Core Points - Apple announced updates to its developer policies to comply with the EU's Digital Markets Act (DMA) ahead of the June 26 deadline to avoid fines [1][2] - The updates include new "anti-steering" rules allowing EU app developers to link to alternative payment methods outside the App Store [3] - Apple introduced a new fee structure, replacing the Core Technology Fee (CTF) with an initial acquisition fee of 2% and a store services fee of either 13% or 5% based on developer tiers [4][5] Developer Fee Structure - The new fee structure includes a Core Technology Commission (CTC) that will replace the CTF, with a 5% commission for developers on standard terms in the EU [6][7] - Developers previously paying the CTF of €0.50 per app install after 1 million downloads will transition to the new rules by January 1, 2026 [7] - Tier 1 developers have limited access to App Store services, while Tier 2 developers gain access to more comprehensive tools and features [5] Industry Reaction - Epic Games CEO Tim Sweeney criticized Apple's compliance with the DMA, labeling it as "malicious compliance" and arguing it undermines fair competition in digital markets [8]

Core Insights - Huawei Developer Conference 2025 (HDC 2025) showcased advanced technologies and the future of the Harmony ecosystem, indicating significant ongoing effects for users and ecosystem partners [1] Group 1: AI Empowerment and Browser Upgrades - Huawei Browser introduced multiple innovative features and upgrades, leveraging AI technology for a comprehensive upgrade in information search and service [2] - The new browser information assistant utilizes authoritative content sources and cross-platform integration to become a reliable information partner for users [2] - Users can ask questions to the browser, which employs AI's multi-dimensional semantic understanding to provide logical and organized responses, enhancing information retrieval efficiency [2][4] Group 2: Advantages Over Traditional Browsers - Unlike traditional browsers that merely list search results, Huawei Browser's AI summarization extracts key points from lengthy articles, saving users time and effort [4] - AI recommendations tailor content based on user browsing history and preferences, broadening the scope of information available to users [4] - For sensitive users like investors and tech enthusiasts, the browser offers in-depth analysis and professional insights, aiding in informed decision-making [4] Group 3: Empowering Ecosystem Partners - The content strategy upgrade provides ecosystem partners with broader content distribution channels and more precise user targeting capabilities [5] - Huawei Browser optimizes algorithms to deliver high-quality visual content to target user groups, enhancing exposure and engagement [5] - Video content distribution has been improved, supporting high-definition playback and intelligent recommendations based on user habits and device types [5][7] Group 4: Enhancing Content Creation and Distribution - Video creators and content providers can present their work more attractively, reaching a larger audience within the Harmony ecosystem [7] - The browser offers dedicated spaces for audio content, expanding distribution channels for partners like audiobook platforms and radio stations [7] - AI capabilities enhance content distribution efficiency through deep user behavior analysis, leading to higher conversion rates for partners [7] Group 5: Localized Digital Economy Development - Huawei Browser collaborates with partners to build local services, injecting vitality into the localized digital economy [9] - The focus on user experience and ecosystem development is crucial for product success, with Huawei Browser's innovations providing better services for users and more opportunities for partners [9] - The browser's upgrades position it as a leading tool in the digital age, enhancing its role in the internet landscape [9]

RA团队 发自 凹非寺 量子位 | 公众号 QbitAI 总是"死记硬背""知其然不知其所以然"？ 奖励模型 训练也形成了学生选择标准答案的学习模式，陷入诸如"长回答=好回答""好格式=好答案"等错误规律之中。 北京大学知识计算实验室联合腾讯微信模式识别中心、William&Mary、西湖大学等机构提出的 RewardAnything 突破了这一瓶颈 ——通过让奖励模型直接理解自然语言描述的评判原则，实现了从"死记硬背"到"融会贯通"的范式跃迁。 RewardAnything降低了传统模式针对不同场景需要收集偏好数据训练奖励模型再进行RL的高昂成本，能够直接利用自然语言作为 RLHF的标准。 其作为奖励模型，仅需一句话描述的准则即可刷新传统Benchmark的SOTA，在RABench上展示出了与GPT-4.1等顶尖模型相媲美的 原则跟随能力与泛化能力。 | Model | Domains | | | | | Principle Categories | | | Overall | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- ...

