Here are some predictions about self-driving cars, pulled together from my tweets over 2016. They are based on the assumption that vehicles which can navigate public streets without a human driver are now technically feasible and will be deployed in many parts of the developed world over the next decade or two.
- Surge in private car purchases
There has been a common type of thought piece in startup-land which asserts that the widespread rollout of autonomous vehicles will cause people to use taxi-app services like Uber instead of owning cars. This flurry of publishing has happened because the people who write think pieces about autonomous vehicles live in urban tech centers and are heavy users of taxi-app services.
The reality is that many more people in the U.S. live in geographies where owning a car is the most practical solution. When self-driving automobile technology is available it will cause these regular commuters to buy a new car sooner than they would otherwise because, if you drive in traffic multiple times per week, a car that can drive itself is 10x better than one that can’t.
2. Eventually, cars will look different
The look and shape of vehicles that we see on the road will change as the market for autonomous vehicles matures, eventually becoming quite different from what we’re used to today. This is because design constraints are very different for a vehicle without a human driver.
There is no way to accurately predict what autonomous vehicles will look like, they will evolve through selectional feedback as a variety of different concepts are tried out in the market and the successful ones are rewarded with more time and money.
3. More commuters
As autonomous vehicles take over the most unpleasant parts of commuting they lower the holistic cost of commuting (i.e. the cost including psychological and other factors). People generally find a commute that takes over 1 hour to be unpleasant, also they find driving in heavy traffic to be unpleasant. Self-driving cars remove the latter factor. This lower cost will increase the aggregate appetite people have for commuting, leading to more commuters.
4. A societal reference point for interacting with AI
There was much talk in 2016 about artificial intelligence reaching an inflection point where it matches the general cognitive ability of human beings. The reality is that current applications of artificial intelligence are “more optimized for specific problems than most people realize”, to use the words of Mark Zuckerberg. Still, the embodiment of an AI into a universally recognized physical form (the car) will make autonomous vehicles a popular reference point for society’s understanding of human interaction with AIs.
As autonomous vehicles develop more advanced features, e.g. recognizing and remembering things about their occupants and environments, they will eventually enter the set of things we recognize as possessing consciousness, regardless of whether we understand the science behind consciousness or not. That will be a major societal inflection point in its own right, surfacing previously abstract moral questions in the real world.