It started an hour and two months late, but it happened. Tesla, the company that led the way in the EV revolution, is taking another leap of faith. Elon Musk is a master at figuring out where the trends are and marketing to those. And AI is hot right now. He wants to keep people thinking of Tesla as the company that is pushing the envelope and is really a tech company, not just an automaker. Elon Musk essentially decided that the future of Tesla was not about cars. Like a thing with four wheels and a steering wheel that you drive. That the future was autonomy. The most important thing, as he sees it, is the Robotaxi, or what Tesla is calling the Cybercab. And its gambit targets not only the car industry but ridesharing as well. Though some big kinks need to be worked out first. Watching the event, my first impression was - is this just another Musk rug pull?

The human experience, as reflected in driving, is more complicated than people like Elon Musk thought. It has turned out to be a very hard problem. Next, Tesla’s CEO Elon Musk unveiling a Robotaxi. Really a difficult moment. Make or break for Tesla. The moment of truth. That's putting it mildly. Ahead of the October 10th announcement, Tesla had a market value greater than the likes of Toyota and General Motors, despite selling fewer cars. Its shares had been trading at more than 80 times forward earnings. If that means nothing to you, consider the firmly established AI winner Nvidia trades about 36 times and General Motors in mid-single digits. Basically, expectations couldn't have been higher. Investors were really piling into the stock in the lead-up to this event. As of April, Tesla shares were down more than 40% for the year. They've come all the way back from that. Now, the promise of the self-driving car has been in the public sphere for ages.

Speed control and guidance become automatic with the flick of an instrument. What I think of as a self-driving car is a car where you could take your hands off the wheel, go to sleep, have a drink, whatever. And it takes you where you are going with zero intervention. Which is exactly what Elon Musk envisions for Tesla's new vehicle. But some form of this is already here. Tesla has one of the best driver assistance systems in a car. Tesla sells software known as autopilot. It also sells another version of software that's called FSD for full Self-Driving. This is a system that will essentially allow you to make fewer of the sort of minute-by-minute decisions of driving as long as you are there, ready to take over at any time. All these car companies basically have some version of this. But Musk's new vision takes things further. An autonomous car that doesn't just drive on its own,

but can operate without you, the owner, even being in it. The idea is like a combination between Uber and Airbnb. If I own a Tesla, it can drive me to work. But if I'm on vacation and I want to make money, I can like lend my Tesla to this shared autonomous robotaxi fleet. So instead of your car being this like depreciating asset, it's actually becoming, you know, like more and more valuable. Tesla, because they are basically designing most of the hardware in the car and writing most of the software, has an ability to essentially like, send software updates to their cars. The same way that Apple does with their iPhones. So the idea is that at some point, you would get this over-the-air update that would turn your car into a robotaxi. The Model 3 and Y will achieve unsupervised full self-driving with permission in wherever regulators essentially approve it in the US, and then to follow in outside the US.

The end customer is the passenger. Essentially, it's the person who presses the button on the still, you know, not released Tesla robotaxi app and pays money for that ride. And then how the money gets divided up is kind of an open question, but it's probably some combination of the person who owns the actual hardware and then Tesla itself. And if this sounds a little too good to be true, maybe it is. Someone is going to have to maintain these cars. I mean, these vehicles are going to need to be cleaned. They're going to need to be serviced. They're going to need to be charged. To like, have a national, a car that could go anywhere at any time at like the kind of scale that Elon Musk is talking about, like, this is going to be a huge and complicated problem that regulators have not even, like, begun to really grapple with. The regulatory regime in the United States,

There are sort of what's considered five levels of autonomous driving. Level one being the lowest level, level five being the highest. Tesla cars that are on the road today have what's known as level two, which means that legally and morally like the human driver is supposed to be in control at all times. The company that is really leading the way in terms of autonomous driving so far has been Waymo, which spun out of Google. Waymo, very slowly and very methodically rolled out their technology in specific geographic areas, and they very carefully mapped the terrain of the cities where they're operating. It's called geofencing. The difference between Tesla and how other automakers approach autonomy is its use of light detection and ranging, or lidar technology. That's like a big sensor suite that helps to kind of map the like immediate environment around the vehicle. Tesla does not use LiDAR. It's frigging stupid. It's expensive and unnecessary.

He says, you know, you got two eyes. There's no laser in your brain that's doing this. Therefore, cameras should be good enough to drive a car. So there's like an ideological component. But I think it's really a practical component because Tesla is selling a consumer good. And these LiDAR sensors, like a Waymo car, would cost hundreds of thousands of dollars if you tried to build it. So Tesla can't do that because it would lose a fortune. So they're using just cameras and software. They basically just unleash their software on their consumers and gain a lot of information from the data that they collect from their fleet. Tesla would argue that their real-world data is more robust than the data that, say, a Waymo collects. You know, not all data is created equal. And Tesla cars are typically more focused on freeway driving, and they maybe have not done as much testing in cities.

But we really don't know, because Tesla has not lifted the veil on exactly how much data they collected. Tesla's approach of collecting real-world data that constantly educates a powerful AI to drive its autonomous fleet is incredibly ambitious, but conceptually it isn't so different from the way humans learn. American drivers are, by and large, not that great. We have roughly 40,000 deaths a year because of speeding, distracted driving, and drug and alcohol use. Tesla and these other autonomous car companies are saying our technology will sometimes make a mistake. It might even be a fatal mistake. But in the long run, humans also make fatal mistakes. The problem there is when a mistake happens, whether a human driver would have done it. Are we, as a society, going to be okay if we introduce a new fatality, like autonomous cars plowing into kids? That's going to be like a complicated ethical argument.

that people are going to have to wrestle with. We've seen a lot of accidents because a lot of Tesla drivers get lulled into a sense of complacency, and they think that the car can do more than it really can. We've seen lawsuits. I mean, one thing that really messed up cruise the GM company and has been a factor for Waymo, is the fact that it's not just about being safe. It's about like not messing up the flow of traffic. So when can the world expect the robotaxi revolution to begin? Well, in his own words: Well, I tend to be a little optimistic with time frames. But end to end in 2026. So, yeah, before 2027, let me put it that way. But the Cybercab's development and deployment roadmap will be a significant new test of shareholder and customer patience. Cybercab was in many ways what we expected.

Investors I spoke to pointed out something important. There was not much discussion of the business model. It was thin on detail and thin on discussion. There was no mention of China and very little about the United States. This was a company that needed more affordable vehicles in order to keep up the growth that everybody was getting used to, and we just don't see any indication that that's actually coming. The $25,000 Tesla that, you know, Elon Musk promised everybody back in 2020 that got scrapped, that got put to the side in pursuit of this Robotaxi. But to the true believers, this is a guy who rescues astronauts from the International Space Station and lands rockets on drone ships. No one ever believed that that was possible. So, you know, there's a lot of adage I hear in Silicon Valley elsewhere like, don't bet against Elon. Love him or hate him, he's definitely somebody who's having a huge impact on the world.

and also who is pretty unpredictable. He's betting the company on this technology. It's a very, very risky bet.

Why Elon Musk’s Robotaxi Is Such a Risky Bet for Tesla

Bloomberg Originals offers bold takes for curious minds on today’s biggest topics. Hosted by experts covering stories you haven’t seen and viewpoints you haven’t heard, you’ll discover cinematic, data-led shows that investigate the intersection of business and culture. Exploring every angle of climate change, technology, finance, sports and beyond, Bloomberg Originals is business as you’ve never seen it.


Why Elon Musk’s Robotaxi Is Such a Risky Bet for Tesla

Scripts