Cruise was the first Autonomous Driving Unicorn, and it had already won $1 billion when General Motors bought it earlier this year. This Memorial Day weekend at Thunderhill might be sunny, but GM's Cruise acquisition was a solar eclipse under which a dozen-odd teams had shown up to test, show off and chat about their products, hoping that among the spies lurked a manufacturer’s emissary to bestow the horned bandana of riches.
“GM got ripped off,” Hotz told me. He’d say it to anyone who asked, but he and the Comma.ai team were the only ones laughing, because in the race to become the next Unicorn, pretty much everyone thinks Hotz is ahead of the game.
“Does anyone believe anything Hotz says?” I asked Handsome Startup Bro and his team over weak coffee and cheap eggs at the free Holiday Inn Express buffet that morning. He and crew were among the two dozen people I interviewed, most of whom wanted to remain anonymous, unless they were talking about their own products, which are always in development, or already great, or both.
“I don’t think Hotz believes everything he says,” Handsome Startup Bro laughed. His team laughed along with him.
“Hey, Spaniard,” I said. “Does Hotz believe everything he says?”
The Spaniard nodded. The Spaniard isn’t in the Self-Driving Car sector, I think, but wants to be, I’m sure. Or maybe he’s just a spy. Why else would he have come to this godforsaken town? I’m not even sure he’s from Spain. I don’t speak Spanish, but The Spaniard likes Gladiator, as I do, and speaks Machine Language, which I don’t, and everywhere we went he asked questions, and people answered, and I listened.
“It doesn’t matter what Hotz says,” said The Spaniard. “Look at what he’s doing.”
Handsome Startup Bro and team fell silent.
Everyone in the tech sector—Google, Apple, Uber, Lyft, Faraday and myriad multi-billion dollar Chinese companies—believes Self-Driving Cars are inevitable. Now that Tesla has thrown down in a big way, everyone in the automotive sector—all of whom claim to have been working on such technology for years—is scrambling to bring Self-Driving Cars to market.
You can’t buy a Google car, but you can buy a Tesla with Autopilot, which operates at what’s called Level 2. (Some would argue it's Level 3. As the current semi-autonomous Cannonball Record holder, I'd call it 2.5). It will drive itself most of the time under most conditions, but the driver must be ready to take over anytime. Tesla’s Autopilot is the best such system available today, but Volvo and Mercedes are allegedly months away from releasing systems as good or better, depending on how one parses press releases.
Until then, it’s Tesla or nothing. The next stop in the road to Autonomy is Level 3, where the car can do all the driving under all conditions, but the driver may still have to take over. Then there’s Level 4, where the car can do everything, under certain circumstances, and the driver can literally take a nap.
While car companies spent the last ten years investing in hybrids or horsepower, the tech sector ran away with Autonomy. Publicly, car companies are clustered at or below 2, and Tesla & Google are close to 3. Everyone wants a piece of the theoretical pot of gold lurking at Level 4, which is why, if you build cars and can read a newspaper, the Build Or Buy question answers itself.
There are only so many startups in the space, almost all of whom are focused on components of Autonomous Driving. If you’re behind, or think you’re behind, and you want to leapfrog to Level 3 or beyond, you want to snap up whomever’s got an end-to-end solution.
Cruise? Their $10K rooftop Autonomous Driving retrofit for Audi A4s never got to market, and GM had, according to my usual sources, “shit.”
Shit. And a billion dollars lying around.
A billion dollars may seem a thick, long, thorny cactus to swallow, but it’s nothing compared to the billions GM and VW paid out for foisting trash on their loyal customers.
Which brings us to Hotz and Comma.ai. For an infinitesimal fraction of the time and money car manufacturers have invested on development and acquisition of Autonomous tech, Hotz, in less than a year, has built a Self-Driving Car. And that was before he raised his first $3.1M angel round.
Hotz’s plan? To offer a $1000 Autonomous Driving retrofit kit by the end of 2016, that “you’re going to be able to install it yourself, if you can set up a piece of IKEA furniture.”
Will it be Level 2? Level 3? Press him on details, and it sounds like Tesla’s Autopilot, but for the car you already own. Ask him how they differ, and Hotz is clear. “A Tesla costs $100,000. Our product will work on cars costing $20,000.”
That’s quite a value proposition, but can he have products in stores by December 31st of this year? As in actually ramp up production and deliver units to customers like my mom? In seven short months?
If that were possible, I’d give him money now. As in right now. But he’s not asking for deposits, which is where Hotz differs from Elon Musk.
“No way,” said The Spaniard.
No way, said the faces of everyone else I asked, all of whom insisted on anonymity.
Which brings us to the genius of Hotz. He’s not selling widgets or even the promise of widgets. He’s selling himself and the collective hilarity, dynamism and genius of his five-strong(!), barely post-pubescent Comma.ai team. He doesn’t need to sell a single unit, maybe ever, to get to his big win.
He certainly doesn’t have the in-house knowledge to rapidly build a widget company, and he probably doesn’t want to. All he needs to do is make enough noise, and wait for bids for the whole kitty. Announce Chffr, their mobile app? Genius. It’ll be in beta until they’re acquired. The year-end deadline? Brilliant. Miss it? Doesn’t matter. No direct competitors. If and when they show up, they’ve got to do battle with George Hotz, the myth. For a kid who claims to hate marketing, he’s the greatest marketer since Steve Jobs. The Hotz myth is a swamp of talented people drowning in his charisma, and his cup overfloweth.
