s also the sort of mentality that a skeptic might say has kept the auto industry stuck using fossil fuels. Google offered a promising compromise with Firefly. It had all the disruption of Silicon Valley boldness, with none of the danger of irresponsible, overconfident development.

          If there is one thing that perfectly describes the Uber menschen, it’s the term “irresponsible, overconfident development.” If Uber were to try to roll out an autonomous car, it would sink like a stone under the tide of class action suits once a half-dozen people are killed in accidents either riding or being near their vehicles. There’s a reason why Google has not yet put driverless cars on the street without professional drivers. It wants to make sure that it doesn’t get buried under lawsuits stemming from accidents involving its cars, and is willing to wait a long time to perfect the tech. This is why Tim Cook of Apple recently dispelled any rumors of an Apple Car anywhere near the horizon. The established players are either willing to stay in for the long haul, or not willing to jump in before the tech is so established it’s got grandchildren.

          • Captain Splendid

            I’ve never been entirely convinced by these “driverless cars need to be perfect” arguments. Safety-wise, a driverless car only needs to better, on average, than a human driver, and that is a lot easier than it sounds.

            Once the point spread between the two is large enough, the insurance companies will then make it more expensive to own a regular car and it’s all over bar the shouting.

            • Richard Hershberger

              The problem I see is not accidents, but daily non-standard situations. If we are talking about true no-controls driverless cars, the sense I get is “We’ve solved 95% of the problems. How hard could that last 5% be?”

              Here’s a scenario that happens to me routinely. My commute is mostly over secondary roads with one lane in each direction, mostly with a double yellow line down the middle. Come trash day, which can be any day of the week depending on which part of the commute we are talking about, it is not uncommon for me to find myself behind a trash truck. The way everyone handles this is to carefully nudge to the left to take a peek, followed by a quick and thoroughly illegal pass around the truck. The trash guys will often helpfully make hand signals to warn you of oncoming traffic.

              So how does our driverless car handle this? We are routinely given pious assurances that they would never dream of breaking a traffic law. So does this mean that a large stretch of my commute will be traveling at 3 MPH?

              I’m not saying that there is no solution to this scenario. But I never see any of the enthusiasts talking about stuff like this or the innumerable similar scenarios.

              • Shantanu Saha

                I suspect that close to 95% of the cost and effort of developing driverless cars will be devoted to ironing out that last 5% of edge cases.

                • hypersphericalcow

                  There’s a saying the for AI, the first 90%is easy, the next 9% is hard, and the last 1% is impossible.

                  Unfortunately for self-driving cars, the last 1% will leave a lot of dead bodies.

                • SIS1

                  “Unfortunately for self-driving cars, the last 1% will leave a lot of dead bodies.”

                  As opposed to the tens of thousands of bodies each year now?

                  If automated cars led to a 4000 deaths a year, that would mean a 90% drop from current levels of a regime of human drivers.

                • If automated cars led to a 4000 deaths a year, that would mean a 90% drop from current levels of a regime of human drivers.

                  And if automated cars led to 44000 deaths a year, that would mean a 10% increase from current levels of a regime of human drivers (possibly involving new and exciting ways to die!). Do you have any reason to think that your hypothetical is more likely than mine (without the parenthetical)?

                  One of my reasons for thinking that mine might well be more likely than yours derives from some of my experiences in the robotics community between 2004 and 2013. Notably, at the Robotics: Science and Systems 2006 summer conference, a keynote speaker from Japan announced that the Japanese government and all the top Japanese firms involved in (or adjacent to) robotics (with the signal exception of Sony…) had as a national high-priority goal the robotic provision of 90% of all elder care by 2016. No one there (and an awful lot of the top people in—what seemed to be—the most important subfields of robotics, including autonomous vehicles, were present and schmoozing both among themselves and even with superannuated topologists like me) gave any indication that they thought this was at all unlikely. (And I was taken in too, of course.) Yet, here we are in 2017, nowhere near that goal, nor is it even talked of (as far as I can tell, though I’m now well out of that loop).

                  The response to Fukushima was of course a big setback to the morale of the Japanese robotics community (the robots that had been developed specifically for dealing with the conditions at the plant after the tsunami turned out to be completely unfit for purpose, and the first—delayed—robotic response had to be improvised with equipment and technicians from iRobot). Certainly no such setbacks could occur to the autonomous vehicle project!

                  I’m also old enough to have been a hanger-on in the penumbra of the AI Lab shortly after Marvin Minsky had (allegedly) suggested to one of the graduate students that computer vision might make a good summer research project. Again, our present-day Great Men are not going to be affected by hubristic myopia, now that the benign influence of hard-nosed venture capital has prevailed!

              • Captain Splendid

                From a tech standpoint, this particular problem is easy to solve with GPS and/or alternative routes.

                Also, when did any corporation really care about the remaining 5% when it has the other 95% locked up?

                • lunaticllama

                  Haha. You suggested above that my self-driving car could be used for camping and off-roading. How is GPS going to help identify unmarked roads or other drivable terrain?

                • Richard Hershberger

                  Alternative routes? Your solution is that any road with a trash truck on it will be avoided by any driverless car? Seriously? This is the best you’ve got?

                • Captain Splendid

                  Easy to froth when you’ve only addressed one of my points. But really, the sad truth is Big Corp doesn’t care about your edge case. Go spew your venom at them.

              • Linnaeus

                To add on to this, one thing to keep in mind is that as we reduce or eliminate certain risks with technological development, we incur new ones that (hopefully) are still collectively lower than the prior risks that the technology in question is supposed to address. It will take some time to get people used to that.

                For example, security is obviously going to be a major concern when you’re dealing with hundreds or thousands of autonomous vehicles that are networked by necessity. Someone, somewhere is going to try to hack that network (vehicles can even be hacked now) and eventually they will succeed. I’m not an IT security expert by any means, so perhaps this is not a serious risk and the measures being taken to deal with this will be sufficient to make this a very low risk. But that’s the kind of thing that’s not unimaginable that could give a lot of people pause.

                • cleek

                  as a programmer, i will be the last one in line for a car driven by programs.

              • efgoldman

                So how does our driverless car handle this?

                Almost all commercial trucks of all sorts use GPS now (for tracking, not directions). Can that information be integrated in such a way that the AI chooses an alternate route?

                • Richard Hershberger

                  As I noted above, but more rudely, “alternate route” means avoiding any road with a trash truck. This more or less means sticking to roads that are large enough to have two lanes in each direction. Vast swaths of the nation have just been taken off limits, at least when trash trucks are on the road (which is pretty much always).

                • aab84

                  What about those of us who are just luddites/privacy freaks and turn off any location-based or GPS device we possibly can because we don’t want a record of where we’ve been.

                  It does not seem as if people have thought through the privacy implications of self-driving cars.

              • addicted44

                It’s not that hard. Any autonomous car will have the ability to do manual overrides to handle the 5%.

                And once the benefits of self driven cars are so obvious, it won’t be very long before driving a car on public roads will basically be banned, and driving will be reduce to an expensive hobby for the rich.

                • Richard Hershberger

                  In other words, you are saying that these will not in fact be autonomous cars. But this isn’t the bill of goods being advertised to us. Google talks about a car with no driver controls at all. If your point is that this is vaporware bullshit, then we are in complete agreement.

                  This raises two points: How does this “all but 5%” car mesh with the brave new world you posit in your next paragraph in which driving is banned? And how does “all but 5%” mesh with summoning a car to your door, riding it to you destination, and sending the car on its way. That’s great if during its unmanned periods it only hits the easy 95%.

                  Or do you have the “remote call center” model in mind? Because it isn’t all that hard to think up problems with that solution.

              • SIS1

                IN response to Lee Rudolph, since this reply system won’t work for me….

                “And if automated cars led to 44000 deaths a year, that would mean a 10% increase from current levels of a regime of human drivers (possibly involving new and exciting ways to die!). Do you have any reason to think that your hypothetical is more likely than mine (without the parenthetical)?”

                Let me count the ways:

                Automated cars can’t become drunk,drowsy, or distracted.
                Automated cars can’t become frustrated, angry, annoyed.
                Automated cars can react to adverse driving conditions faster and don’t panic
                Automated cars will more accurately gage the actual distance and velocity of other vehicles near them.

                We already have cars with automated systems to parallel park and sensors to stop the car if something comes in front of them. We have these systems because computers hooked to sensors measure reality better than we do, not only because the actual sensors can be made superior, but because computers lack the editing that human brains do, an editing system created long before we had the ability to move at more than 30 MPH.

