*Robots will replace players one day* Robots:

Red team’s just a bunch of Arsenal players

Seems they are programmed solely based on these duo 🤕

At least they played better than Man U

lol 😂… not in 3001 years

I think they can represent Argentina, now that Paredes is suspended

They have no soul, but neither do they have our limitations. They're already here, and although we may find it amusing to watch them fail now, they'll soon surpass us at everything. We'd better make sure they always remain at our service

Robot cannot do this 😭

Robo red card after celebrating too hard 😂

Robot already practicing dives before its first match

This robot replaces Neymar Jr. very well

You would be surprised how fast they adapt and develop Chill, until you see them in humanoid form without the jerkings and awkward moves

They are butter than Manchester united😭

“Ai will take your job” Me to Ai: This you🌚

China football +made in China robot🤭🤭

No different to having statues . Some of our players don't move all match.

player 5 acting like GTA san andreas player😂😂😂

Wait till they have super intelligence and we have terminators moving around.

@NobaraUvU

Mhm.hmm

Do not make a stuff for pointless reason. If robot replace human for high risk physical jobs, or repetitive job, i am ok. But not for all of thing we use ai/robot, that is stupid

Did you notice there was messj and ronaldo on the pitch??

Penalty a favor del real madrid

Bro I would pay to watch this shit 😂 they just need to label it robot MMA meets football

Hier sind Özbeks Sarr Transfers 😂 @GaddarAbi

THey all look like Bruno Fernandes

Wait for some years and see how the programming is evolving and then come back

Lol 😂 😂

Pues Santiago Jiménez se cayó solito… igualito que los robots. El Layun patea igual la pelota que el robot azul.

😂

They play better than ferran

These robots still have better decision-making than a certain Ballon d'Or winner in big games. PR machine.

One of them robots have already replaced a few players

These humanoid robot soccer matches look like chaotic fun, but under the hood, they represent the ultimate physical stress test for Embodied AI. Moving from rigid, handcrafted code to autonomous athletic locomotion requires solving extreme dynamic physics in real time: 1. Sim-to-Real Domain Randomization Instead of training on physical grass, neural networks play millions of parallel simulated matches with randomized friction, lighting, and sensor latency. Uploaded to physical hardware, the vision pipeline tracks a fast-moving ball at 60 FPS up to 20 meters away with >90% accuracy. 2. 500 Hz Whole-Body Control (WBCQP) Striking a ball on a 30 kg bipedal frame with a leg swinging at 5.5 m/s creates intense angular momentum. Quadratic programming calculates center-of-mass counter-torques 500 times per second, dynamically twisting the upper body and arms to preserve instantaneous single-leg balance. 3. Decentralized Multi-Agent Coordination & 2.5s Fall Recovery Without a central supercomputer, robots coordinate tactical passing via lightweight localized coordinate sharing learned through multi-agent self-play. When collisions occur, passive energy dissipation and ground-to-stand neural networks return the robot to a full sprint in under 2.5 seconds. These athletic breakthroughs are already spilling over into agile warehouse logistics and disaster rescue navigation across unstructured rubble. 👇 Full deep-dive video breakdown in the reply:

Terrible players 😅