原标题：防范学生过度依赖生成式人工智能 《使用指南》明确了五大应用原则。一是坚持育人导向，强化素养本位。引导学生在生成式人工智 能技术使用过程中形成适应智能时代发展的价值观、必备品格和关键能力。二是坚持教育公平，尊重学 生差异。确保技术工具平等地惠及不同地区和个体，无障碍地服务有身体或认知障碍的特殊需求群体。 三是坚持价值引领，确保技术向善。营造健康有序、向上向善的人工智能教育生态。四是坚持需求驱 动，持续稳步推进。结合学校实际，有目的、有计划、有组织地选择、匹配与应用生成式人工智能技术 工具。五是坚持底线思维，保障安全可控。构建覆盖数据安全、伦理审查、内容监管和风险防控的全链 条保障机制，筑牢生成式人工智能技术在教育领域应用的安全底线。 为防范学生因过度依赖生成式人工智能而弱化独立思考能力，《使用指南》从制度规范、教学引导 与角色定位等维度建立系统性防范机制。一方面，明确禁止学生直接复制人工智能生成内容作为作业或 考试答案，并限制在创造性任务中滥用人工智能，从源头上杜绝"代劳式"使用行为。另一方面，强化教 师的引导职责，要求教师在教学实践中积极开展批判性思维训练，培养学生对技术输出内容的质疑精神 与甄别能力 ...

南都讯记者吕虹发自北京12日，南都记者从教育部获悉，《中小学人工智能通识教育指南(2025年版)》 和《中小学生成式人工智能使用指南(2025年版)》(以下简称《使用指南》)正式发布。其中，《使用指 南》明确，禁止学生直接复制人工智能生成内容作为作业或考试答案，并禁止小学阶段学生独自使用开 放式内容生成功能。同时该指南强调，教师不得将生成式人工智能作为替代性教学主体。 据教育部相关负责人介绍，此次施行的两项指南是科学规范推进人工智能全学段教育、进一步落实培养 具有人工智能素养的创新人才的重要举措。《使用指南》重点围绕生成式人工智能在中小学教育中的应 用场景，明确各学段使用规范，确保技术安全、合理、以有效地辅助教学、促进学生个性化学习、推动 教育管理智能化。而《中小学人工智能通识教育指南(2025年版)》则旨在构建一套科学完备的人工智能 通识教育体系，通过螺旋式课程设计以及"课程重构、资源融通、评价创新、师资赋能"的联动机制，推 动人工智能教育从局部试点转向全域覆盖。 南都记者了解到，《使用指南》明确了AIGC在中小学教育中的三大类核心应用场景，即促进学生成 长、辅助教师教学和支撑教育管理。 针对《使用指南》提 ...

Macro Insights - China continues to impose export controls on rare earths, which raises concerns about supply shortages in the U.S. as over 90% of rare earths require processing in China [2][25] - The Chinese government has initiated a special action against the smuggling and export of strategic minerals, including heavy rare earths [25] Company Updates - Xiaomi faces a significant trust crisis following a fatal accident involving its vehicles and allegations of misleading marketing practices, with CEO Lei Jun acknowledging the company's most challenging period [13] - Zhang Haochuan resigned as Vice President of China Merchants Securities for personal reasons, effective May 12 [12] - CEG liquidators have requested the Hong Kong High Court to withdraw the transfer of shares, indicating ongoing legal complexities for China Evergrande [7] - Shanghai Shimao Construction reported significant litigation and overdue debt issues, with over 6.4 billion yuan in unpaid bond principal and interest [8] - Jin Hong Holdings announced a net loss of 210 million yuan for 2024, exceeding 10% of its net assets at the end of the previous year [9] - Qujiang Culture Holdings reported a net loss of 5.811 billion yuan for 2024, attributed to declining gross profit and increased interest expenses [10] Financial Sector Updates - Regulatory authorities have mandated improvements in the internal control systems of brokerage firms to enhance operational standards in investment banking [21] - Fitch Ratings maintained Shenzhen International's credit rating at BBB but downgraded its outlook to negative [5] - Moody's downgraded AVIC Capital's credit rating from A3 to Baa2 [6] International Developments - U.S. President Trump signed an executive order requiring pharmaceutical companies to lower drug prices, aiming for reductions between 59% and 90% [16] - Morgan Stanley's strategist warned that the U.S. stock market is not yet secure, citing the need for further economic indicators to support a sustained rebound [16] - Nissan plans to cut over 10,000 jobs globally as part of its restructuring efforts, adding to previously announced layoffs [18]