                That robots haven’t come as far as some people thought they would a decade ago has no bearing on whether we can know that automated cars, when they come, will be safer. You are talking about date for getting into market, which is a different issue.

                • Shantanu Saha

                  Automated cars can be hacked.

                  Automated cars can’t become drunk,drowsy, or distracted.
                  Automated cars can’t become frustrated, angry, annoyed.
                  Automated cars can react to adverse driving conditions faster and don’t panic
                  Automated cars will more accurately gage the actual distance and velocity of other vehicles near them.

                  Automated cars can be confused by machine vision anomalies.
                  Automated cars can fail to respond to adverse driving conditions that have not been envisioned by their programmers.
                  Automated cars can fail to respond when the actions of a human-driven car deviate from expected patterns.

                  We already have cars with automated systems to parallel park and sensors to stop the car if something comes in front of them. We have these systems because computers hooked to sensors measure reality better than we do, not only because the actual sensors can be made superior, but because computers lack the editing that human brains do, an editing system created long before we had the ability to move at more than 30 MPH.

                  And that’s all we have after the DARPA Grand Challenges were successfully completed more than 10 years ago? Doesn’t that tell you something about how hard actual autonomous vehicles are to create? AFAIK, NOBODY, including an AI assistant, parallel parks a car at 5MPH, let alone 30MPH.

                  That robots haven’t come as far as some people thought they would a decade ago has no bearing on whether we can know that automated cars, when they come, will be safer.

                  The hubris in this statement is breathtaking.

            • The Temporary Name

              Safety-wise, a driverless car only needs to better, on average, than a human driver

              No. You can blame an individual human driver, and that liability takes pressure off the manufacturer. Without without a driver failure is all on the company.

              • Captain Splendid

                True, but if the accident rate is as low as they’re hoping it to be, liability costs will just be another line item.

                • twbb

                  If the accident rate is sufficiently low enough, legislatures might step in and take even that off the books.

                • The Temporary Name

                  Yay tort reform!

                  Even if the accident rate is low, a sufficiently spectacular accident could spell doom.

                  And Uber still hasn’t made any money.

              • SIS1

                “No. You can blame an individual human driver, and that liability takes pressure off the manufacturer. Without without a driver failure is all on the company.”

                That will depend on how the laws are re-written. You can basically have a universal no-fault insurance regime in which each owner of the car remains responsible for any accidental deaths. You could adapt general product liability laws for any identifiable manufacturing issues with the vehicle.

            • cleek

              humans tolerate accidents caused by human error.

              but they will freak the fuck out if autonomous machines start killing people.

              • Captain Splendid

                Maybe. But if you can put a solid dent in the 30K fatalities per year we’re apparently OK with, the insurance companies won’t care.

                • cleek

                  the people who are expected to ride in or around them will care.

                • Captain Splendid

                  You’re saying that like insurance companies care what the little people think.

                • cleek

                  if insurance companies were the ultimate arbiters of what we can do in society, motorcycles would have gone extinct years ago.

              • dogboy

                +01 sez Loki Stormbringer

            • Scott Lemieux

              Safety-wise, a driverless car only needs to better, on average, than a human driver, and that is a lot easier than it sounds.

              This is almost certainly not true. Driverless cars will have to be much safer than human drivers for the technology to gain widespread acceptance. Whether this is rational or not is beside the point.

              • Captain Splendid

                How much safer? Even with your stricter threshold, that ain’t exactly far off.

                • Hob

                  This is hand-waving. No one in this discussion has any basis for determining what is or isn’t “far off”; you’re going on a vague feeling that you don’t think certain technical problems sound difficult.

                • Captain Splendid

                  It wouldn’t be an LGM thread without a little condescension.

            • Spider-Dan

              No, merely being better than a human won’t cut it.

              Human error is an accepted necessary evil in driving, because to eliminate it would require prohibition of driving by humans. But elimination of AI error is easy: just don’t allow AI to drive!

              In other words, preventing human error would involve savage curtailment of basic freedom, while preventing AI error requires something akin to enforcement of the status quo. The latter is indescribably easier than the former.

              • Captain Splendid

                All things being equal, I’d agree with you (and I’m certainly in no rush as I’m one of those fools who enjoys driving). But there’s a ton of other issues that have a bearing on this that I don’t think can easily be ignored.

              • SIS1

                “In other words, preventing human error would involve savage curtailment of basic freedom,”

                Driving is not a ‘basic freedom’ – and society has every right to think about a system that leads to at least 30,000 dead citizens annually.

                • Spider-Dan

                  I would argue that the overwhelming majority of Americans consider driving a basic, essential freedom. But at this point, we are debating semantics.

                  Whatever term you want to describe human driving – right, privilege, freedom, responsibility – the bottom line is that an attempt to disallow human driving writ large would make alcohol Prohibition look like an amusing trial balloon. It’ll never happen in the lifetime of anyone participating in this discussion, that’s for sure.

                • SIS1

                  I find the notion that some activity that did not even exist in 1900 in any real sense can’t possible disappear for “cultural reasons” a century and a half later due to technological change actually absurd. There are plenty of other cultural norms that have gone by the wayside as the material conditions of people change.

                  There also won’t be any need to ban human driving – the mechanism will simply be pricing, ie. to have a car, you must insure it. If you make the expense of insuring a vehicle to be driven by humans high enough while automated vehicles are cheap, given their much greater safety, then people will vote with their pocketbooks.

                • cthulhu

                  Well, yeah, smart gun tech doesn’t have much market penetration so far and suggestions of legal requirements and/or insurance manipulations are met with extreme push-back. So consider me unconvinced that people will rank public safety above FREEDOM!!

                • liberalrob

                  Nobody here seems to realize that “antique automobile clubs” are actually a thing.

                  http://www.aaca.org/

                  Also, where are the self-driving motorcycles in this brilliant future?

                • Spider-Dan

                  I find the notion that some activity that did not even exist in 1900 in any real sense can’t possible disappear for “cultural reasons” a century and a half later due to technological change actually absurd. There are plenty of other cultural norms that have gone by the wayside as the material conditions of people change.

                  You must have some pretty unpopular opinions about net neutrality!

                  In any case, what we are discussing is not the organic attrition of human drivers via lack of interest; we’re talking about passing laws to prohibit human driving. This is not a plausible outcome in our lifetime, and as long as that is the case, whether AI is merely safer than humans is a meaningless target.

                  Additionally, insurance rates are unlikely to play a significant part. As long as these cars are traveling on public roads with (other) human drivers on them, the ability of insurance companies to reduce rates for autonomous cars will have a limit, and that limit won’t be that much lower than the rate for human drivers. In other words, you can’t get a high-enough number of autonomous cars on the road to enable massive AI rate reductions when those same rate reductions are the intended mechanism for getting the cars on the road in the first place. Catch-22.

                • SIS1

                  “You must have some pretty unpopular opinions about net neutrality!”

                  Most users of the internet are casual, so this regulatory scheme (which I support) could easily go away without some mass revolt.

                  And I believe you are utterly wrong about the insurance issue. Your first claim, that autonomous vehicle penetration will be minimal, is outright wrong. Every year models come out with more automated features, which are popular, and as the population ages, more people will want more automation. People who like driving grossly overestimate the “love” people have for driving. People who like driving are like comic book fans, utterly oblivious to how their obsession is a minority one. If you gave the vast commuting population the ability to read, eat, watch videos on their phones, etc. for that time between going home and going to work (and back, they will take it. You know, just like automatic transitions now dominate because gear-heads aren’t a majority.

                  So as automated features expand and insurers change their rates to accommodate the change, people who demand to do their own driving (at a much greater risk to themselves and everyone else) will likely be penalized. And the penalty will grow with time, enough to change behaviors just like making cigarettes expensive has significantly cut into smoking rates in places like NYS. Because again, most people don’t love driving – they like having a car, but not driving.

                • SIS1

                  “Well, yeah, smart gun tech doesn’t have much market penetration so far and suggestions of legal requirements and/or insurance manipulations are met with extreme push-back. So consider me unconvinced that people will rank public safety above FREEDOM!!”

                  Last time I checked, it is not mandatory to get insurance if you own a gun. It is mandatory in almost every state to get insurance if you own a car. So, the starting field is completely different. On top of that, registration requirements for driving are also universally accepted, while registrations for guns, well, do I even need to say it?