Core Viewpoint - The article discusses the innovative AI applications developed by Google Labs, emphasizing their fun, practical, and diverse nature, moving beyond traditional model updates and parameters [3][5][90]. Group 1: Google Labs Overview - Google Labs has over thirty AI products that are either open or will soon be available, showcasing significant innovation [5]. - The project aims to help users learn in more engaging ways, enhance productivity, and integrate AI into daily life [5][90]. - Google Labs was initially launched in 2002 to foster creativity among employees, allowing them to dedicate 20% of their time to innovative projects [92][94]. Group 2: Featured AI Products - **National Gallery Mixtape**: Generates music based on a given painting, enhancing the understanding of art through sound [10][11][28]. - **Learn About**: An AI-assisted learning tool that structures knowledge acquisition, providing a framework, interactive learning methods, and self-assessment [29][32][54]. - **Little Language Lessons**: A practical language learning tool that focuses on relevant vocabulary, local dialogues, and real-life applications [54][60][66]. - **Stitch**: A design tool that allows users to create UI interfaces using natural language or image references, streamlining the design process [67][72][78]. - **Portraits**: A workplace tool featuring a virtual expert that provides advice on professional challenges, simulating real-life interactions [80][82][86]. Group 3: Google Labs' Evolution - Google Labs was shut down in 2011 as the company shifted focus to core business needs, but innovation efforts continued through other projects [94][96]. - The resurgence of Google Labs is characterized by a focus on small-scale projects, rapid iteration, and future-oriented product development [100][104][106]. - The approach aims to maintain vitality in a rapidly changing AI landscape, emphasizing innovation as a key productivity driver [108][110].

Core Points - Apple is in negotiations with European regulators to avoid escalating fines related to the EU's Digital Markets Act (DMA) [2][3] - The company has already incurred a fine of 500 million euros for violating the DMA [2] - Apple is expected to offer concessions regarding its "steering" provisions that limit user access to offers outside the App Store [2][4] Regulatory Context - The European Commission (EC) has mandated Apple to revise its rules within two months of the fine, with a compliance deadline set for July 26 [3] - If Apple fails to comply, fines could escalate to 5% of its average daily global revenue [3] - The EC is working closely with Apple to ensure effective compliance but has not speculated on the outcome before the deadline [5] Financial Implications - Discussions have included Apple's "Core Technology Fee," which charges developers for annual installs after 1 million downloads [4] - The company faces a growing list of regulatory challenges in Europe, including a 1.8 billion euro fine in 2023 for practices harming music-streaming competitors [6] Industry Landscape - Other American tech giants, such as Google, Amazon, and Microsoft, are also facing significant regulatory scrutiny and fines in the EU [7]

Core Viewpoint - The article discusses how AI is transforming the search experience, moving from traditional keyword-based searches to more complex, agent-driven interactions that can handle intricate queries and tasks [1][4][10]. Group 1: Evolution of AI Search - The evolution of AI search has progressed from early AI summaries to advanced deep research capabilities, expanding both the breadth and depth of search functionalities [2]. - The introduction of agent technology has elevated search capabilities, allowing for the resolution of complex problems, research tasks, and content creation needs [3][5]. - The "Nano AI Super Search Agent" is a new product that combines various AI capabilities to enhance search experiences beyond traditional methods [5][14]. Group 2: User Behavior and Search Needs - Traditional search engines often struggle with user queries exceeding 20 keywords, leading to poor answer quality; over 60% of user needs are now question-based due to the rise of large models [10][11]. - The average length of user queries has increased significantly, with some reaching thousands of words, indicating a shift in user behavior towards more complex inquiries [11]. - The majority of user needs in AI interactions remain focused on search, with a growing trend towards longer, more detailed questions [19]. Group 3: Technical Challenges and Solutions - The development of the "Super Search Agent" faces three main challenges: task decomposition, model scheduling, and ensuring high-quality output from diverse data sources [23][25][26]. - The architecture of the AI system is designed to allow for the stable execution of complex tasks, utilizing a combination of over 80 models to enhance reasoning and execution capabilities [27][28]. - The AI browser developed for the Nano AI system enables deep content retrieval from both domestic and international sources, overcoming limitations of traditional search engines [30][31]. Group 4: Practical Applications and Future Prospects - The AI system can generate complex outputs, such as investment reports or multimedia presentations, by integrating various AI tools and models [22][37]. - The ability to analyze large volumes of documents and generate tailored reports or presentations significantly reduces the time and cost associated with traditional methods [29][37]. - The future of AI search and agent technology suggests a trend towards more personalized and efficient user experiences, potentially transforming individual capabilities into "super individuals" [38].

