TechCrunch · AI6565NVIDIA CEO Huang Renxun demonstrated the Microduck robot at the GTC Berlin conference. The robot, produced by Pollen Robotics which was acquired by Hugging Face and is now part of NVIDIA, can perform actions through voice commands. Pre-orders have exceeded 20,000 units, and deliveries are expected during the Christmas period.
LangChain Deep Agents releases00LangChain Daytona v0.0.9 released, fixes a bug where some output is retained after a request timeout. By applying a request timeout mechanism, it ensures that partial results can still be obtained when a timeout occurs. This update was supported by community contributors and improves the stability of the component.
IBM Research BlogAll nine winning teams used IBM Quantum computers in their research for the NEDO Quantum Challenge, covering three tracks: Social 5.0, Quality of Life (QoL), and Cool Japan, with applications including disaster prevention, drug discovery, and gut microbiota analysis. IBM provided participants with quantum computing resources, technical guidance, workshops, and educational content support. Of the 65 teams (about 81%), 65 chose the IBM Quantum platform, running 11,159 tasks and using over 5,900 minutes of quantum computing time.
Claude · YouTubeSame story · 3 outletsClaude Duet is a new app where Claude proactively contacts users only when needed, sending a brief summary of their primary tasks every morning. Users can delegate more complex tasks to Claude through voice notes and wait for confirmation before sending, paying, or signing anything. The app is currently in an early research preview, and Claude is learning how to better assist users. US Pro and Max plan users can join the waiting list through claude-duet.
How the accounts differ:The Verge mentions Claude Duet is for professionals and small business owners, while Claude states it's for Pro and Max plan users.
Cohere Blog5858Cohere Model Vault introduced confidential computing on Microsoft Azure using NVIDIA GPUs in its public test version, designed to reduce the plaintext data infrastructure customers must trust and provide cryptographic evidence for customers to verify before sensitive information is released. The system protects the computation itself through hardware-enforced isolation, uses Oblivious HTTP to encrypt sensitive request content, and employs a remote attestation mechanism to ensure the environment processing requests meets expectations, thereby protecting the entire process from request transmission to computation execution.
Cohere Blog5858Cohere Model Vault launches confidential computing feature in public test, bringing hardware-level protection and verifiable security for hosted enterprise AI. The feature runs inference on both CPU and GPU, protecting workload memory through hardware-enforced isolation and using NVIDIA confidential computing technology to encrypt GPU memory and traffic between CPU and GPU. Customers can verify the processing environment, including hardware, firmware, and software configuration, before sending sensitive data, ensuring data is protected during the reasoning process.
SemiAnalysisAmazon Trainium has significantly improved its training experience with native PyTorch, enhancing its competitiveness in the training field. Trainium3 supports the standard PyTorch API, allowing most PyTorch code to be migrated to .to('neuron') without XLA annotations, but the BF16 training experience is still not fully refined. SemiAnalysis's engineering team testing shows its performance is close to the CUDA ecosystem, but there is still room for improvement.
Databricks Blog2222Databricks Integrating Spark Structured Streaming Real-Time Mode, Lakebase Postgres, and Databricks Apps
iFlytek released the Xingchen TaaS platform, which transforms Token consumption into business value such as simultaneous interpretation translation and digital employees through intermediate capabilities like model scheduling and intelligent agent engines. It provides solutions for industries such as enterprise services, finance, and healthcare, ensuring that Tokens are not wasted. The Spark X2.6 edge model is open-sourced, with an expanded context window to 1 million Tokens and an edge-cloud routing accuracy of 96.1%.
LangChain Deep Agents releases3535deepagents-code 0.2.1 released, default Anthropic Auto classifier set to Claude Sonnet 5.5, new skill range MCP tool support added.
LangChain Deep Agents releases1515LangChain Deep Agents 0.7.31 released, fixed bugs such as omitting asynchronous task IDs from user responses by default, retaining filesystem permissions in SDK middleware overrides, and added a feature to expose execution output metadata retention in tracking.
The Verge · AI6565Google updates its privacy policy, allowing targeted ads based on user's Drive, Gmail, or Google Photos data when using personalized AI features. Starting October 16, if users connect these services via Google's personal intelligence features to get better responses or recommendations, their data may be used for personalized ads, but advertisers cannot access personal identity information, and ads will not be based on sensitive categories such as race, sexual orientation, etc. Users can disable personalized recommendations through settings.
AWS Machine Learning Blog3434Bayer Crop Science built an intelligent decision engine (ADE) on Amazon Bedrock AgentCore to solve user problems through multi-agent and natural language interfaces, with the first phase solving about 80% of user problems without engineer intervention.
AWS Machine Learning BlogAmazon Quick Sight now supports drawing points and routes in GeoJSON files, previously only supporting custom shape layers. Points are used to mark specific locations (such as stores or facilities), and routes are used to display paths between locations (such as delivery or flight routes). Quick Sight is Amazon's BI feature, offering interactive dashboards, natural language queries, and machine learning insights.
Ollama releases6565NVIDIA Nemotron 3.5 Bolt E4B is a visual language model based on Gemma 4, supporting a 128K token context window, including DFlash speculative decoding, used for local chat, image understanding, and tool use.
Amazon ScienceAudible's technology team has developed a new controllable speech synthesis technology, by combining large language models (LLMs) and three innovative technologies to enhance the experience of audiobooks: 1) Develop a data synthesis pipeline to generate speech data with voice feature descriptions, train the model to understand the descriptions and generate matching audio; 2) Create an automatic system to analyze book characters and identify dialogues, with an accuracy of 97%, to provide stage directions for the model; 3) Improve the dialogue discrimination by 3.7 points in MUSHRA tests, and significantly improve the sound experience in English and Spanish tests by 80%.