                • Spider-Dan

                  Well, first off, the reason for the recent near-monopoly of the automatic transmission has nothing to do with convenience or “love for driving.” It is strictly because automatic transmission technology in the 21st century has advanced to the point where it is a) more fuel efficient and/or b) better performing than a manual transmission. Why do you think it took three-quarters of a century for automatics to seal the deal?

                  Secondly: if increased insurance rates for less safe cars was really an effective market deterrent, sportscars would have ceased to exist long ago.

                • cthulhu

                  “Last time I checked, it is not mandatory to get insurance if you own a gun. It is mandatory in almost every state to get insurance if you own a car. So, the starting field is completely different. On top of that, registration requirements for driving are also universally accepted, while registrations for guns, well, do I even need to say it?”

                  Wait, what? I thought your point was people care about public safety enough to accept a massive change in societal norms? Gun deaths and auto deaths are about the same at this point. There exists smart gun tech, right now, that could reduce those deaths, but so far, gun owners are not willing to trust in the tech such that their gun might not fire 100% of the time. But, sure, these same people will be willing to accept a computer driving them and their kids around at 70mph with no one at the wheel?

                  I am actually pretty supportive of the autonomous car concept and look forward to it especially as a mode of intra-city travel but if you don’t think there is a substantial resistance to such a complete change-over, you should maybe check out the comments sections on many highly visited auto websites. People like all sorts of “driver-assist” tech but are not too keen on giving up driving completely.

                • jim, some guy in iowa

                  just hope Wayne La Pierre doesn’t decide the NRA needs a new revenue stream

                • SIS1

                  test

                • SIS1

                  I hate this comment system….

                  @cthulhu

                  Wait, what? I thought your point was people care about public safety enough to accept a massive change in societal norms?

                  no, that technological change is enough, with the change being the ability to have a car without having to bother with driving. If that change is made financial beneficial, then it would change even faster.

                • cthulhu

                  “no, that technological change is enough, with the change being the ability to have a car without having to bother with driving. If that change is made financial beneficial, then it would change even faster.”

                  Your expectation that humans are generally “rational actors” is cute.

                  And the size of the performance/pseudo-off road/luxury/etc. car markets would argue that owning a car is more than just about getting from point A to B. If you are only ordering a ride, certainly you could have luxury trim lines but, if private ownership remains an option, you can bet people will still want the first two with driver control an option. You would have to forcibly regulate them out of existence – the market would save them otherwise (consider jet skis, speedboats, etc.)

                • SIS1

                  “Your expectation that humans are generally “rational actors” is cute.”

                  No such belief is even mentioned. Most people don’t actually like driving, unlike what driving fans claim.

                  “And the size of the performance/pseudo-off road/luxury/etc. car markets would argue that owning a car is more than just about getting from point A to B.”

                  Given that that market is a minuscule part of the whole car market, I think it backs me up completely.

                • cthulhu

                  “No such belief is even mentioned. Most people don’t actually like driving, unlike what driving fans claim.”

                  “Given that that market is a minuscule part of the whole car market, I think it backs me up completely.”

                  Please provide evidence for what you are claiming. For example, the Ford Mustang had 74% growth in sales last year…in China.

                  Apparently the auto manufacturers will gladly give up profits due to aggressive market segmentation to serve their betters at Uber/Lyft/etc. Got it.

            • Warren Terra

              The liability situation is completely different. If I tragically run down your grandma, the circumstances are unclear, I doubtless made a bad decision in the spur of the moment, and all you can hope to get from me is my insurance coverage ($250K?) and maybe also my assets (much less).

              Now, consider what happens when the robot runs down your grandma. It had sensors galore all recording – perfect recollection of nearly perfect information. It killed your grandma not because of a spur-of-the-moment error in judgment, but because its code always dictated that in the right circumstances it would kill grandma. And the responsible party has put tens of thousands of these killing machines on the road, and has deep pockets.

              Maybe the law could be written that will manage to limit the car maker’s or marketer’s liability – but right now it appears they have incredibly deep poskets and will be absolutely filleted for unleashing killer robots, no matter how rare the events.

              • sibusisodan

                Oh, goodness. We’re going to be able to choose our vehicle AI based on our preferred solution of the trolley problem, aren’t we?

                • Warren Terra

                  There are versions of the trolley problem that absolutely could happen. A car could have to decide whether to swerve, lose control, and injure its driver in order to avoid slamming into a kid or three. The moral decision is almost certainly to injure the driver – who paid for the car. Will the driver pay extra to have their life valued more than passersby?

                • COnrad

                  “Would you like the Moral car or the Monster car?”

                  I’m betting the ‘save driver at all costs’ models will be the runaway bestsellers.

            • Gwen

              Two points…

              1. Anything less than perfection will be rejected by some consumers. Compare to “smart guns” or even just dumb trigger locks. A lot of gun owners care more about their guns being reliable “when I need it” than the risks of accidental shootings. Arguing that smart guns means fewer dead kids will not persuade them unless the system is 99.99999% flawless.

              A lot of drivers will insist they are safer than “some damn computer” even if statistics prove the opposite.

              2. Liability. If a human driver crashes a car it is their problem. If a self-driving car crashes it is the manufacturer’s problem.

          • This is why Tim Cook of Apple recently dispelled any rumors of an Apple Car anywhere near the horizon. The established players are either willing to stay in for the long haul, or not willing to jump in before the tech is so established it’s got grandchildren.

            The technology behind robot cars is not all or nothing. The A.I. developed through this research will be introduced to the market long before robot cars ever become commercially viable. Siri will become more intelligent, we’ll see more robots in service applications, and so on. By the time the A.I. advances to the robot car level, we may not even be using cars anymore. But the A.I. will be there.

            • This doesn’t really match with what I know of applied AI research. Although there’s generalized technology that can be used for various applications (like Siri or self-driving cars) there’s also a great deal of application-specific technology that has to be developed. Most of what’s common between Siri and self-driving cars is already well-established tech.

              • If you are talking specialized A.I., then sure. But a generalized A.I. that can respond in real time to non-predictable events will be a tool user, just like us and many of the other intelligences on this planet.

        • erick

          Exactly, when we had a thread on this before I and others made variations of the same point:

          For a viable driverless taxi service you need:

          – A fleet of driverless cars
          – a distribution network to get those cars deployed in every city you want to operate in
          – the ability to service those cars, ideally in every city you operate in
          – an app for customers to use to order and pay for rides

          Uber has the app, the most trivial and easily recreated thing.

          The car manufacturers have factories to build cars, distribution networks already in place to get them all over the world and a dealer network with locations within a relatively short distance of probably 90% or more of the population, with service departments, driverless cars eliminate the need for salesmen and the dealer infrastructure can easily be repurposed into the local offices for their car sharing services. (While we’re looking a jobs eliminated by driverless cars, with no more widespread car ownership how many car salesmen are there now? And used car dealers pretty much go away)

          Then there are also rental car companies who alread have the infrastructure and could easily convert to driverless cab services.

        • ggrzw72

          And if the goal is to win on tech, why are they setting money on fire instead of, say, investing it in tech?

      • kped

        I assume their bank accounts are already healthy, with all the VC funding they’ve gotten. The CEO is valued at 6B right now (although I imagine a lot of that is shares, but he could sell that in a heartbeat to one of the investors if he chose).

        • Richard Hershberger

          “he could sell that in a heartbeat to one of the investors if he chose”

          I don’t know that this is true. I have heard rumors of buyer’s remorse from the current investors.

          • kped

            Never underestimate rich people’s self regard. “He is fucking this up, but I can make this into the next Google” or some such bullshit.

            Or “all this company needs is some real leadership, that only our VC can provide. We’ve already sunk 19B into this, what’s 6 more for total control?”

      • prognostication

        My understanding, though, is that Uber is actually dumping a lot of money into the technology side of UAVs as well. If they get some lucrative patents out of that, it might do for them what AWS does for Amazon.

      • ap77

        Yeah, this is one of Atrios’s pet projects – the idea that “self-driving, except when it isn’t” is effectively useless. It has to be fully self-driving, no user input at all to really make a difference. And that seems to be a long time away indeed (if it ever works).

        • GoBlue72

          He’s right. Unless the cars are 100% self driving with NO driver in the car, then the model doesn’t work to save Uber. They need to get rid of the labor cost associated with their current model in order to make it pencil. To the degree that a driver still has to be in the car (and thus, paid for their time), then any cost expended to make the cars partially (but not fully) self-driving just makes it harder, not easier, to turn a profit.

        • But it isn’t useless all they need is to add remote control. if they get 99% autonomy, and if they have a call center staffed with 1000 drivers that can establish a connection to any uber car on the road and take control, they can control 100,000 cars.

