Now, three years later, it still hasn't happened. There is no shortage of talk in the automotive industry about this happening very, very soon -- and Alphabet Inc.'s (GOOGL - Get Report) Waymo is widely considered to be in the lead for deploying driverless cars successfully. But ... it hasn't happened yet.
There is no shortage of good demos. There is also no shortage of geo-fenced demos, where the car will perform the driverless task in a very confined ("safe") geography. Furthermore, there is also no shortage of demos where there is a "safety driver" behind the wheel, ready to take over on a split second's notice. Waymo fits very well into both of these key constraints.
However, there is simply no driverless car yet, in the relevant sense: A car that will drive you -- or a blind 5-year-old or 105-year-old -- anywhere, anytime while you're sleeping in the back seat. That's really the only meaningful definition of a driverless car. It's one where there is no human being required to "sometimes" intervene in the driving task.
It is true that we're getting closer to more capable robot cars. It would be strange if we weren't, right? After all, computers are getting ever-faster (GPUs) and sensors are getting better every year (cameras, radars and LIDARs). Match those with improving software, and we are indeed getting closer.
However, look back three decades for a sobering historical analogy: Apple's (AAPL - Get Report) Newton. Pen-based computing was on the cusp of taking over, we were told. All the initial product needed was another hardware and software revision or two and the Newton would have changed computing forever.
While no analogy is perfect, this one is sobering because until we really know how the final product -- in this case, the driverless car -- will perform and how it will be received by passengers and its surrounding traffic participants alike we really don't know if it will be successful.
Let's assume that we get it to "work" -- a driverless car participating in general traffic just like all the other cars. Maybe people simply will think the product is inferior to a regular taxi with a driver. Perhaps it will drive poorly and cost more. Investors seem to take it for granted that the product will be excellent and that people will love it -- just like the Apple Newton three decades ago.
The companies involved in developing driverless cars have huge incentives to exaggerate how close we are to achieving the goal, and the benefits that await at the end of the rainbow. It really should be obvious, but I will spell it out: Without an achievable goal, not too far into the future, the realistic prospect of making money from an IPO or sale to another company, are suddenly a lot lower. There's your very human motive!
That's why you hear endless driverless car optimism from every startup and every established company, including some automakers such as Tesla (TSLA - Get Report) and Waymo. The driverless car will be here soon -- just around the corner! Perhaps in a few months, a year or two for sure. Only one more round of funding at a higher valuation, before we can cash in!
Final nail in the coffin: What happens if we actually get there?
Let's say the product (driverless car) is deemed good or at least acceptable for commercial service, at a price that can generate a profit. Will this be the Nirvana that the driverless car techno-geeks have hoped for?
For an analogy, let's look to drones. Drones were made to work. Are you allowed to fly them everywhere?
No, you are not. Drones are heavily regulated.
Why are they so regulated? Because they are safety risks. No, not in terms of traffic accidents. They are safety risks in the same way we don't allow Al Qaeda to fly airplanes over U.S. territory.
For example, you are not allowed to fly drones near airports. Here is what happened on Thursday, Dec. 20, 2018, at Gatwick Airport in London. You can easily see how even more extensive drone use could halt not just one airport, but all of our major cities. The calls to ban drones even more will likely follow in the near future.
Drones are a threat to security because, among other things, they can contain some sort of a weapon, explosive or chemical/biological weapon. Yet, drones are -- today -- fairly small. They carry a few pounds of payload.
But what about a driverless car or truck? They can carry many thousands of pounds of ... bad stuff.
Terrorists have one major difficulty in attacking American cities: Lack of people on staff. It's not easy to ask someone to blow themselves up in a car bomb.
However, if we ever get to driverless cars, it will be a bonanza for every evildoer in society -- domestically and abroad. That driverless car could be controlled from a basement in Connecticut or from a cave in Afghanistan. Maybe it's not a bomb -- but just something that gums up the Holland, Lincoln and Midtown tunnels entering and exiting Manhattan. The possibilities are endless for this tool to be misused.
See that total traffic halt at Gatwick Airport? Yeah, that's peanuts in comparison to how driverless cars could make drones look like a walk in the park.
Therefore, if driverless cars are ever made to "work" and become operational, it probably won't take all too long until they are banned. It may not happen right away. But once this technology becomes a tool of every foreign power, every terrorist, every other evildoer, the government will have no choice but to ban them.
Driverless cars -- we will have hardly known you!
This is not to say that many of the technologies that go into driverless cars will not continue to become more useful. Cameras, radars and GPUs will continue to enhance the human driver's ability to drive better and more safely. Those investments and deployments are likely going to increase with no evident end in sight.
However, those companies which are basing their business models -- and valuations -- on a truly driverless future are going to be up for a major disappointment: If their product is ever made to work, it will most likely be banned.
That's a problem. And a major investment write-off.
Alphabet can probably shake this off. But can Tesla, Uber and Lyft, among others? Many of these investments are likely to go up in smoke.**
At the time of submitting this article for publication, the author was short TSLA and long GOOGL. However, positions can change at any time. The author regularly attends press conferences, new vehicle launches and equivalent, hosted by most major automakers.