Welcome to the "AI Daily" section! This is your guide to exploring the world of artificial intelligence every day. Every day, we present you with the latest content in the AI field, focusing on developers to help you understand technology trends and learn about innovative AI product applications.
Fresh AI products Click to learn more:https://app.aibase.com/en
1. Microsoft announces native support for the MCP protocol in Windows 11
Microsoft announced an upgrade to Windows 11, which now supports the MCP protocol natively, promoting the deep integration of AI agents and enhancing user experience.
AiBase Summary:
🧠 The MCP protocol allows safe connection between AI models and local applications, improving automation efficiency.
🤖 New feature "Experiential Agent" can adjust proactively and provide personalized services.
🌐 Microsoft's open platform attracts developers, promoting diverse AI application development.
2. Douyin's second headquarters is located in Shenzhen Nanshan, and will also build ByteDance AI Lab research center
The article details the convening of the 2026 Shenzhen construction site progress meeting and the concentrated start of the first 222 key projects. Among them, the second headquarters project of Douyin was launched in Yuezhi Street, Nanshan, marking the significant achievements of Shenzhen in attracting top enterprises and promoting industrial development.
AiBase Summary:
🚀 Large scale: Shenzhen has concentrated the start of 222 key projects in 2026, with a total investment exceeding 157.3 billion yuan, and Nanshan district leads the city in investment.
🏢 Headquarters expansion: Douyin's second headquarters is located in Yuezhi Street, Nanshan, focusing on video R&D and AI innovation, forming a dual-headquarters structure with the already activated Houpai Center.
💡 Industrial agglomeration: By leveraging the headquarters economy effect, Shenzhen is continuously expanding new productive forces by attracting global leading enterprises' key links through policy empowerment.
3. Lenovo unveils AI glasses concept machine at CES 2026! 45g lightweight design, first-time computer connection, targeting productivity scenarios
Lenovo showcased its new concept-level AI glasses at CES 2026, entering the AI wearable market with a lightweight design and the first-time computer connection capability, emphasizing cross-device collaboration and productivity enhancement, targeting programmers and designers.
AiBase Summary:
🧠 Lenovo launched a 45g ultra-lightweight AI glasses with dual Micro-OLED displays to enhance information density.
💻 First-time AI glasses direct connection to computers to strengthen productivity scenario applications.
🌐 AI functions include real-time translation, image recognition, and multi-source information aggregation, combining edge and cloud models.
4. Liquid AI releases LFM2.5: A family of small AI models for edge devices
Liquid AI released LFM2.5, a new generation of small base model family, focusing on edge devices and local deployment. The model includes various variants, such as text, vision-language, and audio-language, and performs well in multiple benchmark tests.
AiBase Summary:
🧠 LFM2.5-1.2B-Instruct is a general-purpose text model that performs well in multiple benchmark tests.
🇯🇵 LFM2.5-1.2B-JP is optimized for Japanese, with enhanced performance for Japanese tasks.
🖼️ LFM2.5-VL-1.6B is a vision-language model, supporting practical applications such as document understanding and user interface reading.
Details: https://www.liquid.ai/blog/introducing-lfm2-5-the-next-generation-of-on-device-ai
5. Anthropic releases Claude Code desktop preview version, supporting multi-session parallelism and Git isolation
Anthropic released the desktop preview version of Claude Code, introducing a multi-session parallel workflow that allows multiple independent Claude sessions to be opened simultaneously within the same code repository, each with an isolated branch workspace, solving the risk of code conflicts. At the same time, the desktop version demonstrates high integration, including cloud collaboration, automatic environment synchronization, and independent version management. This release marks the evolution of AI programming assistants from simple "chatbots" into full-featured "automated programming control panels."
AiBase Summary:
🐙 Multi-session parallel workflow: Supports opening multiple independent Claude sessions simultaneously within the same code repository, each with an isolated branch workspace.
💻 High integration: Perfectly integrates Web version features, supporting cloud collaboration, automatic environment synchronization, and independent version management.
🚀 Evolution of AI programming assistants: Marks the transformation of AI programming assistants from simple "chatbots" into full-featured "automated programming control panels."
Details: https://code.claude.com/docs/en/desktop
6. Self-developed world model works! AutoNavi launches "Flying Street View," allowing you to tour thousands of stores without leaving home
AutoNavi launched "Flying Street View" using its self-developed world model, achieving a leap from flat digital navigation to high-fidelity, continuous dynamic real-world restoration, building a more intuitive and intelligent digital lifestyle connection platform for users and merchants.
AiBase Summary:
🚀 AutoNavi launched the "Flying Street View" function, achieving continuous dynamic flight view.
🛍️ Users can preview the route to the store on their mobile screens, reducing decision-making costs.
📈 Merchants can use this technology to showcase their store features, becoming a new window for digital marketing.
7. HP launches the world's first keyboard-style AI PC, revolutionizing traditional office experiences!
HP launched the world's first keyboard-style AI PC, Eliteboard G1a, redefining traditional office equipment with innovative design and powerful performance, offering professionals a high-performance, lightweight option.
AiBase Summary:
🧠 The world's first keyboard-style AI PC, integrating the x86 computer system into a 12mm thick keyboard body.
💻 Equipped with AMD Ryzen AI 300 series processors, supporting up to 64GB memory and 2TB storage space.
💼 Weighs only 750 grams, with built-in battery life of about 3.5 hours, suitable for modern office scenarios.
8. Google Gemini introduces "Guided Learning" new feature: AI personal tutor-style step-by-step guidance, revolutionizing personalized education
Google Gemini's "Guided Learning" feature provides users with a personalized deep learning experience through structured learning paths and interactive dialogue formats, while incorporating educational science principles to improve long-term memory and practical application abilities.
AiBase Summary:
🧠 Breaks down complex topics step by step to achieve personalized learning experiences.
🔄 Supports custom quizzes and flashcard generation, dynamically adjusting difficulty levels.
🎥 Integrates multimedia content to enhance understanding and memory effects.
9. Live streaming e-commerce bids farewell to "rash growth"! Two departments issue regulatory rules, AI hosts are responsible for operators
The live streaming e-commerce industry welcomes a major regulation, with the State Administration for Market Regulation and the Cyberspace Administration of China jointly issuing the "Regulatory Measures for Live Streaming E-commerce," aiming to clarify the responsibility boundaries of all parties and build a more fair and transparent shopping environment. The new regulations clarify the legal red lines for all types of entities, including platforms, live streaming room operators, hosts, and MCN agencies, and propose specific requirements for AI hosts and traffic supervision.
AiBase Summary:
✅ Clarify the responsibility boundaries of all parties in live streaming e-commerce to build a fair and transparent shopping environment.
✅ Regulate AI host behavior, requiring continuous reminders to consumers and clear responsibility attribution.
✅ Introduce traffic supervision methods, taking measures such as traffic restriction and warnings against violators.
10. Open-source version of "Veo 3" is here: LTX-2 officially released, generates 20-second 4K synchronized AI video in one go, runs smoothly on local graphics cards
The open-source release of the LTX-2 model marks an important breakthrough in the AI video generation field, with its audio-video synchronization capabilities and performance optimization providing creators with powerful tools.
AiBase Summary:
🎥 LTX-2 model supports generating the longest 20-second 4K high-definition video in one go, achieving audio-video synchronization.
🔧 Open-source weights and code have been released, allowing developers to freely inspect, fine-tune, and deploy locally.
🚀 Significant performance optimization, reducing computing costs by up to 50%, suitable for local operation.
