Figure 03 Reveals 2026 Humanoid Robot Endgame (AI NEWS)

There are three or four on display cases on the walls. And here’s figure CEO discussing the timeline for robots building other robots on moonshots with Peter H. Diamandis. We will put robots on our bocce lines this year, and then phasing like basically like phasing humans out of there will be a combination of getting more robots there and doing more high volume like automation. And then you know we will be at a point, I hope in 24 months, where all the robots will build all the robots. Then he reveals figure’s upcoming home deployment schedule.

I think by end of year we will be able to put a robot into an unseen home and be able to do fairly long horizon work. In 26 we launched a robot to do like end-to-end homework like an alpha testing like in my home. I think I just want a robot in my home I can talk to you. I can remember things such as my kids, my kids come home like sad from school or something. I went through about to understand that. I have the EQ like self-awareness to see that, talk to them.

Like I think all this is like something we want to spend more, we’re spending more time on now internally. And he says superhuman speed and agility is coming. We basically have another three to five X in terms of speed and actuators that we can run. The software is not enabling because it doesn’t know how to do it yet. So we can run like three to five times faster than what you saw today. Because we have the whole body to run. We look at it like in terms like radiance a second.

Maybe we traditionally look at RPMs. We look at radiance a second here. We have another like three to five times headroom in the actuators that you’re seeing now. Leading to robots performing surgery this year? 26 we’ll be able to do like from a hardware work what surgeons can do. I feel pretty confident by the end of this year you’ll have a hardware system that you know you can basically if you could like teleoperate or something like that you could like basically be able to do like real surgery. He also predicts the future of AI automation in the workplace.

I mean I think it’s relatively straightforward to think that every human should have a humanoid to do all your work. And then we should have maybe an order of like five to seven maybe ten billion in the commercial workforce. If we could solve a general purpose human robot today that could I think we should have a billion today. And even using humanoids in space? I think we really love to run in space. We would be able to put unlimited amount of capacity of humans at it through humanoids. The next 12 to 18 months I think will be like the largest AI transformation we’ve ever seen.

But at some point you’re gonna walk out probably in San Francisco be the first and you’ll see more humanoids than humans. And I think that’ll be an amazing day. Meanwhile here’s the latest from Unitree’s CEO. It’d be interesting to know how well choreographed these performances are. Would this work on any wall or just this one? Plus they also released this video of the G1 doing swarm acrobatics. They say that this is not AI generated. Many people online have been saying that most of these demos are fake but comment down below whether or not this looks to be completely authentic.

So it looks like robots are either moving in the direction of performative entertainment or hands-on real-world tasks. Make sure to comment down below which you would prefer. Meanwhile researchers from Nvidia as well as various universities just unveiled a groundbreaking generalist robot world model called DreamDojo which has been developed to understand and simulate complex physical interactions, dexterous control, and real-world dynamics. So DreamDojo learns its comprehensive physical knowledge by first pre-training on huge human video data sets using continuous latent actions which is a clever proxy to handle unlabeled videos. Then it does post training on smaller target robot embodiment data with precise continuous robot actions but all of it hinges around the DreamDojo HV data set which is the largest video data set ever used for world model training with 44,000 hours of diverse egocentric first-person human video which is 15 times longer duration and 96 times more skills and 2,000 times more scenes than the previous biggest data set for this purpose.

So because of this it covers a huge range of everyday environments thousands of objects and complex human behaviors from grasping and pouring to folding assembling and even handling cluttered real-world scenes which gives the model broad human-like physical intuition. So after post training DreamDojo then generates highly realistic action condition video rollouts for multiple humanoid or generalist robots like the GR1, G1, AGI bot, and YAM. So when it comes to results they’re pretty clear as DreamDojo dramatically outperforms previous baselines and thanks to its massive human data pre-training it handles contact-rich precise interactions much more realistically and in terms of real-time performance DreamDojo achieves fast real-time generation at 10 to 11 frames per second which enables smooth long horizon rollouts up to one minute long simulations without sacrificing any quality and some of the downstream applications include DreamDojo being able to be used for tel operation allowing humans to use telepresence to complete real-world tasks remotely plus it enables reliable policy evaluation and simulation and model based planning.

Altogether by leveraging human scale video data instead of robot fleets DreamDojo paves the way for a truly general purpose robot world model better generalization to new objects and environments faster robot development and progress towards open world dexterous humanoid robotics at scale. But tell us down below which robot you think is the most useful and check out this video of Reflex HQ’s robot doing what could be the most useful real-world tasks so far and tell me in the comments down below how much would you be willing to pay for this and here’s the robot working in a factory and in just a second you’ll see it working with a human as well anyways like and subscribe and check out this video here for more of the latest in AI and robotics news
[tr:trw].