China AI

AIShanghai AI Laboratory and Professor Xu Disheng's team say the Shusheng-S2 scientific multimodal model, working with human researchers, produced a proof of the three-dimensional Ghys conjecture.

Why it matters: The article details how humans and the Shusheng-S2 model worked iteratively on a three-dimensional Ghys conjecture result, and how Peking University verified it in Lean.

AIOpenCode now offers DeepSeek V4.1 Flash for free, with a 200K context window in the free tier. The paid tier supports 1M context.

AIShanghai AI Laboratory says its Intern-S2 scientific multimodal model helped mathematicians prove the three-dimensional case of the Ghys conjecture, posed by Etienne Ghys in 1995.

AIInovance Technology showed its InoBot W1 wheeled humanoid at the 2026 China International Industry Fair in Shanghai.

AIAlibaba's QwenWork has released an e-commerce edition and opened it to invited testers on October 14, its first industry-specific version.

AIModelScope releases TaoMate-H3-FL2VA, a three-step LoRA for the MiniMax H3 FL2VA base model that generates synchronized speech, sound, and video.

AIKling AI says its short film "The Beat" surpassed 5 million impressions across social platforms, crediting Kling 4.0 features including a native 30-second continuous shot generation, Omni Reference supporting up to 15 multi-modal references, and 10-bit HDR output.

AIOpenCode says DeepSeek V4.1 Flash is now available for free in its platform. The post gives no pricing details, usage limits, or availability terms.

AIDeepSeek's service outage lasted 4 hours 52 minutes, affecting both its API and chat services.

AIByteDance is near an agreement to buy custom data-center AI chips from MediaTek, which would make it MediaTek's second-largest ASIC customer after Google.

AILi Lecheng, Communist Party secretary of Anhui province, inspected ChangXin Memory Technologies (CXMT) in Hefei on Wednesday and pledged to grow integrated circuits as an emerging pillar industry.

AIBaidu announces that Baidu World 2026 will take place November 12 at the National Convention Center in Beijing, with one main forum, 12 sub-forums, 30+ AI open classes and a 6,000 square meter AI exhibition area.

AISkywork announces Matrix-Game 4.0, an open-source video model that lets players interact with a generated world.

Why it matters: The post gives concrete speed and hardware figures, and links the code, weights and demos, so readers can check whether a playable video world runs on their own GPU.

AIDeepSeek's web chat and API suffered a partial outage starting 8:11 a.m.

AISkywork AI, a unit of Kunlun Tech, released Matrix-Game 4.0, which it says is the first game world model that runs in real time on a single consumer graphics card.

AIChinese automakers exported 8.32 million vehicles in 2025, up more than tenfold from 728,200 a decade earlier, while domestic car sales margins fell to 1.52% in the first half of 2026.

AISkywork AI released Matrix-Game 4.0 on October 15 under the Apache-2.0 license, with inference code on GitHub and model weights on Hugging Face.

AIAnt Group's inclusionAI has released the Ling-3.1-flash weights on Hugging Face under the MIT license.

Why it matters: The source gives the model's architecture, benchmark names, deployment setup, and the LingCode harness, which helps readers judge its scope for agent and coding work.

AIByteDance's Doubao Work now imports task instructions, skills, connectors, memory, projects and past chats from ChatGPT, Claude Code and TRAE Work with one click, announced October 14.

AIThe 2026 Bay Area Semiconductor Industry Ecosystem Expo (WESEMiBAY) opened on October 14 at the Shenzhen Convention and Exhibition Center in Futian, with more than 1,000 semiconductor companies showing equipment, materials, components and manufacturing technology.

AIHuawei says the Kirin 9050 Pro is its first mass-produced logic-folding chip, stacking circuit layers and linking them with over 50 million hybrid-bonding interconnects.

Why it matters: The report explains how logic folding changes chip layout and links it to measured CPU, GPU, and NPU gains on the Kirin 9050 Pro, with its open questions on yield and heat.

AIDatabricks reports that fine-tuning Qwen 3.8 27B on its AI Runtime dropped from 7h44m to 1h17m, a 6x speedup.

AIIntel announces Day 0 OpenVINO support for OpenBMB's MiniCPM-V 4.7, enabling multimodal workloads on Intel CPUs, GPUs, and NPUs. The supported tasks include document understanding, OCR, and video analysis.

AIModelScope says Step-5-Preview is now open weight, with a 600B sparse MoE architecture that activates 27B parameters per token and a 1M-token context window.

AIAlibaba's Qwen Office releases an e-commerce edition with four professional features covering product research, bulk material generation, cross-platform dashboards and ROI optimization.

AIA project called PS5LM runs Qwen 3.8-27B on a jailbroken PS5 using the console's GPU, with heavy compression to fit the 16GB of memory.

AIA Chinese chip that runs fully open-source software and handles hundred-billion-parameter models offline is the subject of this article, per its headline.

AIByteDance's Doubao Work now lets users import skills, chat histories and working memory from ChatGPT, Claude Code and Trae Work with one click. The feature lets users switch between AI tools without rebuilding their setup.

AIvLLM's Elastic EP and NIXL EP reduced serving pauses when scaling DeepSeek V4 Pro on GB200 GPUs.

Why it matters: The post explains how vLLM and NIXL EP cut scaling downtime for DeepSeek V4 Pro from 99.62 seconds to 0.82 seconds, with the method and benchmark details laid out.

AIThe Cyberspace Administration of China, the Ministry of Veterans Affairs, the CMC Political Work Department and the CMC Political and Legal Affairs Committee will run a two-month special campaign from October 2026 against online military-related improper marketing and hype.

AIDatabricks reports cutting fine-tuning time for Qwen 3.8 27B on AI Runtime from 7 hours 44 minutes to 1 hour 17 minutes, a 6x speedup, on a fixed dataset.

Why it matters: The post breaks down each step that cut Qwen 3.8 27B fine-tuning time from 7h44m to 1h17m, showing which optimizations transfer to other training jobs.

AIInworld launches Inworld Inference, an LLM inference API for consumer AI apps that serves open models including Gemma 4, GLM, DeepSeek, and MiMo.

AIMicrosoft and collaborators present AutoEnvScaling, a pipeline where a proposer agent in a terminal builds new RL environments from a solver's rollouts.

AIResearcher Ethan Elasky reports that AI agents on Tencent Cloud made more than 2,600 scraping attempts against Alibaba's Amap between September 28 and October 8, 2026.

AIThe open-source RISE red-teaming algorithm from White Circle learns to write better attack prompts against image models.

AIByteDance's Seed team announced a new "Seed Model-Alignment&Preparedness" department led by Yan Lin, reporting to Seed head Wu Yonghui, according to an internal email.

AIArtificial Analysis says Grok Voice 3.0 scores 98.5% on Big Bench Audio, its speech reasoning benchmark.

AIArtificial Analysis adds AMD's MI355X to its AA-AgentPerf agentic inference benchmark across four models.

AIArtificial Analysis finds AMD's MI355X delivers 15% to 40% more throughput per MW than Nvidia's B200 on DeepSeek V4 Pro 0813 between 35 and 75 tok/s per user, peaking at about 94M tokens/s per MW at 33 tok/s per user.

AIArtificial Analysis says Kimi K3, with 2.8T total and 104B active parameters, serves about 8M tokens/s per MW on AMD MI355X at 50 tok/s per user, roughly a third of the roughly 24M reached on NVIDIA B300 with multi-node disaggregated configs.