End capitalism! I don’t even vote! Machines will be better at everything! Human Driving?
Cute! Quaint! What would I do with money? It’s not about money. It’s about winning.
Every outrageous statement is unconscious but totally calculated, stepping stones on the path to a payday he’s positioning as one of the biggest of all time.
Let’s do the math on Hotz’s plan to win.
Cruise raised $19M before their $1B buyout.
Hotz has only raised $3.1M, and would appear to be close to Cruise’s level of development at the time of their acquisition. If Hotz can privately demonstrate a working retrofit kit without raising any more money,and it actually is as easy to install as a piece of IKEA furniture, Comma.ai is worth at least $1B.
At least. And maybe a LOT more.
The mind bends. The jokes stop.
If he shows it publicly via livestream—totally the kind of thing Hotz would do—he might even get two billion, and we’ll see how long he lasts in a corporate environment. Something tells me old-think corporate culture would be glad to see him go, and he’d be glad to leave. I hope it’s the Germans. I’d love to watch his first post-firing interview. They won’t have to fire him.
He’ll walk first.
Or maybe the conservative minded car manufacturers will find him exasperating, and the deal will fall through.
No bids? He does another fundraise, hires some adults and starts building kits. Legal issues delay release? He can license out the tech until hurdles fall. Someone will need it. Who else is building retrofit kits? No one I’ve heard of. While everyone else is trying to build component parts or whole cars, Hotz is working on his win-win scenario.
All Hotz has to do is keep up the press and win the small battles. Battles like getting his Acura around Thunderhill 100% Autonomously, as publicly as possible.
“Hey, George,” I asked before I got into the backseat, “what about the spies?”
“I’m sure there are plenty of spies, but as long as they keep the mentality that they have, they’ll be incapable of competing with us. If Honda sends a spy here and says ‘that’s our car, how did they do this?’ don’t worry, they’ll never ship.”
“They’re just keeping tabs on the space,” said Jake Smith, COO, Head of PR, Chief Test Driver and catch-all expert whenever Hotz ran out of breath or laughed too hard. “They’re probably there to realize how hopeless they are.”
“But,” I said, “suppose these guys sniffing around the car figures it all out?”
“Naaaaaaaaaah,” said Hotz. “The OEMs think in a five year development cycle. OEMs are actually incapable of shipping any technology that is newer than five years old. It’s true. They think in that development cycle, so, you know, maybe five years ago that’s what autonomy was, but five years have passed.”
It was time.
“This is just for fun,” said Hotz. “We’re here to race for fun.”
From anyone else that would have been a lie.
But Hotz was telling the truth. Hotz could get a job anywhere. Consult anywhere. But here he is, up against Brobdingnagian forces, without a care in the world. He’s past money. Look at any picture or video of him. Listen to what he says. He genuinely doesn’t care. When Zuck or Musk or Gates show up in public, we’re getting the distillation of a thousand stylists and media handlers.
Hotz? He’s an icepick in the face of public relations.
$1000 for one of Hotz’s kits? I’d pay $1000 to watch a Hotz reaction video every time Mary Barra does a shareholder call and utters the word disruption.
“Hey, George,” I said, “Are we gonna get 100%?”
“Realistically? Probably not.”
“Did you ever play any driving games growing up?”
“Need for Speed back in the day. I had the wheel. I understand the basics.”
“So let me understand this,” I said. “Yesterday you drove one lap, and on that lap you taught the car how to drive this track.”
The Acura rolled out of the pits, Hotz at the wheel. “Something like that.”
“And do you have any actual racing experience?”
“None at all,” said Hotz. “Three, two, one...engage. We are on Autonomy mode.”
He removed his hands. The steering wheel began to shake in five degree oscillations, the invisible hand of early-stage Parkinson’s upon it, and the car gingerly began crawling around Thunderhill West at 35 mph.
“Staying pretty close to the center of the track,” said Hotz. “This is obviously not the racing line.”
It wasn’t. Not even close. But for a car whose $50,000 GPS had failed, it did a pretty good job of getting around the track...like a terrified old woman going blind and being forced to give it a shot, at gunpoint.
“Hmmmm,” said Hotz. “I think I’ve seen six-year-olds drive around racetracks probably similar to this.”
We approached the final turn, the car slowed and...put two wheels into the dirt.
I was deflated.
Hotz wasn’t. “Come on, car!”
The same thing happened the second time.
“Come on, car!”
Hotz only needed it to work once. With all four wheels on the track. At any speed. To “win.”
Then, one un-air conditioned, never-ending, stifling, head-lolling, cheap Japanese velour backseat in the Summertime lap later—just as I was about to vomit in my helmet—it did.
The lap time? Nobody cared.
Hotz? Not Charlie Sheen.
Hotz and the merry band at Comma.ai will return in The Chronicles of George Hotz, featuring the adventures of Eder, Jake, Riccardo, Niel and Sam, whenever I get around to it.