Posts by BenWilson
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Speaker: Confessions of an Uber Driver…, in reply to
In the taxi industry we have rules around this issue, the rules apply to contractors, employed drivers and shareholders etc. Failure to carry out an accepted order unless there is a justified reason fines apply. The fine is usually in excess of the fare value.
Uber do it simply as a numbers game. Cancel too many and you might be timed out or lose your right to drive for them. There’s not some dispatcher out there making sure orders are met, it’s all algorithms, and the decisions are all made by algorithms. We don’t know what these algorithms are, nor do any staff. When they decide something, it can’t be appealed (according to them). If you attempt to, it’s got to be done as a support call, as if the software or process broke down, rather than as a person speaking to another competent person organizing a business. It’s just not that sort of relationship. Your support message will be answered by a “robomuppet” (as we call them), someone overseas who will not really read it properly and will answer it entirely using a boilerplate response, at least the first time. You’ll have to respond with follow ups pointing out the missing overlap several times until eventually, if you’re persistent enough, it will go a locally based person. Usually their modus operandi is to appease both parties, regardless of the facts of the dispute, unless it’s something highly noteworthy, in which case they will act arbitrarily and with as much prejudice as they can muster.
I don’t think many drivers ride around with both apps on all the time, though. It’s more likely that they put themselves offline on A when a call comes in on B, and vice versa. But they have to do that by hand, every time, and it might get missed. Then you find yourself driving to an Uber and getting a Zoomy request coming in. Do you really owe the Uber passenger their average $7 ride when the Zoomy has a minimum $10 fare, and you’re not an employee of either shop, but instead “responding to a lead”? I can see why people jump between. If the ONLY right you actually have is to choose when to work and which organization to work for, why should you not choose the better lead?
Yes, it’s not the way taxis work. They’re designed around a much more traditional boss/worker relationship. As in, you have one, and are the other. With the gig economy, it’s cutthroat by it’s own design.
-
Speaker: Confessions of an Uber Driver…, in reply to
I am of the opinion once one has accepted a job that person has made a contract with the hirer, the only reason for cancellation is a break down and then the hirer should be contacted to be informed why they have been cancelled.
There's far too many issues built in here to unpack easily. Firstly, Uber would maintain that there is no contract with them whatsoever. In fact, Uber NZ claims that there is no contract between them and drivers at any time, that the contractors in NZ have no contract with an NZ based agent of any kind. The contract, if it is a contract at all, is with Uber BV, based in the Netherlands.
Secondly, they (Uber BV, via their proxy in NZ that claims not to represent them) claim that the nature of their business is not to sell rides to customers at all, but to sell "Leads" to drivers, at which point the driver meets the passenger and a contract is formed between those parties to which Uber BV (and certainly Uber NZ) is a stranger. Uber provides only the the lead and the service of processing the payment, they claim in the preamble to their contract.
So if this is to be taken at face value, no contract is formed with the passenger until they get into the car. Since the driver is only responding to "leads" there is no reason for them not to choose the more promising of 2 "leads", if they come at the same time.
However, because this is at total and utter variance with all common sense and everyone's practical experience of riding in an Uber, and how Uber market and sell themselves, they also write into the driver's contract with Uber BV many pages dictating exactly how the driver must respond to and conduct themselves during these contracts that Uber is supposedly a stranger to. Most of the contract reads like an employment contract, except for the start of it, which specifically excludes Uber from:
1. Operating a transport business
2. Hiring any drivers
3. Having any contract with passengers
4. Having any liability whatsoever for anything that happens to anyone anywhere anytime
5. The jurisdiction in which the actual business is conducted applying to them.In some ways, the way they conduct business could also be seen to be making good on that distancing:
1. There is no phone number or email address for anyone to ever call in NZ, ever advertised on their website or anything in print.
2. They do not keep any regular office hours. Their office seems to exist only for signing drivers up, and they only very, very grudgingly will deal with operational matters
3. You can never, expect to speak to their management. It's not clear if there even is any management in the country. There are only spokespeople and marketing types, if you believe their claims in court, rather than their public announcements.