          • ap77

            I think we’re a long, long way from anything like this.

            • lunaticllama

              Yeah. In the situations where you need a human driver, you probably need to be able to easily look in 360 degrees and change viewing direction immediately as human drivers do now intuitively.

              • ggrzw72

                VR technology is already fairly sophisticated, it’s hard to imagine that the technology won’t be there in time.

                • ap77

                  I find it very easy to imagine that none of this technology will be developed as, and in the time period, its advocates contend.

                • Hob

                  Like so many of the comments on this post, this is technically uninformed hand-waving. I don’t mean that as a put-down – it’s totally natural to look at a thing that seems “fairly sophisticated”, and that seems obviously related to the thing people are discussing, and conclude that “it’s hard to imagine” it won’t all be solved soon. But that’s a totally meaningless statement that falls apart once you start looking at the details.

                  For instance, in this case, what would be needed isn’t just “VR technology” in the sense of being able to give someone a video headset for looking around at stuff; you would also need 1. high-bandwidth real-time video and audio streaming with no perceptible delay and no service interruptions, 2. sufficient camera coverage to allow for visibility from any angle that a human driver could get their head at, and 3. a human operator with such superhumanly fast reflexes that they could make accurate decisions half a second after being in effect teleported from an office cubicle into the car.

                • Yeah, I can’t imagine a situation where a self-driving car can’t safely continue operating, recognizes it in time to call in a human, and the human is able to assess the situation and react usefully in time. If the situation doesn’t involve impending death or injury, the car will probably have a way to deal with it. If it does, you wouldn’t be able to get a human connected in time — and what person would take that job?

                • ggrzw72

                  @Hob, In regard 1, the comment that I responded to presupposed reliable, high-speed, data–it was about making the use of that data intuitive. And not only is 3 not a VR problem, and thus beyond the scope of my comment, but also it’s something that Tesla has already had to address–when the car senses a problem, such that it needs to end autonomous driving, the driver has 5 seconds to take over, or the car pulls itself off the road. If the car isn’t able to realize that a human needs to take over in time for the human to take over, that’s an AI problem.

                  2 is the problem that I was referring to, and I find it hard to believe, that by the time we have developed cars capable of driving 9,999 of every 10,000 miles without human assistance, we’ll still be struggling with integrating panoramic camera coverage into a VR experience and/or perfecting the robotic, binocular camera with the same range of motion as the human head.

              • so-in-so

                More to the point, you may need the driver engaged quickly. Harder to do when he’s been bored to sleep for the passed hour or so.

          • DocH

            Control gets transferred quickly and seamlessly to a remote operator who must immediately evaluate a situation that has caused the AI to tap out and respond to it appropriately? Given that getting the in-vehicle nanny-driver engaged quickly enough is a bit of an issue already, I’m going to need to see this before I believe it.

          • how_bout_never

            Except that even low (less than 100ms) amounts of latency in the connection between the car and remote operator will have large impacts (ha ha) on how hard it is to drive a car from the low bid IT sweatshop that has the contract. Adding in the amount of video data that lunaticllama notes does not make it any better.
            Remote operators located in the middle of nowhere, bad.
            Congested cell network, bad.
            Slow routing somewhere on the internet, bad.

          • Lost Left Coaster

            Stress-wise that would be like being a 911 operator times 1,000.

          • cthulhu

            Yeah, if Uber tech is going to end up dependent on the US’s craptastic OTA IT infrastructure, especially in the boonies but even in many sections of large cities, they won’t be rolling out their cars anytime soon.

      • Shantanu Saha

        Consumer-owned cars, no. But it’s plausible for some fleet uses, like long-haul trucking, where you can use the software to handle the long, boring highway driving, allowing a trained driver to nap while being ready for the edge cases.

    • sigaba

      But it’s a general belief that once vehicles become autonomous, the hailing systems to share them out will become the principle mode of transportation for most people. That’s potentially billions of dollars in revenue, daily.

      It’s such a huge pie that VCs are willing to string out people like these knuckleknobs for a decade or more. The up front losses are chump change, everyone remembers how Amazon lost money for a decade.

      • Murc

        But it’s a general belief that once vehicles become autonomous, the hailing systems to share them out will become the principle mode of transportation for most people.

        I’m not sure about this.

        Like… okay. I’d love to be able to hit a button to summon an autonomous car. Sounds great.

        But in my experience, many car owners have a huge worry about their edge-use cases. They worry about that once or twice a year they drive five hundred miles to a vacation destination, which is why they’d never consider an electric. They worry about that blizzard their town gets at least once a winter where they really do need a lot of weight and torque behind them, which is why they’ll never consider one of those really tiny ultra-efficient cars with a little engine. That kind of thing.

        With self-driving, I suspect you’ll get a lot of “what about this thing I only do four times a year that hailing will make hard, I want to still do that.”

        Having said all that, money is a HELL of a motivator. People might want to own their very own self-driving car, but if going full hailing can move their transportation costs from fifteen grand a year to five grand, you’ll get a LOT of takers.

        • sigaba

          Investors are looking at the overall trends. Auto sales in the US are cratering and the number of licensed drivers in the new cohorts coming up are declining.

          Meanwhile: a lot of people have terrible credit and can’t get loans or leases, and a lot of people are scratching their heads and trying to figure out exactly how we can fleece these people if we can’t get them attached to a loan.

          • Scott Lemieux

            Auto sales in the US are cratering

            This is ludicrously false. Sales are down a bit in 2017 but “cratering” is a ridiculous description.

        • efgoldman

          I suspect you’ll get a lot of “what about this thing I only do four times a year that hailing will make hard, I want to still do that.”

          Or you could live in a fairly congested but spread out suburb, around twenty minutes outside a small to medium sized city, with neither dependable taxi service (if there’s any at all) nor ride hailing [raises hand]

        • ggrzw72

          I think the fact that people are going want to own their own self-driving cars illustrates why Uber will never be able to corner the market. A nontrivial number of consumers are probably going to buy their own cars, and they’ll be happy jump into the self-driving livery business if the money is right. All they’ll need is an app that’ll let them hire their car out when they’re not using it, e.g. enabling them to become “Uber drivers” sans the actual driving.

      • As the saying goes, every few months someone in Silicon Valley comes up with the revolutionary idea of public transport.

        • Murc

          I feel like driverless, hailed buses will end up being a thing. I’ve heard the model described as “the bus’ll come, and it’ll get you where you’re going within X number of minutes, but you have zero say over the route and you’ll be sharing it with a load of other folks. The tradeoff is that it is cheaper than hailing your own private ride and much quicker than ‘the bus only comes through your neck of the woods once an hour.'”

          • Read Bruno Latour, Aramis.

            TL;dr turned out no one wanted to be out in the middle of nowhere stuck in a train car with a stranger and no driver around, even if the technical problems could be solved

            • lunaticllama

              There is also a security aspect. What happens when physical/sexual assaults occur in these vehicles? I think these vehicles would have a lot less appeal than driverless car advocates now believe.

              • I think you could get the benefit of an autonomous bus while still having an attendant for security and to aid elderly and disabled riders. A huge part of the advantage would be plotting out optimal routes, and no human driver is going to want to be assigned several new routes a day.

                A colleague of mine implemented an automated system for sending out cable repairpeople, and the output of that system at first was a Taylorist hellscape. The employees had an interest in a fairly efficient system since they were paid by the job, but the original algorithm so heavily favored time in transit from job to job over “dead” time spent neither at a job or in transit that it would send people across the city to avoid waiting 20 minutes for the next clump of jobs in the area. The employees hated it, and it took awhile to adjust the algorithm to give tolerably decent itineraries.

                That kind of thing is less of an issue for automated vehicles.

            • N__B

              To be fair, the Aramis project that no one wanted to be stuck in the middle of nowhere in a train car with a strange frenchman.

          • Shantanu Saha

            It’s called a Jitney.

            • craig

              CEO Travis’ talk at TED last year was all about the Jitney. It’s not a good talk.

        • sigaba

          If you’re looking for sweet pile of monopoly rents, public transport is a pretty good place to be.

          • Lost Left Coaster

            But I thought Uber was all about competition and breaking monopolies!

      • corporatecake

        The assumption that our car culture is going to change entirely the second autonomous cars and ride sharing are available strikes me as bizarre.

        • lunaticllama

          I don’t get it. The time-horizon for such a drastic change has to be a couple of decades at least. What’s the business proposition for the next 20 years?

        • Scott Lemieux

          Exactly.

