VQV Signal

OPEN SOURCE WATCH TECHNICAL

Post on X claims Apple researchers steer a Unitree G1 robot with a local Qwen model

According to a post on X by @QwenAIFans, an Apple research team attached a Mac mini to a Unitree G1 humanoid robot. The robot is said to crawl under tables, squeeze through doors and find a chair, with its default behavior directed by Qwen3.6-35B-A3B running locally with no network. The post links no Apple source, so the claim rests on the fan account alone. What is new is the claim that an open-weight model controls robot movement entirely on-device.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-11-08T05:35:00+00:00 Detected 2026-11-08T18:22:04+00:00
View original source

According to a post on X by @QwenAIFans, an Apple research team attached a Mac mini to a Unitree G1 humanoid robot. The robot is said to crawl under tables, squeeze through doors and find a chair, with its default behavior directed by Qwen3.6-35B-A3B running locally with no network. The post links no Apple source, so the claim rests on the fan account alone. What is new is the claim that an open-weight model controls robot movement entirely on-device.

AI-assisted summary based on the listed source.

Fan account QwenAIFans says an Apple research team strapped a Mac mini to a Unitree G1 humanoid that crawls under tables, squeezes through doors and finds a chair, directed by default by Qwen3.6-35B-A3B running locally with no network. The claim comes from the fan post; no Apple source was linked. Why it matters:...

If the setup is accurate, it would show that open-weight models can handle robot control without a network connection, which matters for developers building embodied AI. Robotics teams and open-model users would have a concrete example to examine. Watch for an Apple publication, a linked demo, or a statement from the Qwen or Unitree teams that confirms the details.

Signal Strength 88% Technical label WATCH Public Interest 48 Category OPEN SOURCE Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 73 Practical Impact Score 8 Novelty Interest Score 70 Consequence Score 18 Curiosity Score 52 Shareability Score 58

VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Scobleizer News Pulse (X).