4. Who the hell else would suggest to people that they break the law for a living except for people completely distanced from any responsibility for their actions?So yes, in the presence of these factors, it's hardly surprising that Uber drivers do not feel that they owe Uber, or Uber customers anything, except when they actually get into the car. Anyone who has dealt with them in any way that has ever gone outside of the plain vanilla ride quickly comes to realize that they are not dealing with a properly constituted organization at all. It is a law unto itself, and the longer term drivers begin to act more and more like that the longer they keep doing it, as contact with Uber is something self respecting people would want to keep to a minimum - if only to avoid the anger and frustration that comes with all official dealings with them, and to avoid smacking one of the staff over one day, something that becomes more and more tempting every time they are dealt with. There is a reason that one of the 4 employees that Uber has in its front office is a security guard, and that all the other employees (if there even are any employees at all) are hidden behind big locked doors.
-
Speaker: Confessions of an Uber Driver…, in reply to
It’s a well known problem. I’ve written algorithms that attempt to solve it without doing a full Dijkstra walk through the road network for each location pair.
Me too, and that was in the 1990s. Now they can do it with a thousand times more computing power than I had. Which begs the question of why they haven't. I suggest it's because they're actually not as competent at this stuff than their public image would suggest. Their genius is not in app development, certainly not in algorithm design, and absolutely not in robotics. It's in marketing and worker exploitation. I don't see that translating into self-driving cars. It translates into bilking investors until they realize that they're subsidizing vaporware.
-
Speaker: Confessions of an Uber Driver…, in reply to
Surely that information must raise alarm bells with the thousands of other illegal Uber drivers operating in NZ?
It very much depends on whether Uber simply picks up the tab. One this is pretty clear though - left to their own devices, it's pretty much at their pleasure if they do. I expect they probably would want to do so quickly to minimize the damage created by it getting out that this guy is pretty much left to fend for himself.
-
Speaker: Confessions of an Uber Driver…, in reply to
The pickups are frequently at places that are not addresses
Just by way of example, two times I've been directed by both Uber and Google Navigation (which Uber launches if you choose it) to pick people up from the side of the motorway on the approach to the Harbour Bridge going north. There isn't even a road shoulder to stop at. In one case the rider was going to 1 Shelley Beach Rd. In the other they were to be picked up in the Sales bar in Westhaven.
Obviously my local knowledge and common sense picked up a problem when directed from Ponsonby to the Wellington St on-ramp, and I just ignored Google. Similarly when directed to the same on-ramp for a motorway-side pickup, I just called first and simply understood their drunken verbal description of the bar they were in.
Had an automaton been doing these trips, in the first place they would hopefully not have stopped on a corner on the motorway and requested the passengers to get out, but rather driven on to the next available exit (Stafford Rd, or Onewa Rd) and then either got the passengers to get out, or driven them back under direction (not sure how these directions are given to a self-driving car) to Shelley Beach Rd, after which they would need to get a fare review to deal with being driven about 10 times further than necessary.
In the second place, again, I would hope again that the car would know not to stop on the motorway, but exactly how it would find out where the passenger really is, I do not know. It's clearly not practical to just expect the passenger to find their way across 8 lanes of motorway to a vehicle. Presumably they have to cancel the trip and order again, taking more care with their pin location. But care with things like that is exactly the opposite of what you get with Uber customers. It's one of the great selling features of the app that you can be totally lazy about that, don't even need to have the first clue where you are, because an actual human will work it out for you when they get there. And I'm pretty sure the app snaps to what it thinks is the closest road to the physical address anyway, since the passenger to 1 Shelley Beach Rd had literally entered that in the app by typing it. The Uber app simply thinks that address happens to be on the side of the motorway.
Now every issue can be fixed eventually. But I'm not seeing any kind of streaking changes in quality in this app over the year to fill me with optimism. Quite the opposite, both Uber and Google navigation have got flakier. There is certainly no big drive to get drivers to help them out by reporting these issues - in fact every interaction with support is so punishingly cretinous that drivers soon learn to do it only in their own interests, typically to claim money.
So even if self driving cars are coming, I'm really not seeing anything to indicate that self driving Ubers will dominate the world. If anything they look set to break something that is currently reasonably high quality, replacing it with something really crap.
-
Speaker: Confessions of an Uber Driver…, in reply to
If I’m using it in town, I’ll need to pay either parking or a few extra kms for it to drive out on its own to someplace with free parks.
The first solution will work now. It uses less petrol and puts the same amount of cars on the road, but for less time. But you do have to drive it so you can't get drunk. If your economic assessment is accurate, it's pretty marginal whether this wonder of the world will save money, and it certainly will make the traffic problem slightly worse than it is now, along with externalizing your parking onto the inner suburbs, thus making things significantly worse for people who live there.
