Voice‑driven, real‑time arbitrary action generation😁 Using external voice commands, G1 is directly controlled to generate a wide range of actions in real time. This video was recorded in a single take, with on‑site audio recording. Because the actions are autonomously generated by AI in real time, there may be... slight latency, and the smoothness of the movements may be somewhat reduced.show more

If I’m going to be murdered in my sleep by an autonomous robot, it should at least have tits.

To bad it is all chorographed. Ask it to do something it has never dine before. Go find a refrigerator, get me a can of coke and bring it to me. Thats a real trueing test

This looks like cgi. Anyone else?

#Skynet

No you don't, R2D2

And to unbox it you can use our voice agent on smartglasses

If my robot can’t make me a sammich then what’s the point.

Why do none of these ever show the robot actually doing work?

This is insanely cool, but the language understanding still seems a bit rough, when she asked it to propose it first replied “okay, wish you happiness” before doing the kneel 😂

Awesome progress! Another 12-18 months you'll be able to ask it to make your breakfast to order and set the table- time it to be ready at 8am Done!

ChatGPT has a staggering 48% hallucination percentage. But sure let's train robots, with narrow AI. What can possibly go wrong. 😅

Voice-to-action in a single take 👏

Kinda wild how fast it responds, feels like we’re getting close to something actually useful.

Oh look! Our new prison guards, teachers/professors, doctors, vetrenarieans (if there are any pets left), store clerks, maids, factory workers, architects, psychologists, JUDGES, government "entity", doing tricks! NO THANK YOU!! We do not consent!

I just seen one of these bust their shit tryna go up some stairs.

Does it speak English, and can it do something useful like lift a box?

🥱 Wake me up when I can say, “Do the dishes, put away all the toys, vacuum the floors, strip all the beds, do the laundry, fold and put away all the clothes and make all the beds with the clean sheets, then mow the lawn and wash and vacuum both cars.”

Awesome. How long until it is doing chores. Cooking, cleaning, ect. No free lunches round here.

Voice-driven real-time action generation is impressive despite slight latency. G1 shows promise for autonomous AI movement control using natural speech commands. Great single-take demo. #Cybernorse

Maravilloso. Estamos cerca de que la orden sea: “Deja de hacer tonterías, sal ahí fuera y recoge los tomates que ya estén maduros de las tomatera 😅

It may be janky and not useful now, but it is a complex system and it will rapidly evolve in the next few years.

Useless. If you say "clean the table" can it do it? If not, then go back to work and make a robot that can be used for real world tasks.

Its going to come faster than we thought and its going to suddenly hit so hard

Which one is the robot?

LOL Those that make them wish we were more like them because if we were they wouldn't need to make them!

is open SDK available?

it’s impressive The most important part is real time voice to action control. If G1 can hear an external command and generate movement immediately, that is a strong step toward natural robot interaction. People will not want to control humanoids with code, joysticks, or menus forever. Voice is the most human interface.

Yeah but we still dont need him to do push up show me a vid of him cleaning my cat liter or cleaning toilets, then I will buy 3 or 4 of them

希望多一些 中国企业来推特上做一些宣传，让世界多了解一些

This is so AI generated that it isn’t even funny. Go back to the drawing board, use actual cameras with no manipulation next time. And also, stop fetishisibg SE Asian women in short skirts. Thats just gross.

Ok, now G1 punch the Agibot

With the level of advancement happening in robotics and AI, humanity should also focus on building technologies that can help prevent, reduce, or manage pollution more effectively.

"Bust thru that glass door and swim in the pool!"

We're curious as to which company will be the first to make the voice changeable. Voice to action is great but getting the same sounding response every time is draining