![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
landingtreeby Katie J. Wells, Kafui Attoh, and Declan Cullen.
I was trying to start writing a poem! However I have no idea which notebook the first bits of the poem are in, and instead came upon my notes on this book, so I may as well write them up in passing.
I have discovered that what will cause me to finish an urban planning book club book well in advance, though I suspect it only works with fairly short books, is having promised to lend my kindle to a friend so she can read it afterwards. (I may have another NZ recruit to the book club, or then again she may just want to read this one, we shall see).
I was keen to read about Uber because I haven't been paying a lot of attention to them, so have mainly just received a sense of cartoonish corporate villainy via filter-feeding. This book was neither a revelation to me, nor uninteresting: I read it going 'Yes, of course that's how it works'.
The book is based on repeat interviews with forty-odd D.C.-area Uber drivers and a range of other figures on the scene, with the five central chapters - on regulation, race, data, the ideology of AVs, and the conditions of driving for the platform - being divided up between the different co-writers. Uber came to Washington D.C. early, and worked on what has become its playbook for defeating local regulation: getting its customer base to bombard politicians with emails, presenting itself as a solution to problems of racial discrimination, disability access, and stagnation of transit systems; feeding off, and feeding, a sense that we can't expect too much of our cities and that regulation is terribly old hat. This last is the book's overarching point, and the reason for its title: as a solution to problems, Uber makes sense in context, but the context is bad and ought to be changed.
I'm super interested to know how this reads to residents of Washington D.C.
1. Regulation.
The D.C. taxi commission was widely perceived as corrupt and racist, which is one reason Uber could do so well. The examples given here include bribery and guns in meetings. Taxis were rusty, many required cash, few allowed wheelchairs, until 2008 none had meters, and many ignored hails from Black people.
Uber emailed all its DC customers asking them to bombard city officials with messages requesting a removal of mandatory minimum fares from the regulations about to be imposed. This was done overnight. (I'd love to know what the conversation about that sounded like).
Uber was given its own category, free from existing taxi regs.
Taxis in DC were interconnected (a quarter were driven by Ethiopian immigrants with strong social networks) but not centralised: there were a hundred and fourteen companies, with the biggest four not even adding up to a quarter of the total. In New York, taxi companies had a quick, organised response to Uber, but not in DC.
Disability advocates had only recently begun making gains in the DC taxi industry, and were now confronted with a whole new unregulated battlefield in which their work was reset to zero. 2014 regulations did not require from Uber a percentage of accessible vehicles, or vehicles permitting service dogs; it only required a non-binding statement of intent about those things.
Quote: "Uber increasingly sells itself as capable of providing a public service but fiercely fights any regulation that would actually require it to do so."
We don't know if Uber discriminates against wheelchair users, because no one is allowed to access that data: how long do wheelchair-users wait? What treatment do they get in the cab? Uber is exempt from the Freedom of Information Act. They don't have to tell anyone.
Context in which all this could happen? DC government was already committed to innovation and skeptical about regulation before Uber came along. Uber made sense.
2. Race.
Uber's strategy in advertising itself as a solution to racism requires a focus at the level of the taxi-cab. A Black person tells a story about how they couldn't commute home from work because taxis wouldn't stop for them, despite picking up their White colleagues; Uber solved this problem. This anecdote yields positivity for Uber, instead of the question, "Hey, why isn't there bus service to this majority-Black district?" because buses are not the right shape of solution for the reigning common sense.
Also, is Uber's algorithm actually non-discriminatory? Again, no one can make them prove it.
In the late 70s DC got a local, non-federal government for the first time. The first council was majority-Black and had roots in a particular protest group. (I want to know more about this; how did this time end?
Went to add the history of race in DC Chocolate City to the group bookshelf, and found it already there).
When it claims positive social effects for itself, is Uber being sincere or cynical? This book doesn't care. The general effect of Uber is unrelated to the question.
3. Data.
An interviewed driver described himself as helping build the dataset which Uber would use to replace him, in a few years' time, with a driverless car.
Uber offers packets of data to cities, which are often unprocessed, partial, or just plain inaccurate. There is a thirst for data in government, giving way in some of the people this book interviewed to a cynicism: data is only as useful as what's being done with it. Most of this data is currently used to write reports that don't make it into policy. 'Could you please tell us what a solution to these problems could be? We didn't hear you the last thirty-eight times.'
Would knowing all e-scooter movements be a privacy concern? Well, Washington local government already gets public housing applications, ambulance call logs, etc. The decision to give government some highly sensitive data is bedded in. (This makes me want to read a whole book about local government data).
4. AVs
In 2020 Uber sold its scooter, ebike, Automated Vehicle, and flying car divisions. I did not know it had a flying car division.
