CBoston Dynamics just showed off a new hand for Atlas— four fingers, no pinkie, and nearly double the dexterity. Thirteen degrees of freedom.
MNo pinkie? That's a bold choice. But the real number everyone's talking about is 25,000.
CA post on X from an account called Arboryx says Hyundai plans to deploy that many Atlas robots across its plants, including Kia factories.
MThat's an internal target, not a booked order. And this hand is still a prototype. We've seen demos before— remember all those backflips? Putting a robot in a factory is a different game.
CBut this time the hand actually matters. Thirteen degrees of freedom means it can grip tools, turn knobs, handle cables. That's the bottleneck for humanoids.
MAnd they dropped the pinkie because four fingers are meant to be more robust. It's clever engineering, but the rollout target is a big claim, and Hyundai hasn't said when.
CThey said "plans to deploy", not "sold". But Hyundai owns Boston Dynamics outright, so they can set internal goals. And if any company can push robots into factories, it's a carmaker— they already automate everything.
MCarmakers also famously overpromise on automation. Tesla's been promising a robot army for years. Where is it?
COkay, fair— but Hyundai is actually testing Atlas in real plants, not just YouTube videos. That's different.
MThey are testing, I'll give you that. But let's not call 25,000 a done deal. It's a target, maybe a stretch.
CAnd the hand design itself is smart. Removing the pinkie reduces complexity and failure points. Four fingers get the job done for most tasks.
MTrue, but fine manipulation often relies on that little finger for counter-pressure. They'll need to prove it can work with small parts on an assembly line.
CThat's the exact test Hyundai is running. They're not just doing parkour; they're putting Atlas next to workers to see if it can pick up a wrench without dropping it.
MBut even if the hand works, scaling to 25,000 units means producing them at a cost that makes sense. Humanoids are still absurdly expensive.
CCost comes down with volume. Hyundai has the manufacturing muscle. If they commit, they can drive the price down like they do with cars.
MCars are a mature industry. Robots are early. The gap between a demo and a reliable factory worker is still huge.
CThen what's your benchmark? Because by that logic, we'd never adopt any new automation.
MMy benchmark is a pilot that turns into a paid contract. Show me a company buying Atlas units at real scale for real money, and I'll get excited.
CWell, while we're on robots, here's a rumor that'll start a fight. A robot fight MMA account on X is claiming Boston Dynamics is exploring a Nasdaq IPO, hot on the heels of Agility Robotics listing.
MHold on— that's completely unverified. There's no credible reporting on that. Boston Dynamics is wholly owned by Hyundai, and I've seen no filing.
CThe account says "reportedly", so it's a rumor mill thing. But think about it: a pure-play robot stock would be huge for retail investors.
MBut the Agility Robotics listing they mention— that ticker AGLT? I couldn't confirm that either. This is just two unverified claims stacked together.
CEven if it's not true today, the idea has legs. Robotics companies need capital, and public markets are hungry for AI-adjacent plays.
MI'd want to see a company with real revenue and a clear path to profitability, not a headline. Boston Dynamics has cool demos, but Spot and Atlas are still niche.
CThey do have Stretch for warehouses, which is actually shipping. But fine, the IPO rumor is probably fake. Still, the sector is heating up.
MLet's stick to facts. The hand is real, the 25,000 target is real as a target. The IPO is not real yet.
CAnd that's the story: Hyundai's robot push is real, the stock market frenzy is not— at least not yet.