Category: Humanoids
Bipedal and general purpose humanoid robotics news.
-
Tesla’s Optimus Gen 3 is Delayed (But Walking Around the Office)
TL;DR: Elon Musk delayed the Optimus Gen 3 reveal to add finishing touches, but confirmed the bots are already walking autonomously around Tesla offices.
So, the Optimus Gen 3 reveal got bumped. Elon Musk hopped on social media to say they need a bit more time for finishing touches. Standard Tesla timeline stuff, right?
But there is a massive silver lining here. Musk confirmed these bots are already walking around Tesla offices completely on their own. No tethers. No remote controls. Just humanoids navigating hallways alongside engineers.
An official reveal is now expected later this April. If they pull off true autonomous locomotion in a chaotic office environment, that completely changes the timeline for getting these things into actual factories. We are watching this space very closely.
-
Why Your Cloud Data is Now Guarded by Robot Dogs
TL;DR: Data centers are quietly buying up fleets of Boston Dynamics and Ghost Robotics quadruped dogs to handle security and routine inspections.
Here is a wild thought. The physical servers storing your emails are now being guarded by robotic dogs.
Data center operators have started buying up quadruped robots from Boston Dynamics and Ghost Robotics in droves. They do perimeter security. They run thermal scans on server racks. They never sleep, and they certainly do not need coffee breaks.
This is exactly the kind of boring but highly valuable use case that robotics companies have been chasing. You do not need a robot doing backflips. You just need one that can walk a perimeter a thousand times without complaining. That is where the real money is.
-
End of an Era: Boston Dynamics CEO Robert Playter Steps Down
TL;DR: After 30 years, Robert Playter is handing over the reins at Boston Dynamics as Hyundai pushes to turn the Atlas robot into a massive moneymaker.
Thirty years is a lifetime in robotics. Robert Playter spent that entire time at Boston Dynamics, eventually leading the company. Now, he is stepping down.
It makes complete sense when you look at who owns them. Hyundai bought Boston Dynamics a few years back. They clearly want to speed up commercialization. Building cool research robots is fun. But selling them is the actual business.
Atlas is the crown jewel here. Hyundai wants that humanoid working and generating revenue. We will likely see a massive shift from wild parkour demos to practical, boring, profitable warehouse tasks. And honestly, it is about time.
-
Figure 02: The Robot That’s About to Steal Your Chores (And Frankly, We’re All For It!)
TL;DR: Brett Adcock, the man who builds companies like most people make a cuppa, is here with his latest mechanical marvel, the Figure 02 humanoid. This isn’t your grandma’s vacuum cleaner; this is a full-blown, AI-powered bipedal wonder, ready to tackle the mind-numbing *drudgery* of daily life. Forget an AI bubble, he says – we’re on the cusp of an ‘age of abundance’ where these metal chaps will be doing everything from the dishes to ordering your chicken salad. Glorious!
Right, gather ’round, you lot! Because if you thought the future was all flying cars and self-stirring tea, you were only half right. The other, far more exciting half, involves a proper bit of kit called **Figure 02**, and the utterly brilliant chap behind it, Brett Adcock. This isn’t just a robot; it’s a testament to human ingenuity, a gleaming, electric beacon promising to free us from the shackles of… well, laundry, mainly. Adcock, a man whose CV reads like a list of ‘things I’d build if I had unlimited genius and a few billion quid,’ dropped by Shawn Ryan’s show with his mechanical mate, and the chat was, frankly, electrifying. Some say humanoid robots were a decade away from doing anything useful. But Brett? He says, and I quote, “We somehow pulled like 10 years of the future forward.” *POW!* Just like that. Because, apparently, sitting around waiting isn’t in his programming. Now, what makes Figure 02 so utterly compelling? It’s not some clanking, oil-leaking hydraulic monstrosity designed by a committee of health and safety officers. No, this is an **electric humanoid**, reasonably priced (eventually, one assumes), and, crucially, powered by proper neural nets. This isn’t pre-programmed parlour tricks; this is *thinking* machinery, learning to do the useful, tedious, soul-crushing human work that we’d all rather avoid. Like, say, folding socks. The sheer audacity of it! Of course, some perpetually worried individuals (and a Patreon fan, bless him) immediately fretted about a Terminator-style uprising. “Safeguards! What about bad programming?” they cry. Adcock, with the calm demeanour of a man who knows he’s building the future, admits it’s an “incredibly complex problem.” Boiling pots of water, small children wanting to jump on the thing – these are the real-world conundrums. But they’re not insurmountable. Indeed, he calls his campus a “problem fun house.” And honestly, if you’re building machines that can navigate a house *without* setting fire to the curtains, you’re doing something right. He’s had them in his own home for months, kids and all. They even name them! “They love it,” he says. That, my friends, is genuine emotional attachment to a bit of engineering! Then came the inevitable question: “Is AI in a bubble?” *Pah!