Tencent Workbuddy apologizes for not promptly announcing the adjustment of the DeepSeek-V4.1 Flash promotional event. It has decided to settle the credit consumption from November 5, 14:00 to November 8, 23:59 using this model at a 50% discount of the original price and fully refund it.
Multiple fund companies procure computing power servers, enhancing computing power as the core foundation for the public offering industry's AI layout, driving deep AI empowerment of the entire asset management process
Tencent Yuanqi officially stopped its services today, stopping the creation of new intelligent agents and the distribution of API Keys. Starting from January 9, 2027, it will no longer accept data export applications, and users need to complete their data backup before this date. The official will delete all user data except for legal logs, and it is recommended to migrate to Tencent Cloud Intelligent Agent Development Platform (ADP) or WeChat Official Account Intelligent Reply.
The Verge · AI6565Jensen Huang apologized to Nvidia employees for his extramarital affair with a government official, admitting his mistakes and committing to take responsibility; Nvidia's board has launched an independent investigation.
Latent Space4242Stanford University's CS146S course has changed assignments to contributing to open-source projects, letting students use AI code review tools and learn collaboration; Monaco company adopts Devin code review tool, while building its own Agent workspace to maintain technical autonomy, and develops Hal system, allowing engineers and salespeople to solve problems through Agents.
llama-modchip releases a set of ggml/llama.cpp core patches and quantization transcoding tools, compressing the 284.33B DeepSeek-V4-Flash to 56.56 GiB (1.71 bpw), which can run on a pure CPU machine with 64 GB memory, with a peak memory of 57.61 GiB under 256K context and no need for swap.
Why it matters: The project provides a quantization scheme and core patches for running 284B MoE on 64 GB memory, along with measured data for each hardware, which can be used to determine the feasible boundary for pure CPU deployment.
Goldman Sachs has restricted the use of Anthropic's Claude Code and Microsoft's GitHub for its engineers, prohibiting traders and bankers from using these tools to track customer traffic and build market models.
NVIDIA Newsroom77Jensen apologizes to NVIDIA colleagues, the board, and shareholders, admitting to infidelity with his wife and taking full responsibility, promising to continue leading the company and cooperate with the national security investigation, emphasizing that personal choices are unrelated to family.
NVIDIA NewsroomFeatured7878NVIDIA's board announced that, in light of recent reports that founder and CEO Huang Renxun has become a target of foreign intelligence activities, the board has launched an independent investigation. An independent committee composed of independent directors will oversee the investigation with the assistance of external independent lawyers. The board stated there is no reason to believe that national security or NVIDIA's confidential information has been compromised, and the investigation will review all available information; Huang Renxun has expressed full support and pledged cooperation.
Why it matters: NVIDIA's board has launched an independent investigation regarding reports that Huang Renxun has been targeted by foreign intelligence activities, allowing readers to understand the company's response at the corporate governance level.
The Verge · AI6565Suno has been sued by the rights holders of Elvis and Eminem, accusing it of illegally generating music and lyrics that mimic the style of these artists. The plaintiff's lawyers claim that Suno has violated Tennessee's ELVIS Act, which prohibits AI companies from cloning singers' voices.
Mike Rockwell may step down as Apple Siri Engineering Vice President next year, even leaving Apple. The new Siri project he led was delayed by two years before launch, and his departure will cause significant losses for Apple. Amal Subramanian and Sebastian Marino may take over related work, but the succession process for the software engineering director has not yet started.
Apple CEO Tim Cook urges software director Federigi to accelerate AI business development, aiming to upgrade Siri to an AI agent capable of autonomously completing tasks for users. Tim Cook is optimistic about Apple's AI business performance but demands Federigi adopt a bolder strategy, speed up development, and not be satisfied with just completing the Siri feature upgrade. Federigi is seen as an AI skeptic, but faced with the AI wave led by companies like OpenAI and Meta, Apple can no longer stand aside.
MiniMax · Hugging Face new modelsMiniMax-H3-Image-PE uses prompt expansion (PE) to convert short requests into detailed image descriptions and recommends aspect ratios. It supports two workflows: text-to-image (T2I) and image editing, using fine-tuned FP4 weights from Qwen3.8-27B. The online API can be accessed via platform.minimax.io or platform.minimax.cn, and the documentation provides support in both Chinese and English.
Qualcomm showcased its edge-side MoE intelligent agents at the 9th CIIE, driving edge-side AI from single-point applications to intelligent agent experiences. MoE technology enhances the intelligence level of edge-side models through mixed expert modules, achieving higher complexity task completion in the 2 billion to 3 billion parameter scale. Qualcomm collaborated with Infris AI, Changelong, and Jieyu Xingchen to optimize models, inference, storage, and chip platforms, demonstrating a systematic solution for edge-side AI ecosystem synergy.
The Decoder4646The author points out in a Nature comment that research released by Anthropic within the year shows Claude's internal processing patterns are similar to consciousness-accessible information processing, but the company does not view it as evidence of subjective experience; Butlin and other experts report that based on 14 indicators, current AI systems (including GPT-4) are not strong candidates for consciousness.
Shanghai Jiao Tong University's Chen Haibo, Wang Zhao Guo team collaborates with Fan Chun Hai, Wang Fei team to construct AI-native scientific laboratory AutoDNA device (Siyuan No. 1), achieving autonomous molecular biology experiment operations, results published in Nature and featured on the homepage.
Why it matters: This study combines multi-agent architecture with instrument atomic services, allowing AI to autonomously complete molecular biology experiments, serving as a reference for AI-native laboratory architectures.
