// OVERVIEW
Chinese technology conglomerate whose Qwen team developed the Qwen3-VL vision-language model.
Discussed in
- The Pentagon Stops Using Claude
- ⚠️ AI alarm bells
- The Man Who Taught 8 Million People AI Just Said AI Is Terrible for Learning
- OpenAI Says SpaceX Can't Be Trusted
- 152. 领读Kimi K3技术报告：从架构创新聊起，注意力美学、多教师蒸馏和开源MoE
- The Trump Administration Proposes New H-1B Fees
- BREAKING_ Nikesh Arora, Palo Alto Networks
- 179_ 蒸馏风暴：一场无人公开谈论的技术竞赛
- ASCII Art Turns LLMs into VLA Controllers
// RECENT MENTIONS
// SIGNALS
50 SIGNALS
01
mention·arXiv Physical AI·OCTOBER 13, 2026
“Human collection reduces mean labor per demonstration from 140.85 to 33.30 person-seconds (76.4% less labor, or 4.23× as many demonstrations per person-hour).”
Source→02
mention·arXiv Physical AI·OCTOBER 8, 2026
Source→03
mention·StrictlyVC·OCTOBER 6, 2026
“a 'fleet' of AI agents apparently running on Tencent infrastructure and querying Alibaba's Amap service.”
AI-extracted from podcast / newsletter / paper summaries. May contain errors.