Humanoid labor too cheap to meter.
Pittsburgh
Joined April 2022
- Tweets3K
- Following2.8K
- Followers4.8K
- Likes41.1K
Billy Zelsnack 🤖/acc retweeted
MuJoCo 4.0 is out!
• Fast or thin objects no longer pass through each other
• Soft bodies can now deform permanently and rods can twist
• MuJoCo Studio, the successor to simulate, with physically based rendering via Filament
Try it live: mujoco.org
Billy Zelsnack 🤖/acc retweeted
We liquidating the humanoid robots we never fought (in stock in SF):
1 Booster K1
2 Unitree R1 basic
3 Unitree R1 EDU
1 AGIbot X2 Ultra
5 Unitree G1-E EDU
2 EngineAI PM01
1 AGIbot A3
2 Unitree H1-E EDU
1 Unitree H2 EDU
Match any best price, please DM or respond asap, can ship
Billy Zelsnack 🤖/acc retweeted
I'm building a playlist/course for LLM training and inference (WIP).
youtube.com/playlist?list=PL…
So far, we have common topics like distributed training, fine-tuning, RL, OPD, serving, spec. decoding, and basic optimizers.
What else should I include to make this more complete? 🤔
Billy Zelsnack 🤖/acc retweeted
"It’s amazing how much Jedi discourse is just them pretending not to understand things, thus making discourse impossible."
Billy Zelsnack 🤖/acc retweeted
If you have strong intuitions about the future, have high agency, and are creative... you need to put yourself in a position where you can act on those intuitions. Ideally with enough freedom to navigate and follow those intuitions as they change. 1 year in the singularity is worth 10 normal years. Think about the opportunity cost of spending a year in something you don't deeply believe in.
Billy Zelsnack 🤖/acc retweeted
@higgsfield_ai sent us a cease and desist last night, asking us to shut down our Open Source Higgsfield challenge.
Our read on why they sent it. When the apps people pay for can be rebuilt in the open in two weeks, that’s an uncomfortable thing for a closed platform to watch.
So here’s us reiterating our promise to every builder in the challenge and community, and we’re keeping it:
• The challenge stays open
• Final submissions close tomorrow, Oct 30
• Winners are announced Nov 6, with prizes paid as promised
You showed up for each other. We’re showing up for you.
If we win, we all win!
Comfy out 🫡
Last night, @higgsfield sent @ComfyUI a cease and desist (attached)
They want the Open Source Higgsfield challenge taken down within 48 hours.
Their argument is that our use of the Higgsfield name and that using their CEO's likeness is an infringement violation (man, the irony). Claiming that it makes the event look like Higgsfield is associated with it.
This couldn't be farther from the truth, which is that Higgsfield made a promise to open source its product and never did.
For the past two weeks, builders have been recreating popular closed apps as open workflows anyone can run.
So we asked our community to deliver where Higgsfield failed: apps in the open, built by the community, for the community. We've made it clear that the challenge is organized by ComfyUI and not affiliated with or endorsed by Higgsfield.
When we started Comfy, we made the community a promise. Our product stays open, free, and yours. We’ve kept it every day since, and one letter isn’t going to change that.
The challenge stays up. Comfy and open source will win!
Billy Zelsnack 🤖/acc retweeted
Trained the FPM + locomotion policy inside a world model, SuperMimic-style, on my 4070 Super. Far from perfect, the force assistance curriculum didn't carry over, but sparse terrain success surprised me. Not sure about deploying such a policy, but the direction is very promising.
Billy Zelsnack 🤖/acc retweeted
Using KiCad or Fusion through MCP or computer use is like sucking a pizza through a straw
Your AI coding agent doesn’t write code by clicking around VS Code either. It works directly with files, compilers, and other tools.
These MCP integrations are transitional technology. They give people an early taste of what’s becoming possible, but they’re fundamentally constrained by interfaces designed for humans.
tscircuit (the company) has submitted 50 unique PCB designs to order over 2.5yrs, with a median cost of around $220 per order
Evaluating AI designs is expensive, it often takes hours of debugging to discover how the ai made a mistake (when it does). We encode new rules to make sure that future designs don’t have the mistake
Current DRC checks in KiCad are woefully inadequate for fully generated AI designs. The AI will never catch all the mistakes without a framework, the same way AI can’t generate 100k lines of code without running a typecheck. This is why we are a wholly unique MIT-licensed EDA kernel. We encode and force encoding of metadata that KiCad doesn’t support, and provide platform services (such as datasheet and supplier availability APIs) that are required for AI to output complex designs.
It’s open source,there is no lockin. There is no required GUI.
Billy Zelsnack 🤖/acc retweeted
Every day trade workers risk their lives to keep the lights on.
Not anymore.
mundane gives tradespeople a second body that can take the shock or blast instead of them.
