Latest / eMotors: Electric Revolution / How Tesla's FSD Fuels the Robotics Revolution
Transcript
- 0:09When you hear the name Tesla, what comes to mind?
- 0:12For most people, you're probably thinking about those sleek
- 0:14electric cars. Right.
- 0:16Maybe a Gigafactory or some new battery tack.
- 0:18I mean, they're the company that basically forced the entire car
- 0:21industry to go electric. And if that's all you're
- 0:23focusing on, how many cars they ship each quarter, you are, and
- 0:27I think this is fair to say, fundamentally misreading their
- 0:30entire long term blueprint. OK, let's unpack this.
- 0:34Our deep dive today, pulled from their latest strategic briefings
- 0:37and some internal estimates, shows a mission that is way more
- 0:41radical than just a new car in your driveway.
- 0:44The car is almost like a Trojan horse.
- 0:46It's the testing bed, right? Their real game is building the
- 0:49core components, the AI, the robotics for a completely
- 0:55automated world. So we're going to pull out five
- 0:58key insights that really go beyond that surface level.
- 1:01And we should start with the most, I think, undeniable
- 1:04evidence of their AI, the stats from the road right now.
- 1:07Exactly. I mean, the proof that their
- 1:09central bet might be paying off is in this massive pool of data
- 1:13they've collected. And the claim is startling.
- 1:15It is the claim is that their self driving AI is already right
- 1:20now statistically safer than an average human driver.
- 1:24That's a huge claim, but the sheer volume of data they have
- 1:27is, well, it's hard to argue with.
- 1:29We're talking about over 3.6 billion miles driven with FSD
- 1:34supervised. And that's the key part, right?
- 1:36St. supervised, where a human is still meant to be monitoring
- 1:39things. Still paying attention, but the
- 1:41comparison numbers are. Wow, they're pretty striking.
- 1:45The data shows one serious collision for every 5.1 million
- 1:49miles driven with FSD on. 5.1 million miles and what's the
- 1:53number for? Just, you know, regular manual
- 1:56driving. In North America, you're looking
- 1:58at a serious event roughly every 700,000 miles.
- 2:01It's not a small difference, it's not even close.
- 2:03It's an order of magnitude. Their internal estimate, and
- 2:05this is for the latest software, is that the AI is up to 8.5
- 2:09times safer than the average human.
- 2:11Which sort of signaled this massive cultural shift coming
- 2:14doesn't. It it does, if these trends
- 2:16hold, we're getting to a point where letting a human Dr.
- 2:19manually on a highway might actually be seen as the reckless
- 2:22choice. Not the default.
- 2:23Not the default, yeah. But this, the future, this whole
- 2:26idea is built on a very specific strategic choice.
- 2:30It's a huge gamble really. A philosophical one with.
- 2:33A vision based approach. This is where you see the big
- 2:36split in the industry. Tesla is betting everything, and
- 2:40I mean everything on cameras, basically replicating the human
- 2:45eye. And a very advanced neural
- 2:47network to act as the brain, right?
- 2:49Their argument is, well the world is built for human eyes
- 2:52and human brains, so an AI should be able to navigate it
- 2:55the same way. But that puts them in direct
- 2:57opposition to pretty much every other major player.
- 3:00You look at companies in China, BYD, Xiaomi, Huawei.
- 3:03What are they doing? They're doing the complete
- 3:05opposite. They champion what's called the
- 3:06multi sensor approach. It's all about redundancy
- 3:09certainty. Like giving the car multiple
- 3:11sets of eyes. Exactly.
- 3:12They use cameras, yes, but they fuse that data with LIDAR, which
- 3:17gives you these incredibly precise 3D point clouds of the
- 3:20world and radar and radar, which is phenomenal for long distance
- 3:23detection and you know, doesn't care if it's pouring rain or
- 3:27foggy. It just works.
- 3:28So if we break down that trade off, Tesla is choosing
- 3:31scalability, lower hardware cost and these really rapid software
- 3:36first updates. While the competitors, they're
- 3:38choosing reliability, yeah, robustness, they're layering on
- 3:42hardware for safety, redundancy. It's more expensive though, and
- 3:45a lot more complex to get all those different data streams to
- 3:48work together. It is.
- 3:49It's a classic battle really, a software first path versus a
- 3:52hardware centric path to autonomy.
- 3:55But then you have to ask, if the software is already 8.5 times
- 3:59safer, why risk the whole company on a path that seems
- 4:03teams on the surface a bit riskier in say a Blizzard?
- 4:07There has to be a bigger reason for going vision only.
- 4:10There is, and this brings us to the real identity of the
- 4:12company. The reason they chose Vision
- 4:14isn't just to save 10 grand on a lighter sensor.
- 4:17It's because the car isn't the final product.
- 4:20The future of Tesla isn't a car, It's the humanoid robot Optimus.
