In the aftermath of Google's unexpected unveiling of its very own steering wheel-free, the internet is dripping again with opinions about self-driving cars.
Google's self-driving car is a tremendous step in the right direction. Cobbled together by Google's secret R&D lab from the survivors of DARPA's Grand Challenges (three races that pitted self-driving cars against simple courses) the project maintains the sort of 'let's-see-what-technology-can-do' spirit that's made Silicon Valley famous (and powerful). However, it's always been a little bit unclear what Google actually plans to do with the technology.
The company's vision became a little clearer this week, when Google showed off its new batch of prototypes. The basic design of the bulbous, little electric cars speak volumes about what Google wants to accomplish. To call the interior pared down is an understatement. There's no steering wheel, no accelerator, no brake pedal, no glove box; just a button for start, a button for stop, and a screen to display the route. The vehicle's top speed is just 25 miles per hour, and the front of the car is made out of foam in the off chance that a collision doesn't happen. It's also important to emphasise that this is just a prototype. Google's building about 100 of them for research purposes. If Google does eventually bring a self-driving car to market, it will surely be different.
When you step back and look back at the whole package it's easy to identify the three basic tenets of the self-driving dream: efficiency, autonomy, and safety. It's not hard to see how these three principles are central to innovation in all transportation, from the Underground to the Dreamliner. At the end of the day, we all just want to get where we're going. We want to get there in a timely fashion without destroying the Earth too much, and we want technology to help us do so with as little effort as possible. And, of course, we'd rather not die along the way.
Whether Google builds it or not (and Google's certainly in the lead here) the self-driving car is a worthy goal. It's a moonshot aimed at making our most common form of transportation better in every way. The self-driving car points to an idealistic future, where there are no road fatalities, no traffic jams, no wasted time. This is not an easy idea. But it's one worth pursuing.
So when critics say that Google's self-driving car is destined to fail, it feels defeatist. Yes, the challenge of building a car that can drive itself, manoeuvre through the streets without hitting any pedestrians, and respond in real time to moving obstacles from all directions is grand. But half the fun of moonshot is what all the failed attempts teach us. Some of those teachings are already creeping into the automotive market, in fact.
Here's an idea: Google's self-driving car should fail in the short term, but it must win in the end. Indeed, Google's self-driving car already fails on a daily basis. That's part of the point. If you read any of the growing number of stories about riding in these cars, the robotic driver gets confused, and the human driver has to take over every now and then. The Google X team then uses each instance to improve the computers artificial intelligence and to teach it how to be a better driver. Removing the human driver altogether, as Google's new self-driving prototype does, raises the stakes, forcing the engineers to write better algorithms and fail less often.
Then there's the more macroscopic view of this Google-powered self-driving car craze that's been capturing headlines lately. Critics make a good point that the self-driving car is no panacea. Accidents will still happen. Traffic will still jam. Public transport will still suck. Those are all really hard problems. Automotive engineers, urban planners, and politicians have been trying to solve them for years, and you know what? They're not going to go away. But just because self-driving cars can't solve those issues overnight doesn't mean they can't make things better.
In fact, it's hard to imagine how this jaw-dropping technology won't have a positive impact on all forms of transportation. With an estimated 30,000 automotive-related deaths a year, we only stand to benefit from making cars safer in every regard, and since nearly 90 per cent of accidents happen because of human error, adding a little bit of well-tested autonomy into the equation can only help. That same technology will also enable us to be more efficient with how our vehicles consume energy, and how much space we take up on the road. Just picture that motorway scene in Minority Report to get an idea for how traffic jams won't exist in a future powered by self-driving cars.
Meanwhile, public transport is a mess, but it's a mess that self-driving cars can help clean up, since autonomous technology for cars can inevitably also improve how buses and trains work. It's not hard to imagine a future where self-driving cars are a form of personalised public transportation. Or where traffic moves so much more efficiently that buses and carpools no longer feel like a strain.
Self-driving cars in general have a long road ahead of them. They face myriad challenges not only technical and legal but also social and psychological. Google's new cartoon of a prototype is almost certainly not what will drive your kids to school in 30 years; it's a proof of concept, not a final product. But a new survey of IEEE members forecasts that 70 per cent of the cars on the road will be driverless by 2040. Nissan says it will bring a self-driving car to market in the form of a £650 upgrade by 2020. Volvo is right behind them.
Even if Google's self-driving car isn't swarming the roads in a decade, this technology will continue to improve in a way that will only make our lives better. And think of it this way: if we can't hope for a future filled with self-driving cars that don't crash, what's our other option? It's certainly not jetpacks.