Core Viewpoint - Tencent has introduced a new feature in WeChat that allows users to add emoji and images in comments on Moments, currently in a small-scale testing phase [3]. Group 1 - Tencent's customer service confirmed that the new feature is available for users who have updated to version 8.0.60 of WeChat [3]. - Users who meet the version requirements can view emoji and images in friends' comments during the testing phase [3]. - Some users have reported experiencing the new feature, while others are still waiting for an invitation to test it [6][8]. Group 2 - WeChat has previously rolled out multiple updates, including features like earthquake warnings, automatic translation, job listings nearby, chat record backup, and the ability to delete friends who have blocked or deleted you [13]. - The chat record backup feature allows users to back up their chat history to external storage devices like USB drives and supports the creation and management of multiple backup files [14]. - The backup feature is currently expanding its testing range, and users can access it through the settings menu if they are part of the testing group [14].

Core Insights - Apple executives are internally discussing the acquisition of Perplexity AI, which would aid in the development of an AI-based search engine [1] - If the acquisition does not materialize, Apple has a backup plan to integrate Perplexity as an option within the Safari search engine [1] - Meta, led by Zuckerberg, has abandoned its pursuit of acquiring Perplexity and has shifted its focus to investing in Scale AI [1]

Group 1: Google Antitrust Case - Google faces a setback in its attempt to overturn a €4.12 billion ($4.7 billion) antitrust fine imposed by the European Commission in 2018 [2][3] - Juliane Kokott, advocate-general of the European Court of Justice, supports the fine, stating that Google used its Android operating system to stifle competition [2][3] - The European Commission's position is strengthened as it seeks to enforce stricter regulations on major tech companies [3] Group 2: Apple Antitrust Appeal - Apple has filed an appeal against an EU ruling requiring its iOS operating system to be more compatible with other companies' products [4] - The company argues that the EU's interoperability requirements could compromise user experience and force it to share sensitive information with competitors [5][6] - The European Commission mandates that Apple must provide more access to iPhone features for device manufacturers and app developers to comply with the Digital Markets Act [6]

Core Insights - Tencent's AI assistant app, Yuanbao, has seen a significant increase in daily active users, growing over 20 times from February to March 2024, following its launch on May 30, 2024 [2][3] - The app has undergone rapid updates, with 30 updates in 35 days, enhancing its features and user experience [3][4] Product Development - Yuanbao is built on Tencent's self-developed Hongyuan model and integrates various applications to meet core user needs in information retrieval, processing, and production [3][5] - The app's capabilities have expanded to include features like personal intelligent agents, deep thinking, and image understanding, enhancing its utility in daily life [3][5] Model Iteration - The underlying Hongyuan model has evolved significantly since its introduction in September 2023, with parameter scale increasing from hundreds of billions to trillions and pre-training data expanding to 7 trillion tokens [5][6] - The model has adopted a multi-expert structure, improving overall performance by over 50% compared to its previous dense version [5] Strategic Investment - Tencent's commitment to AI development is reflected in its record R&D spending of 70.7 billion yuan in 2024, contributing to an 8% year-on-year revenue growth to 660.3 billion yuan [7] - The company has integrated its self-developed model into over 700 internal scenarios, driving growth across multiple business lines [7]