        • randy khan

          In cities, the car culture is changing already (and never was the same as in the burbs and rural areas).

          I don’t have an opinion as to when autonomous land vehicles will be viable (although I think they will – people drastically underestimate how fast technology develops), but I also am confident that they don’t need to be adopted by everyone to succeed. The existence of car-sharing services and Uber and Lyft demonstrates that there’s a market that autonomous cars could address.

          • Redwood Rhiadra

            (although I think they will – people drastically underestimate how fast technology develops),

            Then where are my fucking jetpacks?

            People drastically OVERESTIMATE how fast technology will develop.

          • corporatecake

            For me it’s less about whether there’s a market for a ride hailing service with a fleet of self-driving cars. Clearly there is. It doesn’t make a difference to the person hailing an uber if it’s driven by person or technology, as long as it works.

            I’m doubtful that such services will replace car ownership any time soon. For people who live in areas where you need a car to get almost anywhere, the independence, privacy, and convenience of owning a car, self-driving or otherwise, seems very likely to outweigh the additional expense, at least for quite some time.

      • GoBlue72

        Yes but Amazon didn’t require any significant technology breakthrough to succeed. Amazon’s problem set to solve was fairly mundane: logistics (scaling up their supply chain, both in terms of # of wholesale vendors, warehousing facilities, delivery contracts), and marketing (getting consumers to shift their shopping habit preference from bricks-and-mortar to online.

        These were not obstacles that requires tech breakthroughs to succeed. Incremental breakthroughs possibly (including the advent of the smartphone which changes the dynamic of the digital divide), but mostly boring business problems.

        Uber’s problem set is wholly different as it requires solving a fairly fundamental technological barrier. A barrier which has the additional challenge that Uber was not established or organized as an automotive company (compare to Tesla) but as a taxicab app company. Thus it needed to essentially outsource its solution. This would be the equivalent of Apple needing to invent the iPhone by hiring Sony to do it.

        • CD

          Exactly. Moreover, sigaba’s “Amazon lost money for a decade” glosses over important differences.

          Uber 2016:

          Gross bookings: $20 billion
          Net revenue: $6.5 billion (excluding China)
          Adjusted net losses: $2.8 billion

          Amazon: http://www.businessinsider.com/amazon-revenue-vs-profit-2016-1

          Long story short, Amazon has had a viable business from the start and has chosen to keep profits close to zero to pursue growth in revenues. Losses there have been in some quarters, but not an Uber-ish scale. That’s not to say Amazon’s strategy has been risk-free, but its cash flow has always made sense.

          Amazon has developed, moreover, a large first-mover advantage, including big-data insight into people’s shopping, that will be hard for others to match. I switched Uber to Lyft some months ago and Lyft is just as good, if not better.

    • rea

      In other words, their business model is t get rid of all the people working for them.

      • brettvk

        I work in retail. This is *every* business model.

      • SIS1

        “In other words, their business model is t get rid of all the people working for them.”

        Isn’t the fundamental basis of Capitalism to maximize capital? A complete replacement of labor by capital is a necessary step.

  • Karen24

    I realize this is tangential to the point, but this guy made an offensive joke in response to a statement by a woman who has a popular liberal website NAMED AFTER HER. She is in the actual business of attracting eyes to articles about how conservatives are terrible. I would demand that the guy be fired for incompetence for having so little discretion.

    • kped

      He resigned today.

    • Origami Isopod

      But he was just telling the un-P.C. truth. Ariana seemed like a chill chick who could take a joke.

      BTW, this is why I am unmoved by whining from dudbros and their enablers that calling out bigoted comedy is “censorship.” Bigoted jokes are a way to push conservative memes in society with plausible deniability. If you can’t tell a joke without falling back on decades- if not centuries-old stereotypes, you are an unoriginal hack.

      • Origami Isopod

        Or what Kimberly Weisul said.

        What sets this incident apart is that the man actually appears to have paid for his dumb comment. Generally, the totally lame defense for these types of comments is that they are intended as jokes, and women should just get over it. In that framing, it’s the woman’s fault for not being able to take a joke, not the man’s fault for being insulting and contributing to a toxic environment.

        • efgoldman

          the totally lame defense for these types of comments is that they are intended as jokes, and women should just get over it.

          And of course, for “women” in that sentence, substitute all and any kinds of minorities.

          • Origami Isopod

            Oh, definitely. Women get the “no sense of humor, must be a man-hater” label, but people of color get the “angry minority” label, and I’m sure there are others I’m not recalling at the moment.

      • N__B

        I am mildly surprised he went with a joke based on that stereotype and not one based on the women-drive-badly stereotype.

        • Lost Left Coaster

          Oh, if he’d had the time…

  • Bloix

    “I don’t see anything about Uber’s technology that solves this fundamental problem.”

    The whole point of the technology is that a driver can be doing something else – eating breakfast, studying for an exam, putting the laundry in the drier – and then get a request and be in the car and on the way to a pickup in a few minutes.

    The current taxi model – lease the cab for a 12 hour shift – requires drivers to make the nut every shift. Uber gets rid of the nut so you don’t have be out on the road or at the airport taxi stand making no money for hours on end.

    • kped

      That’s not true of Uber though. They send the driver closest to you. A guy eating his cereal at home instead of being on the road isn’t gonna show up there.

      And…what the hell are you talking about with airports? The uber driver is most certainly waiting there, otherwise he isn’t getting the fare.

      • Bloix

        Scott is arguing that small markets can’t support Uber. What he doesn’t get is there is a lot of demand for part-time jobs at hours of the worker’s choosing. Students, single parents, younger retirees. The traditional taxi company can’t take advantage of this labor supply because it needs to have its cabs on the road round the clock and therefore it needs to tie its drivers to a fixed schedule. Uber doesn’t care if “its” cars spend 20 hours a day at the curb. Therefore it can rely on casual labor, which allows it to do a better job in providing on-time service than a traditional taxi company in a small market can.

        And what I’m talking about at airports is that in small markets you find dozens of cabbies waiting in lines for hours – far more than the airport needs – because for long periods during the day there are no calls. This is completely wasted time but the taxi company doesn’t care – it’s the driver who’s not getting paid, not the taxi company. With Uber that doesn’t happen. If there’s no business the driver goes home. Does the laundry.

        • GoBlue72

          Yes but your wonderfully perfect theory doesn’t actually turn out that way in practice. Initially when Uber started, and at the margins, it could source its labor supply from “regular folks with other jobs” who owned a car and were looking to just pick up some beer money in their off-hours. Actual, you know, “car sharing”. And if Uber wanted to be just a small niche company, maybe that could have worked.

          But that turned out not to actually work as a model because…reality and physics. The reality being there just aren’t that many people out there interested in picking up occasionally beer money in their off-hours instead of, you know, enjoying their off-hours. (Opportunity Cost, how the F does it work?) And Uber needs to have a steady, 24-7 supply of LABOR in order to effectively provide its taxicab service to meet the 24-7 availability of its service. And that means to secure that labor force, it means getting workers who are foregoing other work or activities for longer periods of time – in other words, a labor force that is primarily working FOR Uber (or Uber/Lyft), whether FT or as primary PT job.

          All of sudden this starts to look like..a taxicab company with same “solve for the nut” problem. Which is precisely the problem Uber has. Its reason Uber developed Uber Line or whatever they call it which strings together ride hails in order to reduce driver downtime.

          • Scott Lemieux

            Right, this is just can opener assumption. The idea that there’s going to be enough people waiting around ready to leave the house at the drop of a hat to make what may well be a negligible amount of money after maintenance and the rake to service demand reliably is silly.

        • efgoldman

          in small markets you find dozens of cabbies waiting in lines for hours – far more than the airport needs

          Bigger airports, too. Plus restrictions on who can wait in line, who can pick up, and who can only answer a call.

        • cthulhu

          The fact is that Uber DOES want extra drivers on the road to maintain quick response times. The drivers app is socially engineered with much gaming of the driver’s motivations to keep them from clocking out. But this means less fares for individual drivers and hence more dissatisfaction with the system. The one thing Uber hides more than anything else is their driver turn-over rates.

    • Richard Hershberger

      I think you have described the idealized image of the Uber driver. You have a car, right? Most of the time you aren’t driving it, right? You have spare time with nothing you need to be doing, right? So combine that unused car with that unused time and make a few bucks on the side. This is a variant on the pitch that all these gig economy startups make. The image is of someone who has a job and is doing OK, and uses the gig to make beer money on the side. This pitch meshes poorly with Uber giving drivers usurious financing deals to buy cars to drive for Uber.