- 4:24And that just changes everything.
- 4:25Suddenly you realize the autonomous car business has been
- 4:28this perfect, incredibly demanding training ground.
- 4:31The. Required first step.
- 4:33To build a general AI that can operate in the real physical
- 4:37world. This is it.
- 4:39They were an AI and robotics company disguised as a car
- 4:42company all along. And their leadership has been
- 4:45very explicit about this. Musk himself has gone on record
- 4:48saying Optimus will eventually be, what, 80% of the company's
- 4:51future value? 80%, just dwarfing the car
- 4:54business. Completely.
- 4:55The goal is it's radical mass production of 1,000,000 humanoid
- 5:01robots. And the purpose is to tackle the
- 5:03global labor crisis. They're targeting those boring,
- 5:07repetitive and sometimes dangerous jobs.
- 5:10The stuff in manufacturing and assembly that's too complex for
- 5:13a simple fixed machine. It's a solution aimed at
- 5:16systemic issues like an Asian workforce or skill shortages.
- 5:19And it leverages their expertise, not just an AI, but
- 5:22in complex high volume manufacturing which they learned
- 5:25from building cars. But here's where it gets really
- 5:27interesting. What makes this whole plan even
- 5:29possible, let alone disruptive, is the cost.
- 5:32This is the jaw-dropping part. The estimate for Optimus at
- 5:35scale is a selling price between 20 and $30,000. 20 to 30,000 for
- 5:40a humanoid robot, that's the ultimate disruption.
- 5:43When you look at the competition, I mean, what do
- 5:45other robots cost? Well, a specialized robot like
- 5:48Figure O2 is estimated at well over $100,000.
- 5:53Some search platforms can go up to 1/4 of a million, so.
- 5:56Tesla is basically using its car supply chain to commoditize a
- 6:00market before that market has even really started.
- 6:03And that low price is completely dependent on that big bet they
- 6:06made earlier, the vision only approach.
- 6:09So it all connects. It's total synergy.
- 6:12Optimus uses the exact same core AI hardware, the same software,
- 6:16the same supply chain that's already in the cars.
- 6:19So every mile a car drives is literally training the robot's
- 6:22brain how to see and interact with the world.
- 6:25Which makes a robot cheaper and faster to develop.
- 6:27And that brings us full circle back to the safety data and this
- 6:31this core ethical dilemma that regulators are facing.
- 6:33The human cost of delay. And this raises a really
- 6:36important question for you, for the listener.
- 6:39How do we weigh the known massive risk of human error on
- 6:44the roads against the perceived risks of a powerful new
- 6:47technology? The statistics are forcing that
- 6:50conversation. You look at independent studies
- 6:52on other systems like Waymo. What did they find?
- 6:55They found an 83% reduction in serious accidents, the kind
- 6:58where an airbag deploys, and a 92% reduction in personal
- 7:03injuries compared to human drivers.
- 7:05When you know that human error causes most traffic deaths
- 7:08globally, delaying something that is demonstrably safer, it
- 7:11means. Thousands of avoidable
- 7:13casualties every single year. It forces complete re
- 7:16evaluation. The challenge isn't just
- 7:18preventing tech failure anymore, it's about actively enabling
- 7:21life saving progress. So.
- 7:23If we tie all five of these threats together, the safety
- 7:25data, the vision only bet Optimus, the low price, the
- 7:28human cost, the take away is clear.
- 7:31You have to look beyond the quarterly car reports.
- 7:33You really do. This is a foundational AI and
- 7:35robotics company. They're redefining what a tech
- 7:37company even is. They're not just trying to sell
- 7:39you a better car. No, they're trying to sell you
- 7:41the future of Labor and mobility itself.
- 7:44Their biggest innovations in the next decade probably won't have
- 7:47anything to do with transportation.
- 7:48It'll be about solving labor shortages with affordable, mass
- 7:52produced robots. So what does this all mean for
- 7:56you? As AI drivers get statistically
- 7:58safer than us, and these affordable robots are getting
- 8:02ready to enter our factories, maybe even our homes, are you
- 8:06truly ready for the world that's coming next?
- 8:09A world where the human driver is the outlier.
- 8:11And your Co worker, the one loading the shelves cost
- 8:14$25,000. It's.
- 8:16A lot to think about. We hope this deep dive gave you
- 8:18a valuable shortcut, a way to get informed about a company
- 8:21that is clearly thinking 10 steps ahead.
- 8:24So thank you for joining us. If you enjoyed this journey,
- 8:26please follow our channel and check out our other deep dives
- 8:29for more insights like this. This rapid progress in autonomy
- 8:32forces all of us to think critically about it, not with
- 8:35fear, but through the evidence and the very high stakes
- 8:38implications for work and safety they're already unfolding around
- 8:41us.