Welcome to the 【AI Daily】 column! Here is your daily guide to exploring the world of artificial intelligence. Every day, we present you with hot topics in the AI field, focusing on developers and helping you gain insight into technological trends and understand innovative AI product applications.
Fresh AI products click to learn more: https://top.aibase.com/
1. Claude4 arrives! Anthropic launches the strongest AI model in the industry, surpassing competitors comprehensively in programming capabilities
Anthropic introduced the Claude4 series AI models at its first developer conference, including Claude Opus4 and Claude Sonnet4, leading in multiple benchmark tests, especially excelling in programming tasks.
[AiBase Summary:]
🌟 The Claude4 series AI models lead in multiple benchmark tests, particularly excelling in programming tasks.
💰 Opus4 and Sonnet4 are respectively aimed at paying and free users, with flexible and competitive API pricing strategies.
🔒 Anthropic has strengthened the security measures for Opus4, ensuring compliance with strict ASL-3 security standards.
2. Apple plans to launch an AI smart glasses expected to be released by the end of 2026
Apple plans to release a smart glasses integrated with cameras, microphones, and advanced AI technology by the end of 2026. It features diverse functions including taking photos, video recording, navigation, and music playback. Siri will be the core experience, with accelerated development progress. Prototype machines are expected to be manufactured in large quantities by the end of the year.
[AiBase Summary:]
🌟 Apple plans to launch AI smart glasses by the end of 2026, integrating cameras, microphones, and AI technology.
🎤 Siri will be the core experience, enhancing user interaction and convenience in daily use.
🔧 Development is accelerating, with a large number of prototype machines expected to be manufactured by the end of the year to ensure product quality.
3. ByteDance and Tsinghua University jointly launched the temporal multimodal large model ChatTS
ChatTS is a new type of temporal multimodal large model jointly released by ByteDance and Tsinghua University, filling the gap in the market for time-series data processing, and possessing strong multi-variable time-series question-answering and reasoning capabilities.
[AiBase Summary:]
🌟 ChatTS natively supports multi-variable time-series question-answering and reasoning, solving the limitations of traditional models in handling time-series data.
📊 The research team adopts the "pure synthetic-driven" approach, building an end-to-end data generation and model training framework to enhance natural language understanding of time-series data.
🔍 The model can identify unseen fluctuation patterns and extract anomalies, demonstrating flexibility and intelligent features.
4. One picture generates everything? AI 'Shenbi Ma Liang' 3DTown, turning a single photo into a 3D city, this operation is too impressive!
The 3DTown framework jointly developed by Princeton University, Columbia University, and Cyberever AI can generate realistic 3D urban scenes based solely on one overhead view without training, solving many problems of traditional 3D modeling.
[AiBase Summary:]
Simplify complexity: 3DTown uses a regional generation strategy to decompose complex scenes into multiple small regions, improving details and alignment accuracy.
Space-aware repair: Through space-aware 3D repair technology, seamless stitching between regions is achieved, ensuring overall coherence.
No-training framework: Utilizing pre-trained 3D generators combined with unique strategies, high-quality 3D scenes are generated, outperforming existing models.
Details link: https://arxiv.org/pdf/2505.15765
5. Farewell wearable devices! OpenAI secretly develops screenless AI device, Altman says it will become the third core device
OpenAI plans to develop a revolutionary screenless compact AI device as the third core device following MacBook Pro and iPhone, led by former Apple Chief Design Officer Jony Ive. Despite some leaks in confidentiality work, its potential could bring the company trillions in market value.
[AiBase Summary:]
🚀 OpenAI plans to introduce a completely new screenless AI device, positioning it as the next-generation core device, surpassing current hardware forms.
🎨 Jony Ive joins, infusing the device with Apple's iconic design genes, enhancing product innovation.
🔒 Leaks in confidentiality suggest that the company needs to strengthen sensitive information management to ensure competitive advantage.
6. SenseTime launches the Yuanluo AI Chess Robot Five-in-One Edition, first cooperation with JD.com Joy Inside
As an AI chess robot designed for children aged 3 to 12, the Yuanluo AI Five-in-One Edition integrates learning and interactive functions for multiple chess types and comes equipped with an English-speaking companion function, providing children with a comprehensive learning experience.
[AiBase Summary:]
🎮 Specifically designed for children, it supports five chess types: Chinese chess, Go, international chess, checkers, and Gomoku.
🤖 Equipped with a conversational agent, it can answer academic questions and provide explanations of encyclopedic knowledge, stimulating children's interest in AI.
