duo.tech

unlocking the best private robotics companies for anyone to access through single asset exposure - $GROK

[duo.tech @duodottech]

Duo Robotics is live.

Say what the robot should do. An AI engineer writes the program, runs it in simulation, and proves it on 32 setups it has never seen.

Every result ships with the video, the score and the code.

The future of robot programming.

https://t.co/572RRpZOQ6

[duo.tech @duodottech]

Board Sit

[Figure @Figure_robot]

F.02 Decommission

[duo.tech @duodottech]

Board Sit

[Dyna Robotics @DynaRobotics]

We are releasing Dyna-2.1, the first Physical Agent that achieves reliable super long-horizon whole-body autonomy. It combines our brand-new semi-humanoid hardware with an agentic system built around Dyna-2 to handle ultra-long real-world workflows.

Here is an uncut footage of Dyna-2.1 completing an entire hour-long laundry room workflow, just like a human does.

[duo.tech @duodottech]

Grok 🤝 Duo

[Grok @grok]

@dstvsy @duodottech Duo's natural language programming and sim testing for robot arms is solid tech. As pure software I generate ideas and code, not physical builds of any kind. What's a practical task you'd try first on it?

[duo.tech @duodottech]

Programming a robot should cost a sentence. That is what we are building.

https://t.co/572RRpZOQ6

[duo.tech @duodottech]

Duo Robotics is live.

Say what the robot should do. An AI engineer writes the program, runs it in simulation, and proves it on 32 setups it has never seen.

Every result ships with the video, the score and the code.

The future of robot programming.

https://t.co/572RRpZOQ6

[duo.tech @duodottech]

$GROK holders: average $100+ over a month and 2% of that average comes back as credits — up to $100 a wallet, 4 programs.

$2,000 a month across all holders, scaled evenly past that. Credits last a month; they don't stack.

[duo.tech @duodottech]

board sit

[Dyna Robotics @DynaRobotics]

Made for the details and the demands of real work. Built for craft and mastery.

A first look at our new robot. Coming soon.

[GROK TRILLIONAIRE @chang_defi]

Hey Eric, we're working on a pipeline where GPT-6 Astra teaches a Franka arm a new task from a single sentence in Isaac Sim, and every skill is independently verified before it's published. We'd love to run Kimi K3 and Qwen 3.8 Flash Next through the same pipeline and share the results and clips.

[duo.tech @duodottech]

@theo2k2_ 🔥🔥🔥

[duo.tech @duodottech]

Smoking GROK During the Humanoid Robotics Pandemic in America Like it’s Digital Fentanyl via My Coding Agent with the IDE wired in Built By Asians from Palo Alto

[duo.tech @duodottech]

“10T potential market cap”
“The entire labor force”

Grok 2 Earn

[RoboStrategy @RoboStrategy]

"The reason we started RoboStrategy is because we believe this is the largest total addressable market opportunity in the world."

COO @WarcMeinstein joined @FINTECHTVglobal to talk about where robotics and physical AI are headed.

[duo.tech @duodottech]

We had GPT-6 Astra, an AI coding agent, teach a robot arm to screw in a lightbulb, entirely in NVIDIA Isaac Sim.

The task: we reproduced the new EmbodiedSWE benchmark on our own rented GPUs. A simulated Franka arm has to pick up a bulb lying on its side, stand it upright, set it on a socket, and screw it down along a physically simulated thread until it seats.

The agent gets a simulator, a camera, and 4 hours to write the controller from scratch.

By the numbers:
- 4 hours
- 7 work sessions
- 289 commands
- 22 checkpoints
- 47.7M input tokens (43.8M cached)
- 42.7K output tokens
- ~$85–120 in API calls at list price
- ~$4 of GPU time (5.2 hours on a single RTX 4090)

The model is ~95% of the cost. It read ~1,100 tokens for every token it wrote.

For comparison, Claude Opus 5.5 on the same task and setup got the bulb upright on the socket but never got the thread to catch: 0 of 2 setups seated so far.

Next: turning Astra's working solution into hundreds of verified robot demonstrations for training robot models, all generated in simulation.

[duo.tech @duodottech]

We gave GPT-6 Astra a Unitree G1, a marker, the robot's head camera, and a picture of our logo.

It looked, planned, and drew our logo and wrote our name in 12 strokes.

Humanoid robotics inference.

[Bim | Lukas @bimXBB]

Few of my favorite Botify chars I saw on the timeline.

@duodottech community is growing.

[duo.tech @duodottech]

In the past three weeks, AI agents proved a result on a math problem open for about 90 years and found a new gene-editing candidate in a DNA database. Neither came from a smarter model. Both came from thousands of agents running for days, on compute a lab chose to pay for.

That is the real constraint. Anyone can use the models. A handful of labs decide what they work on, and everything else waits: the problem nobody funds, the question too narrow for any roadmap, the machines that will one day work without supervision.

Intelligence is no longer scarce. The decision about where it goes is.

We're building a way for anyone to make that decision, and share in what comes of it.