      • Scott Lemieux

        And, again, this isn’t going to work very well in cities that are sprawling.

      • Bloix

        Uber doesn’t need to have every driver be this kind of driver. It just needs enough casual drivers who are willing to come out (in larger markets, maybe just for surge pricing periods) to meet demand – a core of full-time drivers, supplemented by “gig”-type drivers who are a small fraction of drivers in large markets and a higher fraction in smaller markets.

        There are three issues here:

        1) Uber owner-management is a bunch of pricks! True/False
        2) Uber exploits its drivers! True/False
        3) Uber’s business model is not sustainable! True/False

        Each of these is independent and any one could be true without having any effect on the truth of another.

        • Richard Hershberger

          OK, looking just at number 3, right now Uber loses money on every ride. For that to change, either the price of a ride has to go up, or the cost to Uber of that ride has to come down. Uber has not demonstrated the ability to do either, at least not enough to matter. Its long-term strategy seems to be to get rid of the driver to lower expenses. It is not clear that they have the technical expertise to do this, much less to do it ahead of everyone else working on the problem.

        • 1) Evidence says true.
          2) Uber doesn’t have drivers. It has contractors.
          3) Evidence says false.

        • GoBlue72

          Wow! Why did nobody think of this before! That idea totally isn’t what Uber’s business model is that totally isn’t failing.

          Quick, you need to give Travis a call with your amazing idea to save Uber.

        • efgoldman

          Uber’s business model is not sustainable! True/False

          Daughter, who writes on consumer issues for a living, thinks Uber is on a long slow decline to going bust.

        • SIS1

          “(in larger markets, maybe just for surge pricing periods)”

          NYC is one of its most profitable markets, and that “casual” driver model would not work in any way given that “casual drivers” don’t live anywhere near the main business districts that form the profitable core of the whole enterprise in the City.

          • dr. fancypants

            I take Lyft with some regularity around SF (I prefer transit, but I suffer the misfortune of having to rely on the N-Judah for my primary transit needs). One of the questions I always ask my drivers is whether they live in the city. I can count on one hand the number who have responded affirmatively–most come into the city either from the South Bay or, to a lesser extent, the East Bay.

            Which, of course, makes sense–the people who can afford to live in SF aren’t going to be interested in the small-money Uber has to offer. I suspect the same is true in other major markets as well.

      • GoBlue72

        Uber itself had a totally idealized image of who its drivers would be. Which totally idealize image went crashing on the shoals of reality. Turned out there wasn’t this huge pool of casual labor sitting around with nothing better to do in their off-hours than work as occasional taxicab drivers. That model MAY have worked for a nanosecond when Uber was operating as a niche company AND the economy was coming off a horrible financial crisis leaving vast numbers of people unemployed. But, shockingly, turns out there’s an opportunity cost for this “vast” pool of casual labor to go get into their cars to play taxicab driver (instead of remaining on their sofas watching football or hanging out at the corner pub), such that Uber has to subsidize earnings to overcome that opportunity cost hurdle.

        • ap77

          Don’t forget – the “man” also came around and said you’d need things like “insurance” if you were going to be operating as a cab. More job-killing regulations!

          • lunaticllama

            Uber will let you drive without having the proper commercial insurance. They don’t care if you maim some innocent person who is left without a viable insurance policy to provide medical care and compensation.

            • ap77

              Uber would be very happy to let you – that is part of the business model, after all. Sadly for them, many areas don’t agree with that. It’s impossible in NYC, for example.

              • lunaticllama

                Unfortunately, not all polities can pass laws that will protect the general public from companies like Uber when they hire high-paid lobbyists and lawyers to shield them from insuring the vehicles they put on the road.

            • Dr. Acula

              Indeed. An Uber driver ran over and killed a 6-year-old girl in San Francisco a couple of years ago. The driver was logged in to their UberX app and was available for rides. They went so far as to actively lie about the driver’s status at the time of the “accident”.

          • GoBlue72

            And then add to that, Uber itself started setting qualify standards for cars if a driver wanted to qualify for UberX. Now you can’t just source your labor from “regular dudes” with a 20 year old civic and some spare time on their hands. Now you need your casual labor to go out and lease a new or newish car specifically to drive for Uber.

      • Aexia

        One of the interesting things in LA is that Uber/Lyft have become the new go-to sidejob for actors & actresses in between gig. They can do a few rides, go do an audition, do a few rides, do another audition, etc.

        Part of this also is that scheduling in the restaurant industry has become significantly less flexible for the employee.

  • Origami Isopod

    “Actually, what it shows is that it’s much more likely to be more talking.”

    How much does anyone want to bet that Bonderman talks over women all the time, and if any woman tried to cut back in with “Excuse me, I was talking,” he’d be calling her a pushy bitch for months?

    • farin

      Obviously, because women are always talking so how could he say anything if he didn’t interrupt them?

  • sigaba

    Uber seems to be based on the proposition that if you’re the First Mover, literally all other aspects of the business will fall in line.

    • What my MBA degree learned me is that first movers rarely survive. The companies that make it are the ones that sit back and learn from the first mover’s mistakes. Uber is all sexy innovative and disruptive, though, so the rules may not apply.

  • Dilan Esper

    Surge pricing is part of the model too. Licensed cabs can’t do that.

    • Scott Lemieux

      How something Uber can already do will solve the problem that they’re losing large amounts of money on every ride is…not obvious.

      • cthulhu

        Uber does whatever it can to avoid surge pricing, FYI. Their motivation for surge pricing is only that there are enough cars to meet demand rapidly – not as an important aspect of their profit. (meant for Dilan)

        • Dilan Esper

          Price discrimination always starts out mild and gets more severe. That is how you do it.

          Look st MLB ticket prices for a recent example. It started with small surcharges for big games. Now it’s just like airline pricing.

          • cthulhu

            But Uber’s current surge pricing is often not minor at all. But it is not profiteering as much as a shout-out to get drivers to swarm. But as Uber is trying to corner the user market, not the driver market, they actively try to avoid surge pricing by keeping more drivers on the road and positioning them to prevent surges. They’ve gotten better at it and many of the drivers aren’t too happy because surges were a sizable bonus for them.

            Sure, once Uber gets its monopoly, it could really juice some profit from demand pricing but, for now, it is doing the opposite.

  • Bob Loblaw Lobs Law Bomb

    As to Uber’s future profitability, I think it’s a mistake to assume that the company’s long-term plans involve driving people around rather than harnessing the massive amounts of rider data that its current service is generating.

    • Well–but that information will quickly be out of date if they stop collecting it.

      • N__B

        That’s why Uber’s budget has a line item for subcutaneous near-field chips.

  • N__B
    • Ithaqua

      I’ve just decided I like Mariya.

  • Francis

    On the no-path-to-profit, what does Uber spend money on?

    The summary article on Naked Capitalism states that Uber is spending money hand over fist, but I have no idea where the outflows are going. If Uber really is just an app — a radio service for the early 21st century — then the cost of operation should be microscopic.

    In a naive sense, it would seem to me that you could have a profitable company which does nothing but collect 10% off of each ride, and use those funds to support the app and build dynamic pricing models which get enough drivers onto the street so that people will download the app.

    (eg, if density of drivers in sector x is less than y percent, increase ride price. wait 5 mins.)

    • ap77

      I think the idea is the money is being spent subsidizing fares. Even if you get to monopoly under this model, it probably isn’t sustainable, because once you get to monopoly, the idea would be that you’d THEN raise fares.

      But then you run into the problem of what happens when customers don’t want to pay the higher fares and then just stop using Uber. Oops!

      • It just occurred to me that the only way the robot-car model can work is if Uber gains monopoly status, but instead of replacing the drivers with robots, it only hires drivers who own robot cars. The ‘drivers’ still have to maintain their vehicles and pay for gas and insurance, but get paid less because they are not really driving.

        • ggrzw72

          It’s not going to be possible to shift the maintenance costs in any meaningful sense. The “driver” is going to demand maintenance plus a fair return on the use his capital (i.e. car). The shareholder/creditor (whose capital doesn’t require maintenance) is going to demand a fair return on the use of her capital (i.e. money). That’s not externalizing the cost of maintaining the cars, it’s trading one capitalist for another. Granted, the “driver” might not consider his car an “investment” to the same the degree that the shareholder considers her stocks to be, and the difference might be a source of economic profits (profits in excess of a fair return on investment), but economy profits are impossible, over the long term, in a competitive market. And given that the barriers to entry of a competing app are low, there’s every reason to believe the market for app-hailed rides will be competitive if “drivers” own the cars.

