CIX 🦾 Skild demo using vision is very cool. All are executed with a single command at the start, except the obstacle course where someone has to keep nudging the robot to make sure it goes over the obstacles. x.com/TheHumanoidHub…

Jensen NVIDIA just gave our bot a Thor-level upgrade. Pretty sure it is worthy of Mjolnir now. 🔨⚡️

time to train bigger models

Now that's a power up 🦾

This is genuinely a really impressive video of whole-body manipulation for teleoperation. I would not want to drop that Thor, and the control on the little Unitree dog is stable enough they can just drive off with the jetson loosely on top?

Ever wish a robot could just move to any goal in any environment—avoiding all collisions and reacting in real time? 🚀Excited to share our #CoRL2025 paper, Deep Reactive Policy (DRP), a learning-based motion planner that navigates complex scenes with moving obstacles—directly

🚨Excited to introduce a major development in building safer language models: Safety Pretraining! Instead of post-hoc alignment, we take a step back and embed safety directly into pretraining. 🧵(1/n)

We built a robot brain that nothing can stop. Shattered limbs? Jammed motors? If the bot can move, the Brain will move it— even if it’s an entirely new robot body. Meet the omni-bodied Skild Brain:

What makes it so resilient? It's "omni-bodied". The Skild Brain spent 1,000 years walking 100,000 different bodies across simulated worlds. In our world, the Brain treats a broken robot as just another body.

From a big proponent of in-context learning for robotics, I’m excited to see in-context improvement/RL for locomotion. Congrats Deepak Pathak and the whole Skild AI team :)

Skild AI WHY DID YOU DO THAT

Great to see in context learning for robotics!

Life update: After an incredible journey building Parallel Parallel Web Systems with Parag Agrawal and shaping the future of web intelligence, I’ve left Parallel and am onto something new and exciting (with an incredible team)!! There’s so much to do at the data layer for next‑level model

Thanks!! Come chat with us at poster 7 at 5pm! And more results coming soon :)

💡Can we trust synthetic data for statistical inference? We show that synthetic data (e.g. LLM simulations) can significantly improve the performance of inference tasks. The key intuition lies in the interactions between the moments of synthetic data and those of real data

It’s been less than a month since we incorporated, and we’ve already found incredible partners and have so many exciting things in motion. Ritvik Kapila and I have been cooking and will have updates to share soon with y'all! Thanks to everyone who has reached out, and all those

can't catch me now 🏃💨🪚🪚

no robots were harmed in the making of this video :)