It was a woman actually, but that time finally came. Uber’s self driving car will go down in history as the first one to cause a fatality. While Uber should certainly be held responsible for this, judging Uber and its ilk on moral grounds distracts from the real issues at hand.

This incident is likely to be treated as so many others before it: it will cause a commotion and attract attention, some fire-fighting measures will be announced, then it will slowly fade in the background and it will be business as usual.

The governor of Arizona, where the accident happened, has already withdrawn support for Uber and recalled its licence to conduct self-driving tests in Arizona. Others like Nvidia, the company providing much of the technology used in self driving cars, have called for giving Uber a chance, while at the same time holding off further testing on the streets, and rolling out simulations. It may seem preposterous to justify Uber at a time like this, but there are some important points to be made here.

It has been argued that the goal for self-driving cars is not to be perfect, but to be better than humans. This sounds like a pragmatic position. And it is true that no technology is introduced without having its side effects and its wild west period. But this was literally an accident waiting to happen.

An accident waiting to happen

An accident waiting to happen. Image: Reuters

Part of it has to do with the process of developing and introducing new technology, and it can be that in the long run the benefits will outweigh the side effects. But there is another part of it, the wild west part, that has to do with the lack of will and ability to oversee and regulate the use of technology.

Recent research on the deep learning algorithms used in self driving cars revealed thousands of errors. This somewhat expected outcome, given the technology’s breakneck progress and rapid application, seems to have been ignored by companies and authorities alike. In all fairness, the accident that Uber’s car was involved in may not have been related to this.

The fact that this research has been ignored however should be telling. It’s not the first time Uber has been in the limelight, scrutinized and criticized, for all the wrong reasons.

Uber is still operating in London, in you case you did not notice. Uber will continue to do so while a legal appeal process that could take a year lasts. What’s more, the fire-fighting statements and apologetic tone adopted by newly appointed Uber top management seem to appease some, including London’s mayor.

But to focus on Uber’s misconduct and ethics, to lay personal blame and to seek and accept apologies and promises is to miss the point entirely. Uber, and organizations like Uber, are neither good nor bad – they are signs of the time. Even if Uber was ran by Arizona’s Governor or London’s Mayor, it would still have the same defining qualities and effects.

To focus on Uber's ethics is to miss the point entirely; Uber is part of the rising data monopolies. Image: derivative, original by Anya Mooney

To focus on Uber’s ethics is to miss the point entirely; Uber is part of the rising data monopolies. Image: derivative, original by Anya Mooney

Its efficiency is based on optimized and evolving algorithms, clever marketing and big data. Its self centered nature is inevitable, as it has no one to answer to except its shareholders.

Uber may be revolutionary, but not for the reasons you think. A future in which car ownership is obsolete and you can be picked up in no time and driven safely and efficiently to your destination for cheap is something many people would stand behind. Except there won’t be drivers in those cars, and it will be up to Uber to run things as it sees fit.

It’s clear that the combination of big data, processing power and algorithms can progressively automate every task to the point of making it more efficient than what humans are able to achieve. Driving and dispatching is no exception, and that’s what Uber and its ilk are doing.

But that’s only part of the reason why Uber is displacing traditional taxis. The other part is Uber’s employment model. Instead of employing full time, properly trained drivers, Uber will employ just about anyone with a car and willing to spend hours behind the wheel.

These people will be precarious workers with minimum rights and income, be manipulated to stay on the road as long as needed, and be disposed of when self driving technology and legislation are in place – which should not be too long.

In the meanwhile, Uber can sit back and watch the divide and conquer strategy that has played out so well throughout time work in its favor. Uber drivers operating as an army of low-paid disposable contractors before the algorithms take over completely are inadvertently helping dispose of everyone else’s rights and livelihoods as well.

As Wired reports, New York City’s cab drivers are in crisis, and they’re blaming Uber and Lyft. Since December, four taxi drivers have killed themselves, seemingly in response to the intense financial pressures that have accompanied an increase in for-hire vehicles on the city’s streets.

