Five key questions from Tesla's Cybercab/Robotaxi event
I have a lot more, but let's start with these.
At Tesla's "We, Robot" event last night, Tesla CEO Elon Musk rolled out a lot of the same promises he's been making for years, all the while mixing in some new-to-him autonomous tropes that others in the industry have been citing since, well, General Motors Motorama talked about the joys of self-driving in 1956.
GM predicted that would be a reality in 1976. Elon Musk? He says his self-driving concept, dubbed Cybercab or Robotaxi, will ship in 2026.
There was a lot of hype, yet watching the event, and re-reading the transcript this morning, I was amazed at how little Musk actually said.
I was, unsurprisingly, left with more questions than answers. Here are my top five.
Where and when can I use my Cybercab?
"You will be able to buy one, yes exactly, um, and, uh, we expect the cost to be below $30,000."
- Elon Musk
This was among the most interesting aspects of the event, that Tesla will indeed be selling these directly. This is in contrast with the most successful autonomy providers of the moment, Waymo and Cruise, both of whom operate exclusively in a ride-hailing manner.
That is to say, you can't buy their cars.
Why not? Well, for starters, they're way too expensive. Waymo hasn't said what the sensor package on its current machines costs lately, but it's a safe bet that the sensors alone cost more than the $30,000 figure Musk quoted for the Cybercab.
Now, I love the idea of being able to buy my own autonomous vehicle. What I don't understand is exactly how any of this will work. Today, autonomous vehicles are only allowed in few, rigidly defined areas.
Does that mean that the Cybercab will only be available to people living in certain Arizona and California counties? If so, isn’t that a very small market for a new car?
Is Musk betting that his increasingly expensive political ties will force more states to allow more autonomy between now and 2026? If so, doesn't that fly in the face of state's rights?
Also, in what conditions will the Cybercab be able to operate? Will people really pay $30,000 for a car with no steering wheel that leaves them stranded whenever a little snowfall obscures the markings on the road?
And, if that's the case, why wouldn't someone just pay another $10,000 for a Model 3 that works in all conditions, yet will supposedly offer the same autonomy features?
How will any of this work?
"I think there'll be an interesting um, you know, business model where, like, let's say somebody is an, uh, you know, Uber or Lyft driver today, uh, they, where they can actually sort of manage a fleet of cars and, like, a sort of, manage, I don't know, 10, 20 cars and just sort of, you know, take care of them like a, like a shepherd, uh, tends their flock."
- Elon Musk
If we're really going to be creating our own privatized ride-hailing fleets, how will that work?
Will Tesla develop its own app? Will you be able to list your own car as an Uber driver? Who will have liability for your car’s behavior in this situation?
These may sound like minor details in the grand scheme of things, but it's taken years for services like Uber and Lyft to establish themselves in markets across the U.S. and figure out the legalities of it all. While Tesla could surely follow in those footsteps, much like new EV startups are following in Tesla's when it comes to direct sales, putting this infrastructure in place is a non-trivial thing.
Who's going to clean my car?
"The vast majority of time cars are doing nothing."
- Elon Musk
This is very true. Musk said that cars are, on average, used 10 hours a week out of 168 hours. That's actually a higher utilization rate than I've seen elsewhere, but the crux is the same: Cars are a huge waste for the most part.
Musk's seeming solution is adding your personal car to a ride-hailing fleet. But, if you've ever been in the back of a well-used Uber, you know that's not a particularly nice place to be. Will you want your car to smell like that? To look like that?
Tesla showed off a little video of robotic arms vacuuming M&Ms off from seats and purging half-empty water bottles. I'm going to need assurances of more strenuous detailing and sanitizing before I'd let my car run the graveyard shift in a college town.
Haven't we heard all this before?
"If you buy a Tesla today, I believe you are buying an appreciating asset, not a depreciating asset."
- Elon Musk, 2019
Musk has been talking about autonomy for a long, long time now. He's been wrong every single time. I pulled that appreciating asset quote out because in that instance he wasn’t just wrong about timing, he was wrong about everything.
According to KBB, the value of a 2019 Model 3 in good condition, in a private party sale, is as little as $15,700 today. That's far less than half its cost in 2019. Even if priced on the high end, $19,625, you’re still nowhere near turning a profit.
Musk has made all the same claims about safety and other autonomy-related benefits in the past. And, as I mentioned before, others have made the same assertions for decades.
Can Tesla ever make autonomy a key part of its business?
Our goal is, and I feel pretty good about this goal, that we’ll be able to do a demonstration drive of full autonomy all the way from LA to New York, from home in LA to let’s say dropping you off in Time Square in New York, and then having the car go park itself, by the end of next year. Without the need for a single touch, including the charger.
- Elon Musk, 2016
Musk seems stuck in an autonomy rut. Back in 2016, he promised that Tesla cars would drive coast-to-coast by 2017. Now, in 2024, he's saying it'll happen by 2026.
Back then, that was a throw-away line designed to wow the masses. It was successful. Today, with Tesla facing ever-more competition on the EV front, Musk seems intent on pivoting Tesla’s primary business away from being a vehicle manufacturer and into becoming an autonomy provider.
His comments last night, then, weren’t disposable. Musk is connecting the future of Tesla's business and its viability to his autonomy timeline, which has been a year away for nearly a decade now.
There's an old aphorism in software development that seems apt to me this morning: Writing the first 90 percent of the code takes 90 percent of the development time. Writing the last 10 percent of the code takes the other 90 percent of the development time.
Tesla may have 90 percent of self-driving figured out by now, but that last 10 percent can be a real killer.