Welcome to the "AI Daily" column! This is your guide to exploring the world of artificial intelligence every day. Each day, we present you with the latest content in the AI field, focusing on developers to help you understand technology trends and innovative AI product applications.
Fresh AI products Click for more information:https://app.aibase.com/zh
1. DingTalk's 10th Anniversary: Launches Version 8.0, Introduces New AI Office Form - DingTalk ONE
DingTalk launched version 8.0 on its 10th anniversary, introducing the next generation of AI office application form - DingTalk ONE. This product enables human-AI interaction through natural language dialogue, simplifying complex workflows and reducing the learning curve. At the same time, DingTalk demonstrated its market position and user recognition, including data such as the number of enterprise organizations, paid organizations, and AI application numbers.
【AiBase Highlights:】
✨ DingTalk launches version 8.0 and introduces a new AI office form - DingTalk ONE.
🤖 DingTalk ONE enables human-AI interaction through natural language dialogue, simplifying work processes.
📊 DingTalk demonstrates its market position and user recognition, including data such as the number of enterprise organizations and AI application numbers.
2. Alibaba Open Sources Mobile-Agent 3: A Powerful GUI Agent Family
The X-PLUG team released Mobile-Agent-v3, a cross-platform multi-agent framework with powerful features and optimized performance, suitable for GUI automation tasks.
【AiBase Highlights:】
🧠 Mobile-Agent-v3 is based on GUI-Owl, equipped with strong planning, progress management, reflection, and memory capabilities.
⚙️ It has optimized exception handling and reflection capabilities to ensure efficient operation even under interference.
📚 It provides technical reports, demonstration videos, and code repositories to help developers explore its potential.
Details: https://github.com/X-PLUG/MobileAgent
3. WeChat AI Podcast Makes a Shocking Debut! Two-Person Dialogue News Broadcasts, Will Traditional Anchors Lose Their Jobs?
WeChat is testing an AI podcast feature that broadcasts news in a two-person dialogue format, providing users with an immersive information experience. This feature uses AI technology to convert text into natural and smooth dialogue-style broadcasting, changing traditional ways of consuming news.
【AiBase Highlights:】
✨ WeChat introduces an AI podcast feature, using a two-person dialogue format to broadcast news and enhance user experience.
💡 The AI podcast is based on authoritative media and local government information, ensuring accurate and reliable content.
🚀 It may expand to long article reading, chat record reviews, and other scenarios in the future, redefining the way information is accessed.
4. DingTalk Launches Its First AI Hardware: AI Recording Product DingTalk A1
DingTalk CEO Chen Hang introduced the company's first AI hardware product, DingTalk A1, at the 10th anniversary event. This device, only 3.8 mm thick, can help users clearly record voice content and provide highly accurate transcription services. It also includes 36 work and life agent templates, improving the efficiency of meeting records and work summaries.
【AiBase Highlights:】
🔋 DingTalk A1 is an AI hardware device with a thickness of only 3.8 mm, easy to carry.
🎙️ It supports clear audio recording within 8 meters and offers 10,000 minutes of free transcription service with 97% accuracy.
🤖 It includes 36 work and life agent templates, enhancing work efficiency.
5. Apple's Siri May Be Equipped with Google's Gemini Large Model, Enhancing AI Capabilities
Apple is exploring collaboration with Google, planning to introduce a customized version of the Gemini large model for the next version of Siri, scheduled for release in 2026, to enhance its AI capabilities. This move reflects the competitive pressure Apple faces in the AI field and aims to compensate for Siri's shortcomings through cooperation with Google, offering users a smarter and more efficient interactive experience.
【AiBase Highlights:】
🍎 Apple plans to collaborate with Google, introducing the Gemini large model to Siri.
🤖 Aimed at enhancing Siri's AI capabilities and improving the user interaction experience.
📈 This collaboration may have a profound impact on the AI assistant market landscape.
6. Apple Releases an Optimized Version of SlowFast-LLaVA Model: Long Video Understanding Performance Exceeds Large Models
The optimized version of the SlowFast-LLaVA model released by Apple's research team performs exceptionally well in long video analysis tasks, even outperforming larger models. The model uses a dual-stream architecture, effectively solving issues of information redundancy and context window overflow in traditional frame-by-frame processing, providing an efficient new solution for long video content analysis.
【AiBase Highlights:】
🍎 The core advantage lies in the dual-stream architecture, optimizing video processing efficiency.
📈 It performed excellently in the LongVideoBench benchmark test, achieving 56.6 points with a 10 billion parameter version.
🔧 The model is trained and open-sourced on public datasets, providing new ideas for the AI community.
7. Meta Invests Heavily to Acquire Midjourney Authorization! Zuckerberg's AI Arms Race Wins Another Battle
Meta reached a technical authorization agreement with Midjourney, further strengthening its competitiveness in the AI image generation field. This collaboration is a significant step in Meta's AI strategy and reflects the intense competition in the current AI industry.
【AiBase Highlights:】
🔥 Meta reached a technical authorization agreement with Midjourney, enhancing its AI image generation capabilities.
💡 Zuckerberg accelerates the acquisition of AI talent and resources through high salaries and investments.
⚖️ Midjourney faces legal challenges but continues independent operations.
8. Google Drive Video Editing Gets a Major Upgrade! Vids AI Tool One-Click Access, Lowering the Barrier to Video Creation
Google added a Vids video editing shortcut button to Google Drive, allowing users to edit videos more conveniently. This feature not only simplifies the process but also lowers the barrier to video creation, enhancing the user experience for Workspace users.
【AiBase Highlights:】
🚀 Google added a Vids video editing shortcut button to Google Drive, improving video processing efficiency.
💡 Vids has AI-driven features such as automatic script writing and intelligent clip reorganization, simplifying video production.
🌐 Google promotes innovation in productivity tools by integrating AI tools into daily work.
9. Quark Released a Health Big Model Technical Report, Publicizing the Technical Details of the AI "Chief Physician"
Quark's health big model passed the written exams of 12 core disciplines in China, becoming the first large model to complete this challenge domestically. The model performs well in complex medical reasoning tasks and improves performance through high-quality medical data and reinforcement learning techniques.
【AiBase Highlights:】
✨ Quark's health big model passed the written exams of chief physicians in 12 core disciplines, demonstrating high professional ability.
🧠 Trained using three types of medical data to improve model accuracy and reasoning ability.
🚀 Introduced a reinforcement learning phase to optimize model output quality and compliance.
Details: https://arxiv.org/pdf/2508.11894 https://github.com/Quark-Medical/QuarkMed/blob/main/report/QuarkMed_Technical_Report.pdf
10. Global Fish Drawing Frenzy! Millions of Netizens Are Obsessed with This AI Game, Simple Sketches Turn Into Live Fish
The article discusses how the AI game Draw A Fish quickly gained popularity worldwide due to its minimalist design and social interaction mechanisms, showcasing the charm of combining AI technology with human creativity.
【AiBase Highlights:】
🐟 AI technology transforms simple sketches into dynamic virtual fish, providing a unique creative experience.
🎮 The game is easy to operate and does not require registration to participate, attracting a large number of users who are immersed in it.
🌐 Social interaction mechanisms have sparked the creative enthusiasm of global players, forming a space for creative display.