Databricks Blog2222Databricks provides an open, unified lakehouse architecture, integrating BI, ETL, stream processing, machine learning, and generative AI, with data stored in customer-managed cloud object storage (Delta Lake, Apache Iceberg™), without proprietary format locking; Oracle relies on proprietary formats and tool chains, with complex costs and difficult migration, its AI capabilities are newer and less mature than Databricks.
Ollama releases6565NVIDIA Nemotron 3.5 Bolt E4B is a visual language model based on Gemma 4, supporting a 128K token context window, including DFlash speculative decoding, and available via ollama run nemotron-3.5-bolt:e4b.
Codex releases1010GitHub Copilot rust-v0.165.0-alpha.1 has been released, version number 0.165.0-alpha.1.
Hugging Face BlogHugging Face releases torch-neuronx model supporting query Space, detecting AWS Trainium compatibility. This Space supports querying whether mainstream models (such as Qwen3-0.6B) with a scale of 175B parameters can run on Trainium without modification, and provides running scripts.
The 2027 Geely Boyue model is officially launched, with an official guide price starting at 85,900 yuan, reduced to 78,900 yuan after a limited-time offer. The new car upgrades to the next-generation GEEA3.0 electronic and electrical architecture, improving five dimensions: appearance, comfort, intelligence, quality, and safety, equipped with the Xingrui AI large model and L2-level driving assistance system.
Mastra BlogMastra's MCP server can now send events to connected clients via push, polling, or webhook delivery methods. MCP Events provide a standard way for both ends to .emit() and .subscribe() events without the need for additional pub/sub services. Events can be used by agents and applications to handle out-of-conversation changes, such as deployment failures, pull request merges, new registrations, or order status changes.
The Valhalla VK 99 V2 series three-mode mechanical keyboard will go on sale tonight at 20:00. The standard version is estimated to be priced at 329.8 yuan, and the MAX version at 429.25 yuan. It uses a 2U0 98 layout, featuring an integrated SkyLight bar with through-hole RGB backlighting, supporting AI creative lighting effects and music rhythm. It includes an internal vertical volume knob.
Sony PlayStation PULSE Elevate wireless speaker starts national pre-sale today, priced at 1748 yuan, and will be launched on November 12, 2026. The speaker focuses on immersive desktop audio, equipped with two built-in planar membrane units and a speaker with a bass woofer, supports Tempest 3D audio technology, features an AI noise-canceling microphone, and is equipped with PlayStation Link wireless technology, supporting multi-device connection and dual-device simultaneous connection.
ChinaTalk3535HBM was once popular at the end of 2025, as it could meet AI inference needs through high bandwidth, but the market has matured, single architecture no longer dominates, and different chips perform differently in training, pre-fetching, or decoding.
Amazon has renamed its e-book subscription service Kindle Unlimited to Kindle Plus, adding thousands of new audiobooks in the US region, with the monthly fee remaining at $11.99, and core member benefits unchanged. The new audiobooks include ACX-produced content and AI voice-acted works generated by Virtual Voice technology.
LSTC, Hokkaido University, Fujifilm, GE Healthcare, and imec are collaborating to develop a low-power AI inference chip based on a 2nm process for medical imaging, used for photon counting CT high-speed reconstruction, ultrasound imaging real-time overlay, and real-time identification during surgical microscopes. The goal is to provide 5 times the image processing speed and 10 times the energy efficiency of existing GPUs. It will be migrated to a chip after development on IBM Spyre accelerators and will be used for clinical applications at the Hokkaido University Hospital.
The Stone A30 Pro 2.0 robot vacuum was reduced by 1175 yuan to 1223.93 yuan during the JD.com 11.11 event after additional subsidies, down from the original launch price of 2399 yuan. It offers 25000Pa suction power, 95℃ high-temperature self-cleaning, a 99.9% antibacterial rate in the wastewater tank, and 60 minutes of battery life.
HP launched the ZHAN99 Ultra (ZHAN99 Ultra 16 G3a AI) mobile workstation on e-commerce platforms, equipped with AMD Ryzen AI Max+ PRO 495 processor, 128GB memory and 1TB storage, priced at 59999 yuan, on sale from October 30th.
Developer Black Apple announces that its space survival strategy game 'Terraspace' will launch a limited-time free 'happy hour' event on the Epic Games Store starting today. Players can get it for free before 4 a.m. Beijing Time on October 28. This limited-time offer is initiated by the development team and is not an official Epic event. 'Terraspace' is set on an undeveloped planet where players need to collect resources, build facilities, meet the survival needs of a space team, and gradually build a spaceship to start an interstellar migration journey.
SkyProduction, the AI video creation platform under Kunlun Wudi, has integrated the Kling 4.0 AI video generation model, which is a new generation of native multimodal AI video creation large model launched by Ke Ling. It achieves upgrades in audio-visual performance, creative controllability, narrative coherence, and diverse content generation.
MarkTechPostFeaturedSame story · 3 outlets7171Google Cloud announced Agent Compute Engine at AGNTCon + MCPCon, allocating a hardware-isolated microVM for each AI agent's session, with no credentials stored within the sandbox; a customer preview program is set to open in the coming months.
Why it matters: The text disassembles the design of each intelligent agent's exclusive microVM and zero credential sandbox, and provides a comparison of isolation and idle processing with AWS, Azure, and E2B.
The Honor Magic9 series debuts as the first AI agent phone equipped with a system-level Agent Harness commercial architecture at CEIC, leveraging the YOYO agent to achieve complex cross-application operations with an overall task accuracy rate of 91.8%. Honor showcases a complete AI terminal ecosystem, including robots, tablets, watches, etc., centered around the phone to build an all-scenario intelligent system.