Similarly, you can send your current car off to earn a living right now, if you're prepared to rent it to an Uber driver during the day. Are you? I wouldn't be. It's not practical for the pittance it would save and all the damage and inconvenience you'd be putting up with.
The problem is the puke factor. What percentage of passengers will mess up cars?
One of the many problems. As someone who actually picks up Uber passengers, I can tell you that there's no way the app is robust enough to find most passengers at all, nor is Google navigation good enough to travel NZ streets unmanned. I'm getting sent into no-entry streets all the time, and being taken down routes that require making illegal turns. The pickups are frequently at places that are not addresses, and road works continually mess with what roads are open or closed, with detours that require the ability to follow street signs and the hand directions of human beings. Uber's pinpointing of pickup locations is particularly bad, frequently snapping locations to entirely different streets than the passenger is actually on. Many passengers will verbally direct me to drive along things that are not official roads, like country driveways, and long right of ways. Uber constantly directs me to use things that are not roads at all, I have no idea where they get their road map from, it uses carparks as shortcuts, even when there is no connection through.
Its closest driver algorithm is also clearly not using the road network in its predictions, as a frequent driver complaint is being assigned as the closest car to someone in Downtown when the driver is in Devonport.
Yes, every aspect of this is fixable. But it's going to be the work of a great many years. And always I come back to what the point of it all is, considering just what an enormous undertaking it is. Is the current solution really that broken? Could the same investment not be better spent in so many ways to improve transportation? It really is so much easier to just use solutions that have already targeted what the actual problem is, the way trains, buses, and yes, taxis, already do. Not to mention better town planning, including higher density living in the first place.
-
Speaker: Confessions of an Uber Driver…, in reply to
Oh, hang on, soz. You specifically only meant:
- or I don't own a car at all, and just send for one when I need it
Which is pretty much what I said before. This exists, and is thus not innovative but for the one amazingly unvisionary part of taking out the human driver who might make a living from it. In all other ways, you are describing a taxi service, something that's been around since there have been chariots.
-
Speaker: Confessions of an Uber Driver…, in reply to
So there's a modern version of the same thing, which is to let your car out to a taxi driver. Consider how appealing this sounds now, and multiply it by a factor of illegality, low prices, and no human to look after your car while they're in it.
-
Speaker: Confessions of an Uber Driver…, in reply to
I agree that there are lots of question marks over self driving cars but am less sure that consumer willingness to use these would be the deal breaker so to say.
Yes, and also the question of whether they're technically viable at all, and if so, at what cost, none of which is conclusively answered. It's more than a Moon shot. It's a Mars shot. There are highly risky points of failure all the way along it.
I'm still very much not sold on how "visionary" it is, too. It really would seem to be a play to eliminate an extremely cheap and efficient and high quality workforce, for no other reason than just to do that. Which is typical of AI projects, and also the typical reason that they fail horribly. Instead of reengineering some aspect of our lives, they simply try to outright replicate the most complicated part of it (the human being) to preserve the existing engineering solution. Self driving cars will still be taxis clogging up the road system, hugely inefficient in energy use compared to alternatives that have existed for over a hundred years. They're they kind of technological advance science fiction writers would think of, rather than the people who have to actually make it, or the people fucked up by it.
-
Speaker: Confessions of an Uber Driver…, in reply to
The difference between Uber and Amazon is that Amazon were not relying on the biggest breakthrough in the history of robotics to make their business model vaguely viable (it's still not clear how much those cars are going to cost or who will pay for them all). That is one hell of a risk. Add on to that the necessity that legislators in every territory they want to operate in will need to cooperate, something Uber has an entirely negative history of. Add on to that, that all Uber itself actually controls is an app and servers, something that competitors can easily smash open. I don't think they can rely on self-driving cars to maintain their customer base - that's not what the current customers signed up for. How are you going to downrate a self-driving car? How are you going to interact with it generally? As a driver, I know that 95% of rides do involve actually speaking to the passenger and understanding their responses. When you get in an Uber, you're expecting a thing that you don't have to drive yourself, but instead direct with verbal commands and gestures. Your first contact will probably be a phone call. It's simply not the same thing at all, so there's little reason to think that the customer base will seamlessly switch to it.