* A bubble? This isn’t a speculative trend, you soft-headed optimists. Adcock’s answer was as blunt as a sledgehammer: “Absolutely not.” He confidently predicts the “most transformative events in technology” will happen in the next 36 months. Not in a century, not a decade, but *three years*. We’re not scratching the surface; we’re just hitting the starting line, engines roaring, for a world with “millions, billions, tens of billions” of these synthetic humans, both physical and digital. And what will we do, the actual humans, when these magnificent machines are doing all the “busy work”? Because, let’s face it, no one *wants* to unload the dishwasher, deal with tax bills, or call the car service. Adcock’s vision is simple, profound, and utterly Clarkson-esque: “I want to be like fully free.” He wants his AI to run his “operating system,” handling all the mundane demands of life, leaving him (and us, by extension) “clear-headed” to pursue what we truly love. It’s about delegating all that “manual labour behind computers or in the physical world.” It’s a compression algorithm for life itself! An “age of abundance” where goods and services plummet in price, and we’re free to… well, whatever we want. Probably watch more robots, if we’re being honest. This isn’t just about robots doing jobs; it’s about redefining what it means to be human in a world where the tedious is handled by tirelessly efficient silicon and steel. Brett Adcock is not just building robots; he’s building a future where even ordering a chicken salad can be outsourced. And that, ladies and gentlemen, is a future I can absolutely get behind. More **POWER** to him! Source Link -
Unitree Proves Hardware Is Good Business
TL;DR: Unitree’s recent IPO highlights strong revenue from quadruped robots, even as the humanoid market remains largely theoretical.
Everyone wants to build a humanoid. Unitree is actually building a profitable business. Their recent IPO filings show real revenue flowing from their robot dogs. They cracked the code on making affordable, capable quadrupeds for developers and hobbyists.
Here is the interesting part. While they make money on the dogs, they are using that cash to fund their H1 humanoid project. It is a smart play. The humanoid market is still years away from mass commercial adoption. You need something else to pay the engineers in the meantime.
Unitree’s approach proves you don’t need to burn billions in venture capital to build a successful robotics company. You just need to ship a product people actually want to buy today. We will see if their humanoid can compete with the big players soon enough.
-
Tesla’s Trillion Dollar Optimus Gamble
TL;DR: Elon Musk thinks Optimus will dwarf Tesla’s car business, pushing a massive scale-up for the humanoid robot.
Elon Musk isn’t exactly known for thinking small. He recently projected that Optimus could push Tesla into a ten trillion dollar valuation. That is a wild number. But when you look at their mass production plans, you start to see the vision.
Tesla has a distinct edge here. They already know how to manufacture complex machines at a massive scale. Their cars are essentially robots on wheels. Applying that same factory logic to a humanoid form factor gives them a serious head start over pure research labs.
The challenge now is software. Hardware is hard, but teaching a robot to navigate a messy human environment is the real bottleneck. We’re watching them train Optimus on tasks like folding shirts and walking through offices. It looks clunky sometimes. But the iteration speed is what we really need to watch.
-
Stretch Takes the Warehouse by Storm
TL;DR: Boston Dynamics is pushing Stretch into more warehouses, proving they care about real logistics just as much as backflipping humanoids.
You probably know Boston Dynamics for Atlas doing parkour. But their real money maker right now is a one-armed box mover called Stretch. They just announced new career pathways and scaling operations for this workhorse. And honestly, it makes perfect sense. Warehouses are desperate for reliable automation.
Look at the numbers. E-commerce demand isn’t slowing down. Human workers get tired, injured, and burned out moving heavy boxes all day. Stretch steps right into that gap. It doesn’t need sleep. It doesn’t take breaks. It just unloads trucks.
This tells us something huge about the robotics industry right now. Flashy humanoids grab the headlines. Purpose-built machines pay the bills. Boston Dynamics gets this. They’re heavily investing in making Stretch faster and smarter.
-
Unitree H1 Slashes the Humanoid Price Tag
TL;DR: Unitree is bringing bipedal robots to the masses with the H1, an embodied AI powerhouse that costs less than a luxury SUV.
Here’s the problem with most humanoid robots: they cost as much as a house. Unitree is out to fix that. Their H1 robot is stepping into the ring with a price tag under 90k dollars, completely flipping the script on who can afford to play with embodied AI.
Don’t let the lower price fool you. The H1 is incredibly capable. It moves with a terrifyingly fast gait and packs a 3D LiDAR system to map its surroundings in real time. It’s built for developers and researchers who need a solid hardware platform without bankrupting their departments.
By commoditizing the hardware, Unitree is doing for humanoids what DJI did for drones. They’re lowering the barrier to entry, which means we’re about to see an explosion of new software and applications built on top of these affordable metal frames. The robot revolution won’t be a monopoly.