Uber's stated business model was 1. Dominate the market, 2. Raise prices, 3. Invest the proceeds in AVs.
A driver said, "It feels like we're just rentals."
Even without being deployed, AVs are useful, because "conversations about workers' classification, pay, and surveillance become anachronistic."
One driver's skepticism: how could Uber possibly expect to maintain a high-tech fleet from its starting point of 'Our infrastructure is other people's cars'?
There's a sidenote here on food delivery robots, which were touted as solutions to poor areas lacking walking-distance access to healthy food, and then actually deployed not in those areas, but on college campuses. We are all so surprised.
This chapter on AVs is devoid on information about AVs themselves: it's all about the degree to which they're being put forward as a solution to problems they cannot solve. The reasons why they're hard to make don't really come into it: even if they worked perfectly, it's not clear that would bridge the gap with the rhetoric.
5. The drivers.
Many Uber drivers consider themselves the smart ones who can game the system, while looking at their colleagues with pity. Almost half the drivers interviewed mentioned how smart they themselves were.
There is no Uber depot. Drivers are atomised. (I think of the prison in Andor, its gamification of work; but that at least builds a working group! Uber makes every driver compete with every other).
If a driver breaks a strike, no one can tell: online, you're anonymous. [There is an unverified story of drivers in Penang who did manage to unionize; I'd like to know more about this.]
What with the algorithm's mysterious carrots and sticks, Uber drivers seldom know what they're paid per mile. "One expert on the ride-hailing industry estimates that calculating drivers' earnings requires no fewer than twenty pieces of information." "There is no clear relationship between a passenger's fare and a driver's pay."
An accident of policy made airport parking lots a place where a lot of drivers were waiting, without a pressing need to stay in their cars: they could hang out. This let them develop strategies to game the algorithm, turning off their apps en masse just when a big flight was about to come in, thus making the algorithm raise wages in an attempt to get more drivers on scene.
Most drivers in the airport lots were Black - or at least, this was the perception (derogatory) of many non-Black drivers interviewed.
The airport parking lots were the focus of a 2019 partial strike. It got good media attention - for a minute. Then police presence in the parking lots was doubled; Uber turned the story into 'Greedy drivers are manipulating our algorithm' while firing the drivers who had talked to journalists about it; and Uber capped surge pricing to stop the strategy from working.
I was trying to start writing a poem! However I have no idea which notebook the first bits of the poem are in, and instead came upon my notes on this book, so I may as well write them up in passing.
I have discovered that what will cause me to finish an urban planning book club book well in advance, though I suspect it only works with fairly short books, is having promised to lend my kindle to a friend so she can read it afterwards. (I may have another NZ recruit to the book club, or then again she may just want to read this one, we shall see).
I was keen to read about Uber because I haven't been paying a lot of attention to them, so have mainly just received a sense of cartoonish corporate villainy via filter-feeding. This book was neither a revelation to me, nor uninteresting: I read it going 'Yes, of course that's how it works'.
The book is based on repeat interviews with forty-odd D.C.-area Uber drivers and a range of other figures on the scene, with the five central chapters - on regulation, race, data, the ideology of AVs, and the conditions of driving for the platform - being divided up between the different co-writers. Uber came to Washington D.C. early, and worked on what has become its playbook for defeating local regulation: getting its customer base to bombard politicians with emails, presenting itself as a solution to problems of racial discrimination, disability access, and stagnation of transit systems; feeding off, and feeding, a sense that we can't expect too much of our cities and that regulation is terribly old hat. This last is the book's overarching point, and the reason for its title: as a solution to problems, Uber makes sense in context, but the context is bad and ought to be changed.
I'm super interested to know how this reads to residents of Washington D.C.
1. Regulation.
The D.C. taxi commission was widely perceived as corrupt and racist, which is one reason Uber could do so well. The examples given here include bribery and guns in meetings. Taxis were rusty, many required cash, few allowed wheelchairs, until 2008 none had meters, and many ignored hails from Black people.
Uber emailed all its DC customers asking them to bombard city officials with messages requesting a removal of mandatory minimum fares from the regulations about to be imposed. This was done overnight. (I'd love to know what the conversation about that sounded like).
Uber was given its own category, free from existing taxi regs.
Taxis in DC were interconnected (a quarter were driven by Ethiopian immigrants with strong social networks) but not centralised: there were a hundred and fourteen companies, with the biggest four not even adding up to a quarter of the total. In New York, taxi companies had a quick, organised response to Uber, but not in DC.
