Yahoo Finance

I want to switch gears here a little bit because you you are the the former Tesla president. And Tesla's certainly in the news, uh, robo taxis on the way. John is yelling Texas coming up here shortly. I'm just curious, John, as you think about Tesla's advantages in that market, how would how would you think about that? How would you frame that it as it goes head to head with Waymo, which of course right now is far ahead?

Yeah, Waymo's got, uh, got, you know, has 10 million rides now under their belt. Uh, they're the market leader. Tesla's going to, uh, launch with what Elon describes as is around 10 cars in a limited, uh, street map in Austin, which I think I applaud. That's the right way to start. Start with safety first and make sure that the passenger's safe is the primary, uh, is the primary objective. But your question on what advantages they have, Tesla's got a lot of cars on the road. Uh, they've got millions of cars in the car park. They're gathering a ton of data. Uh, and additionally, uh, Elon's got one of the biggest GPU clusters in AI. Uh, he's got a 100,000 GPU clustered, uh, AI data center outside of Memphis. He mentioned last week, uh, that they're now trying to build a million, uh, GPU clustered, uh, gigawatt, uh, AI data center. And so that data is flowing into some of the most powerful compute in the world. And so that's a really, uh, that's a real serious of advantages Tesla has. But as the head of, uh, autonomy at Tesla described last week publicly, he feels like they're a couple years behind, uh, behind Google, Waymo. And, um, and that's probably right. They've got some, some real ground to catch up because it's much different, uh, running, uh, driver assistance software than it is running a driverless car.

Well, and I'm also curious where you come out on the whole sensors versus cameras argument, you know, the kind of technologies, um, that each company is employing because all the data in the world is great, but you still have you have a piece of hardware that it has to to power and that has to be safe. So do you have any any thoughts on that?

Yeah, I'll kind of give you both sides of the argument. The the one side of the argument, and this has been Elon's, uh, mantra is that humans drive cars with two eyes. And, uh, and so a car with eight eyes is going to do better, uh, with full-time attention on the road. That's, uh, that's really the Tesla position. On the other hand, there are reasons why, uh, the others, uh, Waymo and the leading Chinese autonomous companies have applied both radar and lidar in addition to cameras. And that's because our eyes can't see well through glare. Our eyes don't see through the dark. Our eyes don't see around corners, uh, and our eyes don't see well around weather. And, uh, cameras have that same challenge. They don't see well through glare, through weather, through the dark, around corners. Lidar and radar, uh, do. And so, uh, from a safety perspective, there's the argument to be made that you really do need those extra sensors if you're going to be a lot better than the human eye, uh, and and responsibly address those weaknesses, uh, just in in in vision, um, versus some of the other sensors. Now, the argument that Elon makes, uh, to counter that is, hey, for every sensor set that I add, I now add delay in, uh, in decision making because there are often times where the lidar sensor doesn't agree with what the cameras are saying. And so then the computer has to decide, okay, which one's right and what am I going to follow? And I need to make this decision like super fast to avoid a potential obstacle. And so his argument is much faster compute, much lower energy cost. We ought to be able to do this with eyes. Uh, the others argue, you can't do it with just eyes because cameras have the same, um, the same failings as human eyes.