Autonomous Over-Under: 2020
How soon will we have practical autonomous cars? There are so many factors that determine timing:
- How safe are the systems compared to human drivers?
- How fast will ride sharing companies embrace them?
- How will competition force acceleration of deployment?
- How will regulators enable or block a deployment?
- Will cities compete to be first?
I was chatting with some friends about the space, and we were inclined to place bets on when we might see a real deployment. What is “real”? Here are some rules:
- Available to end consumers, not just a tech demo.
- In at least one American city.
- The car comes to you without a driver in it.
- The car autonomously drives point-A to point-B in the city
There are some enabling technologies that might work under the hood that I don’t think impact the experience. For example, it might be a monitored drive where a remote tele-operator can wrest control. It might be a city that has been completely pre-mapped. The drive might be slightly different, e.g. more diligently obeying traffic rules than humans. It might only operate during business hours in good weather; The general degree of availability in a city shouldn’t be part of a bet.
One of the best ways to bet is over-under. It’s easy to say that this probably won’t happen in 2017. It’s also easy to say this is inevitable by 2067. So where is a the disagreement boundary?
I’m betting before 1/1/2020.
The cars are demonstrably safe for reasonable point A to point B trips already. Cities will compete to signal support for new technology. Many companies are competing in the space. Consumers would love to spend less on rides. Tens of thousands of people are killed or injured by terrible human drivers every year.
This is definitely a risky bet. Where is your over-under?
Before deciding, watch this video of a Cruise autonomous drive through San Francisco at night. Try counting how many human traffic violations like illegal u-turns you can spot.