Tesla is betting on vision-based self-driving without LiDAR, unlike Waymo's cautious map-heavy approach. Elon's plan: deploy semi-autonomous cars to collect billions of miles of data, train on edge cases, then flip a switch to full autonomy. Waymo has logged 250,000 rides weekly since 2009 with two reported crashes.
Questions this video answers
2009.
29 cameras plus LiDAR and radar.
250,000 rides per week.
Two crashes.
Nine cameras worth $1,000 (total).
Chapters
Tesla's got a massive problem with its self driving tech.
I'm in my Tesla on autopilot, going 40 miles an hour towards a fake Wiley Coyote Roadrunner painted wall. Please stop. Please stop.
But no one is talking about what happens when they fix this issue. Because Waymo's success is an illusion. And Tesla is on track to leapfrog them with a strategy to deploy a fleet of robo taxis in every city around the world. All within the next three years, we'll.
Have hundreds of thousands, if not over a million Teslas doing self driving in.
The us Made possible by el playbook that's continuously defied the odds and skeptics. First, let me introduce you to who everyone thinks is the market leader. Imagine you're in downtown San Francisco. You open an app, tap a button and a car glides up. No driver, just empty seats waiting for you. That's Waymo. It started inside Google back in 2009. Today it's also live in cities like Austin and Phoenix. Each car is a rolling super sensor, 29 cameras plus LiDAR and radar. It fires millions of laser pulses every second to build a 3D model of the street. It can see shapes, judge distances and track movement faster than any human. But the real magic isn't the hardware, it's the pre mapping before a single passenger steps in. Waymo vans have already driven those same roads hundreds of times, recording every traffic light, curb and road marking. When a Waymo robotaxi hits the road, it's not just reacting, it's cross checking live data against a blueprint. It's like deja vu for robots. The results are hard to argue with. Waymo has only been involved in two crashes since launch. It's now handling 250,000 rides per week. On paper, Waymo looks like it's already won the race. But the question isn't does it work? It's how fast can they take this global? That's where things get interesting.
In the opposite corner, there's Tesla. No lidar, no radar, just nine cameras worth $1,000. Instead of custom mapping cities, Tesla's cars figure it out on the fly. It's like dropping someone in Rome and telling them to drive. Tesla calls it vision based AI.
The solution that we have is AI and vision, so there's no expensive equipment needed.
It's a system relying only on cameras and neural nets to interpret the world. Unlike Waymo's cautious rollout, Tesla ships over the air updates to users every few weeks, all while collecting billions of miles of real world footage. At first Glance. Waymo's tech looks superior. Cameras have less range than lidar. They struggle in rain, fog, or at night. To be fair, this isn't apples to apples. Waymos are fully driverless. Teslas are still semi autonomous. But that's the point. Waymo is a valedictorian who studied too hard. Tesla is a dropout who's already making money.
LIDAR is a fool's errand. And anyone relying on lidar is doomed.
So how is Tesla getting away with it? The answer is liability hacking. Since Tesla's system is still semi autonomous, it's classified as driver assistance. That means the human behind the wheel is still legally responsible. If something goes wrong, it's your fault. Most of the time it's not Tesla's. If Tesla had gone the Waymo route, they'd be on the hook for every crash. But by offloading liability to the driver with partial autonomy, Tesla gets to avoid approvals, delays and lawsuits. That's not just clever legal work, it's what unlocks Tesla's biggest advantage in AI. Real world data at scale.
It seems like the actual debate should be about like, do you have the fleet or not? So data collection systems. Yeah, do you have a fleet or not is significantly more important. Whether you have lidar or not, it's just another sensor.
Waymo's launching like NASA. Slow, precise and expensive. Tesla's doing it like SpaceX. Ship fast, then learn from failure. Here's Tesla's four step strategy. One, deploy semi autonomous vehicles. There are already 500,000 on the road. Two, collect billions of miles of real world data. Three, train the AI on edge cases lab testing can't replicate. Four, flip the switch, instantly upgrading the entire fleet overnight via a Software update. That's 6 million Teslas turned into fully autonomous vehicles. Right now, Tesla is deep into steps two and three. But step four won't happen until they hit an AI safety threshold. This is where governments agree Tesla's AI is statistically safer than a human driver. Tesla estimates it'll take about 6 billion miles of data to clear that bar. This year, a model Y delivered itself to a customer. Welcome to your new Tesla. The robo taxis went live in Austin, already covering more ground than Waymo Service.
Tesla has rolled out its robo taxis.
In Austin, but would you feel safe getting in? And their semi autonomous fleet has logged 3.6 billion miles. At this pace, they're two to three years away from flipping the switch and leaving everyone else behind. But will the public trust Tesla's technology?
What Elon Musk is doing is endangering public roads that are already a very dangerous place.
Elon's endgame isn't owning a fleet. He wants you and your car to become the fleet. Tap a button and your Tesla joins a pool of robo taxis making money while you sleep. Think Uber, but you don't have to drive the car.
Just like Airbnb, you can rent out your spare bedroom or your house if you're not using it, and make money on it. And that's what we expect Tesla customers to be able to do.
It could turn every Tesla into an appreciating asset, justifying higher margins, pricing, power and long term valuation. This is Elon's masterstroke. When Autonomy arrives, millions of Tesla owners will have money riding on it. They'll advocate for regulatory approval. Maybe not out of belief, but out of profit. It's the same playbook Uber and Airbnb used. Build something people love and earn from, then regulators will have little choice but to legalize it.
And the bigger we get, we know, the more attention we draw from regulators. But I also think the bigger we get, the more regulars say Airbnb are partners to us.
Wall street gets it. Tesla's trillion dollar valuation assumes autonomy works. If it fails, that value disappears. But investors aren't just betting on autonomy. They're betting on a story they've already seen. Tesla's robotaxi strategy follows a pattern Elon has used before. Look at SpaceX. Critics said he couldn't build better rockets than NASA, but that was never the point. He made space exploration cheaper by creating reusable rockets after years of blowing them up and learning. SpaceX now handles 80% of the world's payload. It's the same with electric cars. Elon didn't care that batteries were weak or charging took too long. He bet that if he built a car people wanted, the infrastructure would follow demand. That's why Tesla started with a $100,000 sports car, not a budget sedan. Elon's pattern is that his companies look like expensive failures until they suddenly work. His robo taxi strategy is the same play. Waymo might beat Tesla in precision, but Tesla is betting it can beat everyone in distribution.
