daniel wu retweeted
GEN-1.5 has been training continuously for over 8 months. We left it running because every metric we tracked kept improving with the engine: absorbing more data, scaling more efficiently, boosting post-training, and compounding step-change improvements through algorithmic advances.
daniel wu retweeted
if you’re a guy in your 20s, build a robotics startup
go into debt if you have to
daniel wu retweeted
Human Archive builds datasets to model sensorimotor intelligence for robotics and world models.
HA-Multi is the largest dataset of manual labor tasks aligning egocentric RGB, stereo depth (active IR), tactile gloves, body IMUs, and wrist cameras.
Congrats on the launch, @babugi28, @shloke_patel, @samay_maini15, and @Rush26agarwal!
ycombinator.com/launches/PeP…
daniel wu retweeted
Today, any business owner or worker can get paid to do the same work they already do every day. All you have to do is wear a simple hat. If you or someone you know runs a business with manual labor tasks, fill out this form: docs.google.com/forms/d/e/1F…
daniel wu retweeted
Introducing Clutter Bot
A robot that understands simple english commands and actually executes them.
Natural language → robot action. Made using @GeminiApp and @Vultr
Built with @danielwu28 for the @lablabai AI Meets Robotics hackathon!
daniel wu retweeted
Today we're releasing Isaac 0: our first robot for the home.
And we made a short video for our first customers to share what it’s been like having one in their home 😊
daniel wu retweeted
Very excited to release Contact-Anchored Policies (CAP) 🧢 today! Check out this thread for more details on that and on our in-the-loop simulation evaluations:
heading to sf tmr for fouders inc hw hackathon. was wondering if theres anyone who i can borrow a spare g1 from. thanks!
daniel wu retweeted
how to get started in robotics (software side):
robotics software isn’t just coding; it’s turning math and logic into motion.
start from first principles and build up like this:
learn python + c++
robotics runs on both; python for prototyping, c++ for real-time control.
understand linear algebra + control theory
you can’t move robots confidently if you don’t understand matrices and feedback loops.
master ROS2
it’s the operating system for robots. topics, nodes, messages, launch files; learn how data flows.
build simple projects
line-follower, wall-avoider, or teleop robot using Arduino + ROS bridge. you’ll learn more wiring things than watching videos.
simulation is your friend
Gazebo / Isaac Sim; simulate sensors, motors, and physics before touching hardware.
learn perception
start with OpenCV, then move to camera calibration, LiDAR, depth sensing.
learn planning & control
PID → trajectory planning → model predictive control.
you’ll realize control is where “magic” happens.
structure your code like systems, not scripts
think in nodes, data pipelines, and processes; not while-loops.
once you reach this stage; start integrating ML, reinforcement learning, or swarm coordination.
robotics is just layers of abstraction sitting on top of physics; master each layer, and you’ll start seeing how it all connects.
