After Uber’s self driving car fatal accident, I want to share three observations – the last being by far most important – that in my opinion are missing or under-valued (besides, which is worrying, compassion for Elaine Herzberg’s family and friends).
1) It’s not business news
When the news broke – see the link to BBC above – it was published as “business” news, along the lines of what this may mean for Uber, the share price of any company involved in autonomous cars, of traditional car companies etc. While I understand there must be a bigger reason for reporting a fatal car crash globally, we may all agree that the category should better be somewhere between society, science and future. The story is so much bigger than its implication on some business. And if it’s about business, the most interesting aspect would be this:
2) Ever noticed that Uber is not a car company?
Somehow it became “Uber’s car”. And this “somehow” is software and data.
According to my research, it was a Volvo car. But it could have been a Toyota, Mercedes, Volkswagen or Schmolkswagen, because in this context, who assembled wheels to the vehicle’s axle is irrelevant. If I were a traditional car maker, this would be truly worrying – how quickly it became normal for us to talk of an Uber vehicle.
3) The missing discussion about societal buy-in
It’s not about the technology’s blind spots (we will solve these eventually, I am sure), and the most popular argument is that humans kill way more people, so Elaine Herzberg’s death is collateral damage and simply the price we pay for the advancement of technology. I can see where this comes from – I am just convinced that it is missing the point. As I have tried to explain before (lengthy and unnecessarily laborious), technology that can have such an impact on people’s lives (in this case: life and death) and the way they live together (how cities are designed, how accessible mobility is etc.) will need a certain level of societal buy-in. This doesn’t necessarily mean that political parties and institutions need to be involved in the concrete design of such technology and that a highly regulated market has to be installed before such a market really exists, but it means that societies need to understand the decision making process far beyond ex post facto police investigations and the release of (awful, and I belive borderline inhuman) video footage of the accident.
We, the public, have a right to be part of the process of developing driverless mobility. One part is secured by laws – you can’t just put any vehicle on the roads, but that’s just a “go or no” switch. This can’t possibly be enough. Only if logics and mechanisms are made publically available, and a negotiated portion of results are made available open source, so many institutions and market participants can compete in creating alternatives, accumulate and exchange knowledge and collaborate on refinements, should tests on public roads be allowed. That certainly wouldn’t help the Herzberg family, but it would acknowledge that all of us are impacted by such technologies, so all of us – our institutions, our companies, our universities, our talents – should be part of the process of adopting, and more importantly, understanding and creating what they do, how they do it, and why. We cannot accept a handful of companies to design and decide such impactful technology.
If we don’t get our part in it, we – as societies – will react irrationally to such technologies. That may include to let this technology unfold completely unregulated: compare how we have no clue how exactly a Facebook, YouTube and Google search algorithm works, but we let those decide who gets exposed to which information, and ultimately decide how we construct realities we live in and then vote and act accordingly. But it may also be the opposite, over-regulation and protectionism of old, rather destructive structures and industries. And anything in between.
As much as technology evolves faster and faster and faster as ever before, we as societies need to learn faster and faster as ever before, and adapt accordingly. The discussion should not be: “driverless cars – for or against”, but how a process of societal participation can be organized. Because this will not only be relevant for self driving cars, but for many other areas of our lives, too.