One step closer to recreating the greatest commercial ever piped.video/taxldmY8V-Y?is=3_Ms…

Fantastic progress from @SkildAI.

Robots are on the path to RSI. At Skild, we have deployed them for bringing up real datacenters, automating factories, and building critical infrastructure!

Robots can now play soccer…. 😭😭

AlphaGo, but for humanoid soccer. Skild AI brings self-play to humanoids. Their flagship robotics foundation model, S1, can now be post-trained with no human demonstrations by competing against itself in simulation. The testbed is soccer, which demands balance, agility, ball control, reactivity, and strategy. Given a single objective (score goals), S1 played 140+ years in NVIDIA Isaac Sim and went from falling over to dribbling past defenders, shielding, tackling, shooting, and getting back up. None of these behaviors were hand-rewarded; they emerged because they helped it score. Its opponents (past versions of itself) improved alongside it, creating an automatic curriculum. With 4 agents, passing and coordination started to emerge. Skild says the approach extends to social navigation, collaborative manipulation, and any multi-agent task with a simulator and an objective.

Think self-evolution is really a key towards professional-level sports intelligence…. Really great results!

AlphaGo defeated humans. When will humanoid robots do the same? Skild AI is cultivating the "Lionel Messi" of the humanoid robot world. Interestingly, they had the Unitree G1 engage in self-play for the equivalent of 140 years within a virtual World Cup setting. Then, in the real world, it autonomously plays soccer against humans--intercepting the ball, defending, and dribbling on the attack. Moving from the virtual to the real world eliminates the need for the linear, decade-long training required for a human soccer player; instead, the robot can gain over a century's worth of experience in the simulation in a very short time(perhaps just a few weeks) It’s soccer for now, but imagine the future possibilities if this extends to construction sites, factories, supermarkets, and households… That is what true scalability looks like.

The scale of experience you can generate in simulation is ridiculous. Now the interesting problem is making the simulated world physically representative enough for that experience to transfer.

@SkildAI’s #messinator is cute but here is the actual tea from the blog

Cannot say how much I enjoy seeing this! Now time to up the team size! piped.video/fkPbkJI_IKE?is=OoC0…

Sim2Real is so back baby!!!!

it's always fun to see robots playing soccer. still remember i accidentally tackled someone pretty hard in my last soccer match at cmu iykyk😜

Found the next G.O.A.T. It just needs charging. 🐐🔋

Humanoid football had a big year. For comparison, this is what the world finals of RoboCup 2026 looked like just last month in South Korea. Should we compete next year? ;)

Handheld camerawork is hard work: a workout for you, a contact sport for everyone else. 🎥😬

Whatever the objective, you’re always competing for some limited resource... either with members of your own species or in predator–prey interactions. Treat the entire ecosystem as a single system, and it’s all just self-play! Alyosha, @phillip_isola, @chris_lu, and I spent countless months debating this during my PhD… no conclusion, alas.

When I was a kid waking up at 3am to watch Messi, I definitely didn’t think I’d somehow end up working on robot soccer years later. We originally wanted to get this video out around the World Cup final to cheer for Argentina and Messi. We’re a little late now, and Messi has already said goodbye to the national team. So I guess this can be my slightly late small tribute to the player who made me fall in love with football in the first place. I’ll save most of the technical details for later — there’s a lot more we want to share. For now, just a few thoughts on why I’m excited about this direction. AlphaGo and AlphaStar showed how far self-play can go in digital worlds. One question I’ve always found really interesting is: What happens if we bring the idea self-play into the physical world? With robots, you can’t just come up with a good strategy. It also has to actually work through balance, contact, noisy perception, hardware limitations, and an opponent that doesn’t behave exactly the way you expect. That’s what makes self-play especially interesting to me in robotics. It’s not just about getting better at a game. It gives us a way to study how embodied agents can keep adapting through interaction and discover more robust behaviors on their own. And the bigger vision here is definitely more than a robot soccer demo. I like to think of it as Real-world FIFA: robots actually learning skills, tactics, competition, and eventually cooperation — instead of just replaying motions we designed for them. To me, soccer is a really fun testbed for embodied intelligence because everything comes together in one place: perception, control, decision-making, and multi-agent learning. We’ve shown robot-vs-human. Expect to see 1v1, and then more players. Really excited to see where this goes. More to come. 🔵🔴🇦🇷⚽️🤖

Seeing this makes me excited about the potential for humanoid robots as consumer products. The sheer versatility is exciting. People might not pay $20k for a robot that just does laundry, or just dusts. But imagine it can do a few things, and then every month, a software update ships with completely new capabilities. One month you get the soccer update. The next month you get the landscaping update. Then the pool cleaning update. Etc. For me personally, a skill-level-adjustable soccer playing humanoid is worth at least $5k of value.