Disability advocates had only recently begun making gains in the DC taxi industry, and were now confronted with a whole new unregulated battlefield in which their work was reset to zero. 2014 regulations did not require from Uber a percentage of accessible vehicles, or vehicles permitting service dogs; it only required a non-binding statement of intent about those things.
Quote: "Uber increasingly sells itself as capable of providing a public service but fiercely fights any regulation that would actually require it to do so."
We don't know if Uber discriminates against wheelchair users, because no one is allowed to access that data: how long do wheelchair-users wait? What treatment do they get in the cab? Uber is exempt from the Freedom of Information Act. They don't have to tell anyone.
Context in which all this could happen? DC government was already committed to innovation and skeptical about regulation before Uber came along. Uber made sense.
2. Race.
Uber's strategy in advertising itself as a solution to racism requires a focus at the level of the taxi-cab. A Black person tells a story about how they couldn't commute home from work because taxis wouldn't stop for them, despite picking up their White colleagues; Uber solved this problem. This anecdote yields positivity for Uber, instead of the question, "Hey, why isn't there bus service to this majority-Black district?" because buses are not the right shape of solution for the reigning common sense.
Also, is Uber's algorithm actually non-discriminatory? Again, no one can make them prove it.
In the late 70s DC got a local, non-federal government for the first time. The first council was majority-Black and had roots in a particular protest group. (I want to know more about this; how did this time end?
Went to add the history of race in DC Chocolate City to the group bookshelf, and found it already there).
When it claims positive social effects for itself, is Uber being sincere or cynical? This book doesn't care. The general effect of Uber is unrelated to the question.
3. Data.
An interviewed driver described himself as helping build the dataset which Uber would use to replace him, in a few years' time, with a driverless car.
Uber offers packets of data to cities, which are often unprocessed, partial, or just plain inaccurate. There is a thirst for data in government, giving way in some of the people this book interviewed to a cynicism: data is only as useful as what's being done with it. Most of this data is currently used to write reports that don't make it into policy. 'Could you please tell us what a solution to these problems could be? We didn't hear you the last thirty-eight times.'
Would knowing all e-scooter movements be a privacy concern? Well, Washington local government already gets public housing applications, ambulance call logs, etc. The decision to give government some highly sensitive data is bedded in. (This makes me want to read a whole book about local government data).
4. AVs
In 2020 Uber sold its scooter, ebike, Automated Vehicle, and flying car divisions. I did not know it had a flying car division.
Uber's stated business model was 1. Dominate the market, 2. Raise prices, 3. Invest the proceeds in AVs.
A driver said, "It feels like we're just rentals."
Even without being deployed, AVs are useful, because "conversations about workers' classification, pay, and surveillance become anachronistic."
One driver's skepticism: how could Uber possibly expect to maintain a high-tech fleet from its starting point of 'Our infrastructure is other people's cars'?
There's a sidenote here on food delivery robots, which were touted as solutions to poor areas lacking walking-distance access to healthy food, and then actually deployed not in those areas, but on college campuses. We are all so surprised.
This chapter on AVs is devoid on information about AVs themselves: it's all about the degree to which they're being put forward as a solution to problems they cannot solve. The reasons why they're hard to make don't really come into it: even if they worked perfectly, it's not clear that would bridge the gap with the rhetoric.
5. The drivers.
Many Uber drivers consider themselves the smart ones who can game the system, while looking at their colleagues with pity. Almost half the drivers interviewed mentioned how smart they themselves were.
There is no Uber depot. Drivers are atomised. (I think of the prison in Andor, its gamification of work; but that at least builds a working group! Uber makes every driver compete with every other).
If a driver breaks a strike, no one can tell: online, you're anonymous. [There is an unverified story of drivers in Penang who did manage to unionize; I'd like to know more about this.]
What with the algorithm's mysterious carrots and sticks, Uber drivers seldom know what they're paid per mile. "One expert on the ride-hailing industry estimates that calculating drivers' earnings requires no fewer than twenty pieces of information." "There is no clear relationship between a passenger's fare and a driver's pay."
An accident of policy made airport parking lots a place where a lot of drivers were waiting, without a pressing need to stay in their cars: they could hang out. This let them develop strategies to game the algorithm, turning off their apps en masse just when a big flight was about to come in, thus making the algorithm raise wages in an attempt to get more drivers on scene.
Most drivers in the airport lots were Black - or at least, this was the perception (derogatory) of many non-Black drivers interviewed.
The airport parking lots were the focus of a 2019 partial strike. It got good media attention - for a minute. Then police presence in the parking lots was doubled; Uber turned the story into 'Greedy drivers are manipulating our algorithm' while firing the drivers who had talked to journalists about it; and Uber capped surge pricing to stop the strategy from working.
no subject
Date: 2024-03-05 12:12 pm (UTC)