Send the robot into the danger zone; the operator stays in full control from a safe distance.
Our mission is to amplify the skilled labor force, not replace it.
Billy Zelsnack 🤖/acc retweeted
Sourcing XT-30 wire harnesses has been one of the most difficult and slowest parts of building Axol.
We’ve faced long lead times, unacceptable quality, and suppliers struggling to accommodate the 90-degree bends Axol needs. One supplier even stopped making XT-30 harnesses altogether because it is too slow to build.
We brought production in-house. Now we’re opening that capacity to other robotics teams with @acornharness.
Acorn Harness is here! XT30 and XT30(2+2) wire harnesses, assembled in the USA and built to your exact length. Pick the gauge, end angles and wire colors, and see the price as you go. Every one is continuity tested and ships within a week. acornharness.com
Billy Zelsnack 🤖/acc retweeted
I've got to admit... this is the most fun I've had at work in a long time!
I love this concept, so we put the methodology to work, and it's training now, but right off the bat it far exceeds the performance of our PPO training pipeline!
👇
We found a way to scale humanoid control that reduces GPU-hours by more than 1000×.
Across 129,514 motions, SuperMimic surpasses the state of the art (128 GPUs, 7 days) with 1 GPU in 3 hours.
Project: supermimic.github.io
Demo: supermimic.github.io/demo/
Billy Zelsnack 🤖/acc retweeted
Over the last few years, end-to-end robot learning has moved incredibly fast. Even working at the frontier, I’m regularly surprised by how much it can already do. And yet we still haven’t seen it deployed at scale.
At @enact_intel, we think that’s because no one has fully committed to this new paradigm yet. Most approaches fall into one of two camps:
Humanoid companies start with the form factor. The body comes first, built to look like us, and the AI gets adapted to hardware designed before the model existed.
Pure model companies run on off-the-shelf hardware. Fast and convenient, but it limits the modalities and the quality of the data.
We see it differently: the body shapes the data, and the data shapes the model.
So we’re rebuilding the whole stack: model architectures for multi-frequency, multi-scale intelligence, the data pipelines, and all the way down to the actuators and sensors feeding the model. Our robots sense force and touch through their own actuators, so they feel what’s happening in real time. That makes learning more data-efficient while pushing reliability and performance. It works across form factors and is built for real-world deployment from day one.
Huge thanks to everyone who’s helped along the way: early supporters, everyone who took time to give advice, our angels, and the fantastic team I get to tinker on robots with late into the night.
@redalpine @firstminutecap @Notion_Cap @ylecun @AndrewLonsberry @arthurmensch @Thom_Wolf @mapo1 @HaraldNieder @spencer_crawley @ascribblenau7 and many more...
Billy Zelsnack 🤖/acc retweeted
@kevin_zakka is one of my favorite voices in the robotics sim world, and i loved his take on this in mujoco -- so i thought i'd give the @Tesla roadster launch video to @antiochrobotics's real2sim pipeline. results are pretty sweet.
all automated end-to-end, took about 15 minutes to recreate from the livestream video.
car runs a full mocked control policy over ROS, thrusters + suction all modelled, etc. smoke needs some work clearly, we'll get that fixed :)
The @Tesla Roadster stunts, simulated in MuJoCo! Thrusters are site actuators, the ramp suction uses adhesion actuators, and the builtin box QP decides how hard to fire each jet.
Billy Zelsnack 🤖/acc retweeted
A million hours of what?
If we’re scaling human demonstrations for robot learning, what should each hour contain?
We built S1, our skill capture glove, alongside our robot hand to record joint movements that map to robot targets, with touch and video on the same timeline.
Billy Zelsnack 🤖/acc retweeted
Launching a new humanoid learning course at @CMU_Robotics in Spring 2027!
Every team will get its own (small) humanoid robot for hands-on labs.
Will open-source everything (lecture slides, codes). Stay tuned!
Billy Zelsnack 🤖/acc retweeted
Three years ago I downloaded “a little weekend project” on Instructables: a pair of animatronic eyes open sourced by Will Cogley.
I never stopped iterating around it. A head, then a body, and those eyes became Vita, a robot that moves, talks and streams 24/7 from my workshop.
Now my new robot MOSS, a fully open source build, is featured on the Instructables homepage. From downloading my first project there to sharing my own.
The loop is closed.
Link in the first comment.
Billy Zelsnack 🤖/acc retweeted
New update to my Anatomy Atlas: joint motion!
Flex an elbow, oppose a thumb, or turn the head. Watch the joints move and the muscles involved light up in real time.
Open the "Motion" menu to try it.
anatomy-atlas.pridgenmd.chat…
Billy Zelsnack 🤖/acc retweeted
🎮Magnetra3🎮
#ContactGlove3