Last week, I published a piece of speculative fiction about how I saw Google and Uber evolving as companies and taking advantage of new technology 10 years from now. That “Dispatch From The Future” got quite a lot of attention, in part because more than a few people got very excited about the prospect of driverless cars and the impact that they can have on the way we think about transportation.
Since a lot of people read it as the future is now, and others have said that my ideas about what the world will look like in 10 years might actually be conservative, I thought it might make sense to explain some of my thinking around my outlook for the driverless car.
The story was based mostly on Amir Efrati’s story on Google’s interest in possibly manufacturing driverless cars and so-called robo-taxis, as well as the big Google Ventures investment in Uber that was finally announced the week before. The idea was to take those pieces of news and think about how the companies will evolve over time. So here goes.
Let’s start with Google. My piece last week presupposes not only that Google has built a fully electric driverless car by 2023, but that it’s actually on its third line of commercial vehicles by that point. Note that in my piece, Google doesn’t announce its plans to manufacture its own vehicles until 2018 or so, five years from now — which I think is a fair amount of time for it to figure out the logistics of autonomous vehicles and get them to the point where they have a five-nines safety rating.
You might think that Google wouldn’t want to manufacture the cars itself, as it mostly deals in software. And one could see Google developing autonomous driving software and then licensing it to third-party manufacturers, like Ford or Toyota. But if those manufacturers weren’t interested, it also wouldn’t be completely unprecedented for Google to get into the manufacturing game itself.
Just look at its acquisition of Motorola, or its development of Google Glass, and you see that the company increasingly has developed an appetite for making things. Things you can touch and feel, and not just the software that powers them. When it comes to cars, I could see Google doing the same thing. It’ll probably want to control its destiny in manufacturing the first car that runs its software.
In addition to manufacturing the cars, I believe Google will also build infrastructure to support fully electric vehicles, kind of like Tesla’s supercharger network. But if you have a driverless car, you also need a driverless way to power up. I picture some sort of fully automated Roomba-like dock for cars these PowerUp base stations.
As for the types of driverless vehicles that Google would produce — I don’t know why, but I like the idea of Google making autonomous vehicles built around actual utility — like the Google GX10000 two-seat commuter vehicle or the four-seat robo-taxi model. I just don’t think it’ll want to build the traditional five-seat family sedan that appeals to your usual car owner. But that’s just me.
So what about Uber in 10 years? For one thing, it’s a public company, according to my report from the future. But that shouldn’t come as a huge surprise. The company is just three years old, and is just printing money.
One interesting wrinkle from my story is the idea of Uber as a “transportation and local delivery service” — meaning that somewhere along the way it made at least some of its drivers delivery guys. Of course, there’s no shortage of folks fighting over the local delivery market, including startups like Postmates and Instacart, as well as big Internet players launching services like Google Express and eBay Now.
None of them have built the logistics framework that Uber has for routing rides in the last few years, however. That’s why I don’t think it’s crazy to assume that Uber will also be adding delivery services in several markets over the next few years. It’s already experimented with ice cream, barbecue, and mariachi band delivery. It seems like only a matter of time before it gets serious about partnering with local merchants and retail chains to power on-demand drop-off of goods and services.
Another interesting tidbit about Uber was the future impact that I believe the company will have on traffic and congestion in major cities in the future. I personally haven’t owned a car in at least 10 years, having lived in New York and San Francisco — two cities where the combination of good public transportation and bike friendliness make that doable.
But take a look at Uber’s piece on the economics of car ownership, and you see that going car-less and just relying on Uber isn’t such a crazy idea. I think others will, over time, figure out that it’s actually more convenient and less costly to take Uber than to own their own cars. Ten years should do it.
One of the critiques of my piece last week was that 2,500 driverless cars isn’t a whole lot when you consider the size of taxi fleets in most cities. Uber would surely want more, right? Well, for one thing, it’s a trial run.
But for another — think about the efficiencies that come with cars that don’t take shifts, don’t need to sleep, know the fastest route to pick up the nearest passenger, and can charge themselves in their downtime? I’m guessing that 250 autonomous vehicles could probably do the work of 1,000 cabbies in some cities.
The Regulatory Environment
One of the big knocks against my piece is that the future might actually come sooner than I expect, that 2023 is actually a pretty conservative estimate for robo-taxis. But then I think about all the legal and regulatory wrangling over services like Uber today, I think about the taxi lobby and the insurance lobby, and I wonder if 2023 might actually be too soon for some of these changes to take place.
I don’t doubt that, once the technology is proven and that autonomous vehicles actually prove safer than human-driven cars, there will be states like California that introduce special driverless car lanes on their highways. It only makes sense, right? Get the cars that can be trusted to steer themselves away from those that can’t, and let them go as fast as they can. This would also become an incentive for adoption of the technology, as there are real advantages to driverless cars — fewer accidents on the road, less traffic, etc.
That said, it’s one thing for governments to let users have autonomous cars that could be pulled out of autopilot mode and driven by hand. It’s a whole other thing to have cars without drivers scooting people around town. I think that’s a dicier proposition, and one that will probably be held up in regulatory review for a while.
I have no doubt that robo-taxis will someday arrive, just maybe not as soon as we think (or hope).