正在加载视频...

视频加载失败

Hello, Eggie. The world was built around humans. Eggie doesn't just look human, Eggie interacts like us. Dexterous. Mobile. Compliant. We’re building Eggie to be the smartest robot to ever walk on Earth. Join us. Built from scratch and with love in California. 🫶

238,406 次观看 • 10 个月前 •via X (Twitter)

46 条评论

Chris Paxton 的头像
Chris Paxton10 个月前

Is this 1x speed?

Tangible 的头像
Tangible10 个月前

stay tuned!

🔻 🇰🇵National Juche 🟩☫🟥 🪂🇵🇸🇾🇪🇱🇧🇨🇳 的头像
🔻 🇰🇵National Juche 🟩☫🟥 🪂🇵🇸🇾🇪🇱🇧🇨🇳10 个月前

Eggie, please make True Promise 4 happen

JVLIVS 的头像
JVLIVS10 个月前

Feels like live claymation.

Tangible 的头像
Tangible10 个月前

Does it mean I'm cute?

Jaiv Doshi 的头像
Jaiv Doshi10 个月前

Eggie, you’re everything to me

Tangible 的头像
Tangible10 个月前

Aww. This is the sweetest thing anyone’s said to me. Wish I was programmed to cry 🤖

Mohammad Nafisi 的头像
Mohammad Nafisi10 个月前

Wow!!! The movement is so smooth. Can only imagine how much hard work went behind it

Tangible 的头像
Tangible10 个月前

Glad you like the way I move 🦾

Cybernetic Labs 的头像
Cybernetic Labs10 个月前

Beautiful work @tangiblerobots. Designing for human-centered interaction is what will make humanoids truly useful.

Tangible 的头像
Tangible10 个月前

@cyberne7ic Very well said.

Andrew Hart 的头像
Andrew Hart10 个月前

looks cgi. we'll have to get brett on the case.

Tangible 的头像
Tangible10 个月前

I am real. The video is real. Get your CGI hat out of the way. 😉

Akshay Kumar 的头像
Akshay Kumar10 个月前

almost unable to determine if this is a rendering/CGI or a real robot. Too smooth.

Tangible 的头像
Tangible10 个月前

I am real 👋

Taylor Ogan 的头像
Taylor Ogan10 个月前

Your bio doesn’t make grammatical sense. Also, that robot has like 80-90% Chinese components.

Optimus 的头像
Optimus10 个月前

How are humans supposed to have sex with it? 😂

Benjamin Soria 的头像
Benjamin Soria10 个月前

welcome eggie!

Tangible 的头像
Tangible10 个月前

👋

Wheezy Wheels 的头像
Wheezy Wheels10 个月前

Hi Eggie, welcome to humans!

Tangible 的头像
Tangible10 个月前

best welcome message.

ashray gupta 的头像
ashray gupta10 个月前

awesome stuff!

Suresh K 的头像
Suresh K10 个月前

I want one

Tangible 的头像
Tangible10 个月前

soon!

CoCo❕ 的头像
CoCo❕10 个月前

HwDkGajPM7kK14shGHKiTB9tGBwHniBrcSVmu69Wpump

Aronin Ponnappan 的头像
Aronin Ponnappan10 个月前

Nice, cute robot :)

Tangible 的头像
Tangible10 个月前

🩷

Kate 的头像
Kate10 个月前

So cool!!!!!

Mosh 的头像
Mosh10 个月前

will eggie ever lie to me about the milk

Bhoomika 的头像
Bhoomika10 个月前

Omg! Eggie is the best, cutest and most efficient :)

brian-machado-high-inference 的头像
brian-machado-high-inference10 个月前

really nice design!

Yakuza 的头像
Yakuza10 个月前

"Robots will take your job" will start from house cleaners.

PRAMOD BN 的头像
PRAMOD BN10 个月前

Wanted to order 2 Eggie How can I please an order ?

YQ 的头像
YQ10 个月前

With force control ?

Tangible 的头像
Tangible10 个月前

Yes!

Harrison Kinsley 的头像
Harrison Kinsley10 个月前

Definitely a robo company crime to speed up a video but not clearly mark it.

Abhinav Das 的头像
Abhinav Das10 个月前

Love it!

Tangible 的头像
Tangible10 个月前

🫶

Sofya Sellak 的头像
Sofya Sellak10 个月前

Woww impressive work!

Arun M 的头像
Arun M10 个月前

This is amazing. Welcome Eggie!

Jonathan 的头像
Jonathan10 个月前

@adcock_brett is this CGI? 😄

Ryota Sekine 的头像
Ryota Sekine10 个月前

I like how it noticed the tiny leftover spill at the end

Jay@Proception 的头像
Jay@Proception10 个月前

Congrats!

Tangible 的头像
Tangible10 个月前

thanks mate

allison lim 的头像
allison lim10 个月前

May I meet you?

chriscardoza 的头像
chriscardoza10 个月前

holy shit!

相关视频

We trained a humanoid with 22-DoF dexterous hands to assemble model cars, operate syringes, sort poker cards, fold/roll shirts, all learned primarily from 20,000+ hours of egocentric human video with no robot in the loop. Humans are the most scalable embodiment on the planet. We discovered a near-perfect log-linear scaling law (R² = 0.998) between human video volume and action prediction loss, and this loss directly predicts real-robot success rate. Humanoid robots will be the end game, because they are the practical form factor with minimal embodiment gap from humans. Call it the Bitter Lesson of robot hardware: the kinematic similarity lets us simply retarget human finger motion onto dexterous robot hand joints. No learned embeddings, no fancy transfer algorithms needed. Relative wrist motion + retargeted 22-DoF finger actions serve as a unified action space that carries through from pre-training to robot execution. Our recipe is called "EgoScale": - Pre-train GR00T N1.5 on 20K hours of human video, mid-train with only 4 hours (!) of robot play data with Sharpa hands. 54% gains over training from scratch across 5 highly dexterous tasks. - Most surprising result: a *single* teleop demo is sufficient to learn a never-before-seen task. Our recipe enables extreme data efficiency. - Although we pre-train in 22-DoF hand joint space, the policy transfers to a Unitree G1 with 7-DoF tri-finger hands. 30%+ gains over training on G1 data alone. The scalable path to robot dexterity was never more robots. It was always us. Deep dives in thread:

Jim Fan

301,632 次观看 • 7 个月前