The Decoder5252Turing Award winner Richard Sutton states that raw computing power is insufficient to achieve intelligence, AI science needs new methods and is far from completion. He criticizes the limitations of simply scaling large language models (LLMs), calling synthetic data a 'big mistake' and human expertise the bottleneck for expansion. The industry has sparked widespread discussion, with some researchers agreeing that new learning algorithms are needed, but others point out that the algorithms themselves are crucial for scaling.
Fuji TV plans to use AI in all live-action drama production starting from October 2027, significantly shortening the production cycle and saving 70-80% of the special effects production process. The AI production work will be assigned to the internal lab AI Sandbox, which Fuji TV will establish in September, mainly responsible for CG designers trained by Tokyo production company ANDENT to conduct testing and application experiments of AI technology.
ByteDance is considering expanding the second phase construction of its data center project at the Pesca Industrial Port in northeastern Brazil, with a scale exceeding the $10 billion project currently under construction in the local area, and may also build data centers to serve local customers. This move aims to consolidate its leading position in Latin American computing infrastructure, and the project can enjoy tax incentives by transferring data overseas.
Tencent's Chief AI Scientist Yao Shunyu believes the AI moat exists more at the product level than the model level, using WeChat as an example to illustrate its use of user background and proprietary data to provide irreplaceable services, even if competitors have better models, they cannot replicate this trust and contextual understanding ability.
MarkTechPost6262Google AI open-sources the Elephant-Goldfish Model (EGM) Plugin, enabling AI coding agents to write specifications before writing code. The plugin includes 7 skills, 9 reviewer roles, and 6 brainstorming lenses, enforcing independent specification reviews through 5 zero-context review stages.
Binance announced that its Chief Financial Officer Kai Ze Wu will resign, and the successor will be Justin Yek. Previously, Binance was hacked and stole $6.3 billion worth of cryptocurrency, and about 90% of the loss has been made up.
MarkTechPost3838DeepSeek-V4-Pro, Qwen3.6-35B-A3B, and ten other open-source MoE models differ in five aspects: expert count (8B-384B), top-k (1-8), shared experts (7/10 models retained), scoring function (Sigmoid/sqrtsoftplus), and load balancing (Aux loss/bias).
The DecoderFeaturedSame story · 2 outlets7171Researchers from Alibaba and several Asian universities used real user change requests to build the coding intelligence agent benchmark Co-Design Bench. Among the ten tested models, the best-performing GPT-6 Astra only meets about 70% of the task requirements, but only 26.4% of the tasks are completely correct.
Why it matters: A benchmark display composed of real user requests shows that the strongest model can only fully solve about one-quarter of the tasks, and the score difference for the same model in different environments is nearly 20 percentage points.
ASUS ZenBook 16 Air Snapdragon AI ultra-thin laptop 24GB+1TB version priced at 12499 yuan launched, equipped with Snapdragon X2 Elite Extreme processor (3nm process)
Protests against data center construction erupt in multiple European countries, with concerns over their impact on society and the environment. Around 2000 people gathered in Dublin, and 1300 in Helsinki, as the Stop Dirty Data Centres alliance demands a halt to construction until democratic procedures decide. People worry that data centers will consume land and power resources, driving up energy prices, and are concerned that the interests of large tech companies are undermining democracy.
Apple Mac mini M6 with M6 chip starts at 4799 yuan on JD.com after state subsidies, down by a thousand from the initial launch price, crossing the 5000 yuan threshold; education discount version 4777 yuan; M6 chip version's AI performance boosts up to 4 times, storage and graphics processing speed each 2 times, CPU boosts 40%
ApeX Gaming's WS2600C full-tower server workstation case is launched, priced at 1199 yuan, supporting up to 15 HDD expansions, compatible with 420mm radiator.
The Douyin App has integrated Apple's Smart function, allowing users to quickly retrieve information by taking photos of products with their iPhone camera. After taking a photo, users can search for related products on Douyin to view details, historical prices, AI interpretation, and user reviews.
The MaxxBench team made minor changes to 89 DeepSWE tasks (median reference patch added only 17 lines, 3.5%), re-evaluating 15 model configurations, finding the difference between the highest and lowest scores expanded from 20.9 percentage points in DeepSWE to 48.3 percentage points.
Why it matters: By rewriting the instructions and starting code for 89 DeepSWE tasks, it demonstrates the huge gap in model scores between familiar benchmarks and tasks after modification.
A long-time Claude Code user decided to migrate to Pi due to persistent minor frustrations. Claude Code lacked stability and quality control during rapid feature iterations, such as issues with the AGENTS.md file, VSCode integration limitations, and frequent changes in system prompts, ultimately leading the user to switch to Pi. Pi offers stability, minimalist design, and high customization, with the user already implementing multi-tab text selection functionality in 20 minutes by writing an extension.
Driven by the AI chip super cycle, Samsung Electronics' cash reserves have reached 214.8 trillion KRW, an increase of 24.8 trillion KRW from three months ago, nearly double the 108.5 trillion KRW projected for the end of Q3 2025, with market estimates suggesting its reserves are sufficient to support the shareholder return plan announced in August.
China Mobile Cloud signed an agreement with Zhipu AI, focusing on domestic computing power adaptation and model optimization, to jointly build a Token joint development center. Zhipu's GLM series models will be integrated into China Mobile Cloud's MoMA model service platform. Both parties will jointly build an intelligent product matrix and delve into the financial and government sectors.