💬 Comes with an English-speaking companion function to help children practice speaking at home, enhancing their language skills.
7. Microsoft Notepad adds AI writing function, assisting users in quick creation
Microsoft added an AI writing function to Notepad in the latest Windows 11 update, while Paint and screenshot tools also received several AI upgrades.
[AiBase Summary:]
📝 Notepad adds an AI writing function, enabling rapid generation and expansion of text, boosting user creativity.
🎨 Paint tool introduces a sticker generator and intelligent object selection tool, significantly optimizing image editing experiences.
📸 Screenshot tool includes a "perfect screenshot" button and real-time color picker, making screenshots and edits more precise and efficient.
8. Shenzhen sets up a 7 billion RMB fund to boost the rise of AI hardware startups
The Shenzhen municipal government has established a 7 billion RMB investment fund to support AI hardware startups. Exhibits showcased various innovative results, including Huawei's first foldable tablet PC and UBTech Robotics' intelligent robot 'Meng UU'.
[AiBase Summary:]
🌟 Shenzhen establishes a 7 billion RMB fund to support the development of robotics and intelligent equipment startups.
📱 Huawei showcases its first foldable tablet PC running HarmonyOS, marking an enhancement in technological independence.
🤖 UBTech Robotics introduces the intelligent robot 'Meng UU', and EngineAI displays a quadruped robot, showcasing industry innovation.
9. Google Beam released! Turning 2D videos into 3D immersive experiences with real-time translation and eye contact
At Google I/O, Google launched the revolutionary 3D video communication platform Google Beam. By leveraging AI technology, traditional 2D video calls are upgraded to 3D immersive experiences, featuring real eye contact and real-time voice translation.
[AiBase Summary:]
🌟 Google Beam is based on AI technology, upgrading traditional 2D video calls to 3D immersive experiences, offering a sense of realism without wearing AR/VR devices.
🗣️ The platform integrates real-time voice translation, supporting multilingual communication, opening new possibilities for business collaboration and personal communication.
💼 Initially targeting the high-end enterprise market, already collaborating with major companies like HP and Zoom, planning to gradually expand to the consumer market.
Details link: https://beam.google/
10. UAE launches the "Stargate" plan, fully subsidizing ChatGPT subscriptions for citizens
The UAE announced the construction of the 'Stargate UAE' super AI data center, collaborating with top tech companies to introduce advanced systems covering a 2,000-mile range and providing free ChatGPT services to citizens.
[AiBase Summary:]
The UAE will build the 'Stargate UAE' super AI data center, large-scale and technologically advanced.
The UAE becomes the first country globally to fully subsidize ChatGPT Plus subscriptions for all citizens, promoting technological普及.
The project covers a 2,000-mile range, serving nearly half of the global population, enhancing the country's technological strength and international influence.
11. Mistral releases new open-source model Devstral: Challenging GPT-4.1-mini on notebooks
The French startup Mistral has released the open-source language model Devstral, which has 24 billion parameters and low computational resource requirements, supporting local deployment and device-side usage. It also performs excellently in programming tasks, surpassing many models with larger parameters.
[AiBase Summary:]
✨ Devstral has 24 billion parameters and low computational resource requirements, suitable for local deployment and device-side use.
💻 Performs exceptionally well in programming tasks, particularly adept at handling context-related issues in complex code libraries.
🏆 Scores 46.8% on the SWE-Bench test, far surpassing other open-source models, even leading over GPT-4.1-mini.
12. Anthropic API adds four new features: code execution tool, MCP connector, file API, and expanded prompt cache
Anthropic enhances the functionality and flexibility of the Claude AI by adding a code execution tool, MCP connector, file API, and expanded prompt cache, bringing developers more efficient toolsets.
[AiBase Summary:]
🌟 The code execution tool allows Claude to run Python code, transforming it from a code assistant into a data analyst.
🔗 The MCP connector simplifies the integration of AI with external systems, supporting seamless connections with mainstream platforms.
📚 The file API enhances contextual handling capabilities, facilitating the processing of complex documents and multi-format data.
13. US House passes new bill prohibiting AI regulation for the next 10 years
The US House of Representatives passed the HR1 bill, prohibiting states from regulating AI for the next ten years, drawing attention from tech giants and society at large.
[AiBase Summary:]
🌟 Bill passed: The US House of Representatives passed HR1, prohibiting states from regulating AI for the next ten years.
💡 Benefiting tech giants: Microsoft, OpenAI, etc., benefit from relaxed regulation, driving technological innovation.
🔍 Fund support: The bill allocates $500 million to improve government AI applications and smart governance construction.