HumanoidHub

[HumanoidHub @HumanoidHub]

Neo Beta, making a splash, or, friendly connection 😊

[HumanoidHub @HumanoidHub]

It's slick, yet clunky...kinda childlike in it's learning...Figure02

https://t.co/jPeUBHswHi

[HumanoidHub @HumanoidHub]

From the Chinese...Xbot-L moves along the Great Wall of China, encountering uneven terrain and tunnels. Thankfully, they have plenty of duct tape there, since the guts came loose and it was "Duct Tape to the rescue!"

https://t.co/7wZlZs63IT

[HumanoidHub @HumanoidHub]

Apptronik's Apollo out of Austin, TX, moving some boxes...

https://t.co/n7Tp9BKylS

[Agility @agilityrobotics]

Nothing to look at here. Just Digit working away so reliably that it has become commonplace. #Momentum2024

[HumanoidHub @HumanoidHub]

Here's a previous version of Digit out for a walk...maybe it's a Spring walk in Oregon.

Who knows, maybe next time it will be mowing the lawn.

https://t.co/THWAPhaMs0

[Agility @agilityrobotics]

Let's goooooooo! #FullSendFriday

[HumanoidHub @HumanoidHub]

@agilityrobotics How was the route chosen, or, why did it walk where it walked?

[HumanoidHub @HumanoidHub]

This is Optimus, Tesla's humanoid.

https://t.co/aGncpfvIs3

[HumanoidHub @HumanoidHub]

Agility is located in Corvallis, OR...they created Digit. The latest connection is to a systems integrator, allowing Digit's capabilities to be worked into an entire system of materials handling or logistics applications.

https://t.co/9ZuhujKbee

[HumanoidHub @HumanoidHub]

The current labor U S labor shortage is around 2,000,000 unfilled openings. Filling all these with AI and humanoids would cushion the impact of the coming disruption as the competitors to human labor become increasingly viable.

[HumanoidHub @HumanoidHub]

National Association of Manufacturers estimates a shortage of 'skilled labor' over the next 9 years...scientists, engineers, etc. Maybe AI and robots will fill the gap?

https://t.co/k645OqWSUa

[Jim Fan @DrJimFan]

We trained a robot dog to balance and walk on top of a yoga ball purely in simulation, and then transfer zero-shot to the real world. No fine-tuning. Just works.

I’m excited to announce DrEureka, an LLM agent that writes code to train robot skills in simulation, and writes more code to bridge the difficult simulation-reality gap. It fully automates the pipeline from new skill learning to real-world deployment.

The Yoga ball task is particularly hard because it is not possible to accurately simulate the bouncy ball surface. Yet DrEureka has no trouble searching over a vast space of sim-to-real configurations, and enables the dog to steer the ball on various terrains, even walking sideways!

Traditionally, the sim-to-real transfer is achieved by domain randomization, a tedious process that requires expert human roboticists to stare at every parameter and adjust by hand. Frontier LLMs like GPT-4 have tons of built-in physical intuition for friction, damping, stiffness, gravity, etc. We are (mildly) surprised to find that DrEureka can tune these parameters competently and explain its reasoning well.

DrEureka builds on our prior work Eureka, the algorithm that teaches a 5-finger robot hand to do pen spinning. It takes one step further on our quest to automate the entire robot learning pipeline by an AI agent system. One model that outputs strings will supervise another model that outputs torque control.

We open-source everything! Welcome you all to check out the paper, more videos, and try the codebase today: https://t.co/RwiBT3z78H
Code: https://t.co/ERp4Gl0N36

[HumanoidHub @HumanoidHub]

@benrcrenshaw How did it fall? Force of arms? Force of argument? We need better arguments, perhaps expressed in fictional stories, to captivate hearts, minds and wills.

[HumanoidHub @HumanoidHub]

@Apptronik Here's an article quote:
"That growth will call for even more skilled workers—particularly statisticians...scientists, engineers... industrial maintenance technicians..."

I guess Apollo will have to be pretty smart and well trained to fill those gaps!

[HumanoidHub @HumanoidHub]

SanctuaryAI will be working with Microsoft. So, Gemini will be in Phoenix, while ChatGPT is in Figure01. I wonder why Microsoft didn't go with Digit?

https://t.co/7YwPhxvqcf

[HumanoidHub @HumanoidHub]

@benrcrenshaw We need compelling and persuasive arguments, expressed in philosophy and story that will sway their opinions. One potentially fruitful approach would be debunking the blind faith in progress. Have you seen any good work on that?

[HumanoidHub @HumanoidHub]

@bheater @agilityrobotics @AutomateShow @BostonDynamics @Apptronik @NEURARobotics Excellent! Any chance for live streaming or recording?