CA high school robotics team built a two-year-old an electric wheelchair. For free.
MWait, a whole wheelchair? That's a serious build.
CThe family couldn't afford the twenty thousand dollar chair, and insurance wouldn't cover it. So the team just... made one.
MThat's a story worth paying attention to. Kids with a 3D printer and a motor controller solving a real problem.
CAnd it's not a toy. It's a functional chair for a toddler. The post is from Massimo on Bluesky, and it's been making the rounds.
MI want to know how they handled the safety side. A wheelchair for a two-year-old isn't a science fair project.
CThat's the thing. The post describes mentors, engineers from local companies. Not a bunch of kids winging it.
MStill, the liability alone would scare off most adults. These kids just did it.
CAnd that's why it stands out. The best robotics stories aren't the billion-dollar demos. They're this.
MSpeaking of billion-dollar demos鈥攖here's a forecast floating around that says the US will have hundreds of thousands of humanoid robots by 2030.
CFour hundred thousand? That's a lot of robots wandering around looking for a job.
MThe post cites JP Morgan, via Reuters. I couldn't confirm the exact numbers, so treat it as a claim, not a fact.
CBut the direction is real. Every major player is pushing humanoids鈥擳esla, Figure, Boston Dynamics, plus a whole wave of Chinese companies.
MHere's my problem: we've been hearing 'robots are coming' for a decade. Where are they?
CThey're in warehouses. They're in factories. They're just not in your living room yet.
MBut hundreds of thousands of humanoids by 2030? That's not a warehouse arm. That's a walking, talking machine.
CAnd the cost is reportedly dropping fast. A humanoid that was a million dollars five years ago is now under a hundred thousand, according to reports.
MCost isn't the bottleneck. It's reliability. A robot that falls over in a factory is a lawsuit.
CSo you think the forecast is fantasy?
MI think it's optimistic. And a bank publishing a boom forecast has its own reasons to be bullish.
CThat's cynical. The technology is genuinely improving. Look at what Optimus was doing two years ago versus now.
MI'm not saying it won't happen. I'm saying 2030 is aggressive. Maybe 2035.
CFine. We'll check back in four years. If I'm right, you owe me a coffee.
MDeal. But if I'm right, you're buying.
CDifferent world entirely: Amazon's home robot, Astro, is getting roasted on Bluesky.
MErin Fogg posted that the best Amazon can do is a robot that doesn't vacuum, only spies on you, and may or may not be equipped with missiles.
CThe missiles part is a joke. But the rest is... actually accurate.
MAstro doesn't vacuum. It has cameras. It patrols your house. And people are supposed to pay for that?
CAmazon positioned it as a home companion and security device. But the value proposition has always been fuzzy.
MA robot that watches you but doesn't clean anything. That's a surveillance camera on wheels.
CAnd it's expensive. For the same money, you could get a robot vacuum that actually does something.
MThe joke lands because it's true. Amazon built a robot that solves a problem nobody had.
CMeanwhile, the high school kids built something that solves a real problem for a real family.
MThat's the whole industry in two stories. One side builds for headlines, the other builds for people.
CAnd one more from the robotics world鈥攁 researcher named Eugene Vinitsky posted about self-play motion planners running in the wild, in collaboration with Torc Robotics.
MTorc is the autonomous trucking company owned by Daimler. So this is real-world deployment, not a lab demo.
CThe interesting part is 'self-play.' The AI learns by competing against itself, like AlphaGo did.
MThat's how you get robust behavior. The system has seen every edge case because it created them.
CAnd it's running in actual trucks. That's a big deal for autonomous freight.
MThe future is multi-agent reinforcement learning, he says. I think he's right.
CIt's the same technique that could make humanoids actually useful. Let them practice in simulation until they can't fail.
MThat's the missing piece. Not more motors, not better batteries鈥攂etter brains.
CAnd that's why the JP Morgan forecast might not be so crazy after all.
MDon't start. We already made the bet.