          • ggrzw72

            That should be “economic [not economy] profits.”

      • so-in-so

        That’s where the monopoly comes in – assuming they don’t suddenly run out an buy cars but still have to get places. Have to drive the old-line cabs out of business first.

        • ap77

          But that means Uber must be a necessity rather than something than is merely more convenient than other options. Are there people who used to rely exclusively on cabs to get everywhere because they have no other options but have now switched over to Uber exclusively? There probably are, but that doesn’t strike me as a large amount of people.

          If you’re living in a place that REQUIRES a car to get around, it seems to me that it’s a lot more likely that you’ll just have your own car than rely on cabs/Uber/etc.

      • Francis

        subsidy is a really dangerous word, because it means so many different things to different people. I prefer it when people describe actual cash flows.

        According to the Naked Capitalism article, each driver pays a portion of each fare to Uber. So Uber is (a) receiving money from drivers, (b) spending money on X; and (c) overall, losing money like mad.

        What’s X?

        • Spider-Dan

          X is “paying drivers” (although in reality, it’s also “paying backend employees” and “paying overhead”).

          • Juicy_Joel

            Uber has something like 14,000 non-driver employees

        • SamChevre

          X is “attempting to build a self-driving car.”

          Uber doesn’t pay drivers: drivers pay Uber. It’s a classic middleman, with a side business in building a self-driving car.

          • Spider-Dan

            Uber doesn’t pay drivers: drivers pay Uber.

            …what? I’m not sure what you’re trying to say here, but Uber most certainly does pay drivers. Why would anyone lose money for the privilege of driving other people around?

    • Hob

      “If Uber really is just an app — a radio service for the early 21st century — then the cost of operation should be microscopic”

      On the contrary, keeping a web application up and running all the time to serve an international workforce requiring real-time data is not all that cheap. Servers and storage have to be rented on a large scale with lots of redundancy, because things will fail. They have to be deployed to multiple data centers because you really don’t want an Internet traffic problem in between country A and country B to take out your whole B market. Developers and operations/IT staff have to be paid, with round-the-clock staffing.

      That’s not to say that Uber should be credited with great technical skill, but rather that “just an app” is a misleading idea that belies the amount of infrastructure behind any Internet-based business. There’s no such thing as a software system on this scale that just works with minimal support.

      • Redwood Rhiadra

        I should note that this is one way Scott is *wrong* about Uber’s future. A would-be competitor against a hypothetical Uber monopoly doesn’t have “minimal startup costs” – the costs to build the necessary data centers are MASSIVE, and anyone wanting to enter the market will have to build them.

  • trollhattan

    Uber seeming more like the Bernie Madoff of tech startups all the time. They pantsed investors who are now awaiting magic to happen in order to make their investments whole. Lotsa lick with that, especially if you believe self-driving cars suddenly make the bidnez model work.

    • Yeah. I have to believe that the whole robot-car meme is just smoke to keep the investors forking over that cash.

      • Moondog von Superman

        I hear losses were down slightly last quarter!

    • twbb

      The difference I think is after 2001 anyone investing in a tech startup should be fully expecting that it is more likely or not they will lose all their money. That’s not something that could be fairly said of Madoff investors.

  • paul1970

    The naked capitalism articles are very long but very good. summary: uber’s only (dubious) hope for profit is 100% monopoly running an essentially conventional taxi service, achieved by quitening regulators with techno-librertarian bs.

    • While still pushing the cost of operations out to the drivers. Basically, a more expensive cab company where the cab drivers own and maintain their own cabs.

  • paul1970
  • Bloix

    I got this far in that article:

    Uber’s … biggest challenge [is] the … battle between Uber’s Silicon Valley investors and local citizens over control of the laws governing the urban car service market… [I]ts investors cannot take the risk that cities respond to Uber dominance by reimposing pricing and service requirements, or other steps designed to restore meaningful competition.

    “local citizens” control!
    “Meaningful competition”!

    HAHAHAHAHA!

    Total regulatory capture, government-granted monopolies, fixed prices, exemption from federal and state labor laws, exploitation of immigrant labor, and shitty service! That’s the taxi business.

    • Shantanu Saha

      I’ve got to ask this question: how much Uber stock do you have?

      • Bloix

        Well, I don’t work for Uber, and Uber is not publicly traded, so Uber stock is not available to me. Also, I don’t buy individual stocks so I wouldn’t buy Uber even if I could.
        And I’m certainly not persuaded that Uber is a viable company. I think it might be and the arguments that it is not aren’t persuasive to me.

        Here’s what I know about Uber: I take Ubers. I started taking them when I had to be at work until after midnight many nights after metro closes (I generally commute by public transportation), and I found that they are more reliable, cheaper, and more pleasant than taxis. Maybe I’m riding courtesy of stupid venture capitalists, I don’t know. If I am, that’s fine with me.

        Here’s what I know about taxis, which I used to take: They suck.

        • paul1970

          If you read the whole series, you’d learn that you’re riding courtesy of venture capitalists who probably don’t even know whether they’re stupid or evil. There are no inherent efficiency savings in the Uber business model, nor economies of scale. The only way the company can justify its valuation, and the current enormous subsidies they offer, is if they drive out all the competition and then end your cosy deal.

        • Ithaqua

          Yes, but how long do you think the VCs will let you ride on their dime? And when that end date comes, what happens to Uber? That’s the issue that’s underlying much of this thread…

    • lunaticllama

      My city has great taxi service at a fair price. It costs the same as Uber and it is usually quicker to hale a cab than request one on Uber and stare at your phone for 5 minutes. Plus the taxi drivers know the city and how best to navigate it.

  • One of the first assignments that I have my students do in my e-commerce class is to write a paper on Uber’s business model, and discuss the gig economy and technological disruption. Interestingly, and almost without fail, the students gush over how Uber is the greatest innovation since sliced bread, and how the company has stuck it to those crappy cab companies, and how Uber allows people to make money on the side, and so on. And I’m like, ok, have you not been reading the news for the past year or so?

    Interestingly, too, they tend to discuss the downsides of the gig economy in the abstract, almost as an afterthought. It’s like, hey it’s awesome that we get cheap rides that we can schedule with our apps and people with bachelors degrees can make some cash on the side–and oh by the way no benefits.

    • ap77

      “and people with bachelors degrees can make some cash on the side”

      Are there studies about how many people actually drive for Uber on this “on the side” basis? I would guess it is a very small number.

      • so-in-so

        That was the initial pitch, I think. At least, when I first heard of Uber. Since then I gather you get dropped if you don’t take enough fares.

        The idea of an app that allows me to take a paying rider as I drive around on my daily business is certainly more appealing than becoming a default contractor driver with a minimum mandated work schedule.

        • Exactly. But apparently the hype is still strong, at least with my students.

          • so-in-so

            “It’s an app! On my smart phone! Must be cool!”

        • ap77

          Do they drop you for insufficient fares?

          I think the point is that it’s not really possible to make any money as a driver for this thing unless you’re working a full time cab schedule anyway. In places like NY, you’ve got to jump through various hoops to even be able to do this in the first place, which effectively precludes anyone from doing it other than full time.

          • Redwood Rhiadra

            Do they drop you for insufficient fares?

            Yes, they actually do.

      • rea

        Are there studies about how many people actually drive for Uber on this “on the side” basis?

        Here in Michigan we got a lot of articles a while back about an Uber driver who pursued his day job in between rides. Unfortunately his day job was “serial killer”

    • Moondog von Superman

      Is the gig economy really about “technological disruption” at all? Seems to me it’s just exploiting other disruptions — the destruction of the having-a-good-job economy. (Plus, creating demand for various service gigs, we have an expanding number of well off people.)

  • AMK

    Hard to believe they would not be profitable in major cities if they modestly raised their prices. But that would require them being OK with being a successful taxi business worth $5 billion (Lame!) instead of a $50 billion monopoly-money Silicon Valley fetish object that gets constantly slavered over.

    • Also, on a level playing field, what’s to keep cab companies from responding by creating their own Uber apps?

      • Wapiti

        In Seattle, the yellow cabs have a “TapTheApp” ad for their own app already. Haven’t tried it myself, nor have I used Uber.

        • N__B

          In New York the yellow cab app is Arro. Works fine, although I don’t take cabs much. I expect there are similar apps all over the place.

        • MyNameIsZweig

          San Francisco has Flywheel. I’ve used it once or twice. Works well enough I guess, but I’m more of a public-transit-plus-cycling commuter anyway.