So it’s freelancers versus full time employees, and now Uber sympathizers versus the people and regulators. Uber sympathizers who have signed an Uber petition to keep it in the streets of London are closing the one million mark, citing safety and loss of jobs. Many would probably cite innovation and better service as well.

While these claims are not entirely unfounded, they are hollow. These jobs will be soon lost anyway, and there have been enough reported incidents to undermine security claims. But this brings us to the core of the issue: the emerging data driven monopolies.

Efficiency and safety are both based on a foundation of data. Data collected, processed and used by Uber to power its algorithms in complete opaqueness. By gaining market share, Uber is amassing ever more data, in a reinforcement loop that makes it harder and harder to compete against.

The fact that Uber ditches every notion of ethics and legality in the process, by doing things such as collecting data from user devices without consent even when the application is not runningusing that data to drive analytics that determine pricing and using backdoors to spy on users and apps to evade control is just adding insult to injury.

You can expect data monopolies to operate similarly to good old monopolies, except more efficiently. Image: Anya Mooney

You can expect data monopolies to operate similarly to good old monopolies, except more efficiently. Image: Anya Mooney

But, should not the market self-regulate, and will there not be competition from other innovative companies? Let’s look at another part of the world for answers: Russia.

In Russia Uber was facing stiff competition from Yandex. Yandex is a Russia-based technology giant that dominates its home market in search, cloud services and ride hailing among other things.

Both companies have been using similar approaches to capture market share, resulting in driving prices down and owning a combined near 90% of the local market. Now Uber and Yandex Taxi have made a deal to work together, in essence forming a monopoly. What are the chances of anyone else, let alone independent drivers, competing in this landscape?

Greg Abovsky, Yandex CFO, responded to a request for comment by citing the deal is subject to approval by Russian regulators, and the argument is that since there is room for growth in the market this is not a monopoly.

Yandex is often called the Russian Google, and this does sound a bit like what Google would sound like if they said they are not a monopoly in search because more people will be searching online in the future.

First mover advantage in the big data and AI age will be tremendously important if left unchecked. There’s an interesting implication of this however. These technologies will make the market smarter and make it possible to plan and predict market forces so as to allow us to finally achieve a planned economy.

If you’re wondering where such a bold claim may be coming from, it’s none other than Jack Ma, the founder of another one in the league of giants: Alibaba. Companies of this caliber already dwarf governments in nearly every aspect, including their ability to gather and process data.

Some economists argue that the online platform monopolies resemble central planning institutions, so it would be more “legitimate and rational” for the state to become a “super-monopoly” platform.

This may sound scary and big-brother-ish. But before we get lost in the arguments in favor of one or the other monopoly, let’s think about the real issue: allegiance and control. Where does corporate allegiance lay, and how much control do we have over it? Then what about the state?

In a world that is increasingly becoming data-driven, reinventing algorithms and institutions seems like more than a realistic option – it seems inevitable. The real question is by whom, and for whom. If we want to be actors and citizens rather than users and consumers, it’s time we reinvented our collective identity and started taking control.

This assassination of character is what we should be really worried about.

This article was first published as Keep on Uberin the free world, on the Linked Data Orchestration blog.

Photo by marki1983

2 Comments Mama, Uber just killed a man – or more

  1. Kevin CarsonKevin Carson

    As my friend @LinuxSocist noted on Twitter, any software involved in making autonomous decisions that may cost the life of a human being should required to open its source code to the public for inspection.

  2. George AnadiotisGeorge Anadiotis

    I was also thinking along similar lines. The possibilities there are endless, especially since there’s ethical decisions involved. Would you trust a car with hard-coded accident behavior rules you have no access to? Would you be tempted to hack it? Would towns or districts ban cars that do not comply to certain software standards?

Leave A Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.