        • CD

          The Seattle Yellow Cab app is, sadly, not very good. The Uber and Lyft apps have much smoother and simpler interfaces, and they can generally get you a car faster.

          I’m no expert, but surely the expense to create and maintain the app and its underlying IT infrastructure is pretty large.

  • Bruce B.

    One of the things that’s gradually become obvious to me is how blinkered the conversation about self-driving cars is in a particular way: essentially nobody is talking with disabled human drivers about what they have to do to operate vehicles safely.

    Healthy people drive in a variety of ways that all assume handling a set of inputs in a particular time frame and making the various elements in a set of outputs in a particular time. People with the spread of handicaps available to human beings who can nonetheless operate cars safely – by themselves or with a variety of assists – can do some of these things very, very differently. It’d be nice to see some examination of that.

    Por exemplo.

    • Also, what are their driving statistics compared to the typical hairless chimp? Are they more or less prone to get into accidents, etc?

      • Bruce B.

        My very limited and patchy understanding is “it varies wildly”.

  • Predisent Putinfluffer

    “When you ride with UBER you ride with Trump!”

    That’s right! A Bobblehead of me letting you in on very lucrative investment opportunities as you ride along.

    I can’t promise anything but it just may also sing and dance!
    Still some bugs to work out so I don’t know, but…

    Should make a mint.

  • I’ve only used Uber a couple of times.

    The first time was OK.

    The last time I used them the car was a Chrysler 300 that had seen better days. I wasn’t all that impressed with the driver either.

    Granted I’ve had very mixed experiences with cabs as well.

  • Roger Ailes

    Why is the cretin Huffington on the board of anything? Reason 205 not to use Uber.

  • LexJackle

    Sadly Uber is making mistakes with trying to establish a monopoly. I have experience with the brewery industry (UK) and the big companies undercut to set up a monopoly the right way. The big companies sell their products with razor thing margins. It might be a net profit of less than 1p a can or bottle. Thats with the economics of scale, efficiencies available only to large scale operations and the purchasing power of a big corporation. Any small brewery won’t be able to produce their product as nearly as cheaply (a price we were offered was a loss of 46p per bottle, from the Co-Op) and a profit of (lets say 1p) won’t lead to anything. On 1 million bottles, you get a nice £100,000 at 1p, but for a brewery able to only give 50,000 bottles at maximum capacity, our profit would be £500, or half the monthly pay for a brewer. Not that we’d be able to make things cheap enough to match the supermarkets price point.

    So what happens? The small breweries sell to independent bottle shops and independent pubs, which has a fairly low market cap. The big money from supermarkets is closed to them and hence are casual consumers. Getting big enough to be able to swallow the 1p a bottle price point and make it work is simply impossible. If you want a wider market, you’re forced to sell out to one of the big breweries. And so the big breweries grow via someone work. Or if you don’t sell out, you can enjoy your medium sized brewery, sometimes barely able to pay its staff minimum wage.

    But the thing the big breweries do that Uber doesn’t is that they will always make a profit. It might be ridiculously low, a low exploited to kill of any independent competition, but the profit is there. They are very conscious to make sure that their cash reserves are not dwindled. That way, if something massively bad happens (say a recession or new law), the breweries can deploy their cash reserves to keep to their contracts and keep in business, to then renegotiate their prices when the contracts come up for renegotiation.

    Uber can be the same way. Set their prices so that they make the tinyest amount of profit and sit on their cash reserves. They could then sell their service as a universal cab service. So instead of going to a new city when traveling and trying to find the local cab company (in the UK, its common to walk into a pub and ask for a cab number from the barman/ normally after a fair few), you can just get it through uber. Same abroad (I’ve tried to order a taxi in Nairobi; drunkenly trying to give an address to a non-English speaking Fin (don’t ask) was interesting); I would rather have the app remember the hotel and have a `get me back’ feature. All of which would be highly profitable, possible to do with already existing cab companies (look at just-eat) and not require burning through a finite amount of venture funding.

  • toddzino

    I lurk here and usually agree with 95% of what this blog and its usual commenters are saying about the topics at hand, but I’m compelled to register and speak up on this topic because a LOT of you are failing to see the forest for the trees here.

    I live in Los Angeles, which is the poster child for a “car culture” and that spent years neglecting public transportation only to be furiously playing catch-up now to where we are slightly better than San Francisco (yay?) but still way behind NYC and any major European or East Asian capital.

    The density of the metro area overall is higher than the NYC area despite having no Manhattan or Brooklyn to be the center of the universe. As such, there was no major cab infrastructure here before ride sharing became a thing. Uber, Lyft, and their clones have fundamentally changed the transit culture of Los Angeles for the better, and the effects are already being seen in the 5 years since I moved here from NYC.

    I got rid of my car a year ago and have been subsisting on LA’s Metro system + bikeshare + Lyft. I saved $500 a month in lease premium, insurance, parking, gas, etc. offset by spending about $150 a month on Lyft. I go out 2-3x a week to all parts of LA and feel no loss of freedom or opportunity from not having my car sitting in the garage unused most days. It helps that I am able to commute to my job via light rail, but I am far from the only modern Angeleno dumping my car (or scaling back from 2 cars to 1 per household in other cases)

    When I go to Austin for work every 2 weeks (another city entirely built on car culture that is growing at geometric pace with no public transit to speak of), I am able to use rideshare to my office there and back to downtown Austin for ~$30 a day vs. paying $60+ a day for a rental car.

    Austin and LA are two examples of cities that have a large “creative freelancer” population who would much rather spend their week driving for rideshare at hours of their choosing than working part time at a coffee shop or retail store to make enough money to pay the rent between paid artistic gigs.

    As awful as Uber is for many reasons, and I root for their failure because of this, the concept of Rideshare is not some fad that is going away with Travis Kalenick’s downfall. As for using VC money for price fixing (a terrible practice that has led to Uber’s existential crisis) I would happily pay $5-$10 more per ride to normalize the background check and fair market dynamics and it would still be a better deal for me for most of my traveling and day-to-day, and this would also be better for drivers than the cab medallion cartels that legally hold immigrants hostage and in massive debt in cities like NYC. When Uber (and Lyft!) pulled out of Austin for a year from May 2016 until just a few weeks ago, 3 other rideshare companies swooped in and stood up a market within weeks and I was able to use it at the same level of convenience as before.

    Moreover, the concept of driverless cars is not going away, and plenty of major R&D shops and corporations beyond Uber have a stake in this future — this combined with electric drivetrain vs. combustion engine WILL have a net positive impact for the environment. And yes, it will have existential implications for the concept of “work” for many people, but that is happening with all other forms of automation well outside the realm of driverless cars.

    We should be having an honest conversation about the future of basic income and work opportunities across each level of education/skillset/ambition, but wishing the automation genie back in the bottle is not going to get us there — nor us letting the odiousness of Uber cloud people’s judgement on the actual logistics and benefits of ride hailing and car sharing as a net positive for a country like ours. We have let car culture drive a lot of the worst aspects of city planning, culture wars, white flight, and suburbanization that have brought us to Trumpland.

    • Ithaqua

      You pay $60/day for a rental car??? Last month I went to Seattle and paid ~$15/day (yes, that’s inclusive of everything) for a Prius for 5 days!

      • CD

        Car rental prices vary hugely city to city.

    • searcher

      > The density of the metro area overall is higher than the NYC area despite having no Manhattan or Brooklyn to be the center of the universe.

      Que?

    • “We should be having an honest conversation about the future of basic income and work opportunities”

      Yeah, I’m sure the Kochs and the Silicon Valley Libertarians are going to get right on that Universal Basic Income.

      Paul Ryan is probably drafting the legislation as we speak.

  • Warren Terra

    It’s interesting that we’re hearing about gender dynamics in the executive suites, but not behind the wheel. I’m not at all a heavy user of taxicabs or their app-driven competitors, but I’ve never or almost never had a woman cabbie, and every anecdote I hear about an Uber or Lyft driver says it’s a dude.

    I’m not an idiot; I can understand why women would be unwilling to let strange men into their cars. I can even see why women might in theory prefer Uber/Lyft to taxicabs (no cash, the customer is at least in theory identified by their app registration and payment method) – though on the other hand taxicabs often armor the driver’s compartment against the passenger, which private cars driven for Uber or Lyft don’t. But it’s notable that we seem to be cruising towards eliminating the driver without ever having figured out a way to make even the slightest strides towards gender equity.

It is main inner container footer text