Core Viewpoint - The People's Bank of China, along with several regulatory bodies, has issued a set of 30 key measures aimed at enhancing financial support for the Guangzhou Nansha area, promoting its role in the Guangdong-Hong Kong-Macao Greater Bay Area's development and modernization efforts [1][2][3] Group 1: Financial Services Enhancement - The measures include improving financial services for innovation and entrepreneurship, supporting the construction of technology innovation industry cooperation bases, and encouraging financial institutions to innovate financial products like bill discounting [1] - There is a focus on expanding financial services for social welfare, including the trial of video-based credit card applications and cross-border cooperation among credit institutions [2] Group 2: Development of Specialized Financial Services - The initiative encourages the Guangzhou Futures Exchange to explore electricity futures and enhance the renewable energy futures product system [2] - It supports the development of generative artificial intelligence models for financial applications, including smart credit, investment, and risk control services [2] Group 3: Financial Market Connectivity - The measures promote the use of free trade accounts and other mechanisms to provide comprehensive financial services to foreign institutions [2] - There is encouragement for cross-border cooperation in futures products between the Guangzhou Futures Exchange and the Hong Kong Stock Exchange [2] Group 4: Cross-Border Financial Innovation - The initiative aims to attract diverse investment institutions such as sovereign wealth funds and venture capital firms to the region [2] - It supports qualified Hong Kong and Macao investors in establishing securities and fund companies in mainland China [2] Group 5: Financial Regulatory Mechanism - The measures emphasize the need for coordinated regulatory efforts between central and local authorities, ensuring compliance with unified financial management rules [3] - There is a focus on creating an innovative financial regulatory mechanism tailored to the needs of Nansha's development [3] Group 6: Supportive Measures - The initiative includes policies to support housing, talent, and education, facilitating the development of an international talent zone in Nansha [3] - It aims to create favorable conditions for domestic and foreign professionals in the financial sector to work in Nansha [3]

Core Viewpoint - The recent release of the "Guidelines for Artificial Intelligence General Education in Primary and Secondary Schools (2025 Edition)" and "Guidelines for Students' Use of Generative Artificial Intelligence (2025 Edition)" by the Ministry of Education aims to enhance AI literacy among students and promote the integration of AI into education [1][4]. Group 1: AI General Education Framework - The guidelines propose a layered and spiral approach to AI education, focusing on cultivating core competencies for adapting to an intelligent society [3][4]. - The educational framework emphasizes the integration of knowledge, skills, thinking, and values, forming a comprehensive AI literacy that includes technological innovation thinking, critical thinking, human-machine collaboration, and social responsibility awareness [3][4]. - Specific educational goals are set for different school stages: elementary education focuses on interest and basic recognition, middle school emphasizes understanding technical principles and practical applications, and high school concentrates on systematic thinking and innovative practices [3][4]. Group 2: Integration with Daily Teaching - The guidelines aim for deep integration of AI education with daily teaching through systematic implementation paths, including curriculum development, innovative teaching models, and practical activity integration [4]. - The significance of these guidelines is highlighted as a response to the global technological development trend and the urgent need for cultivating innovative talents for the future [4]. Group 3: Responsible Use of AI - The "Usage Guidelines" stress the importance of preventing students from over-relying on technology, particularly in the context of generative AI applications [6][7]. - The guidelines outline a systematic prevention mechanism to avoid diminishing students' independent thinking abilities, including prohibiting direct copying of AI-generated content for assignments and exams [7]. - Teachers are encouraged to guide students in critical thinking training, analyzing the logical flaws and biases in AI-generated content to enhance their ability to process information independently [7].

本文由中国科学院计算技术研究所研究团队完成，第一作者为硕士生杨晗，通讯作者为副研究员安竹林，助理研究员杨传广。 论文标题：Multi-party Collaborative Attention Control for Image Customization 论文地址：https://arxiv.org/abs/2505.01428 论文代码：https://github.com/yanghan-yh/MCA-Ctrl 背景 近年来，生成式人工智能（Generative AI）技术的突破性进展，特别是文本到图像 T2I 生成模型的快速发展，已经使 AI 系统能够根据用户输入的文本提示 （prompt）生成高度逼真的图像。从早期的 DALL・E 到 Stable Diffusion、Midjourney 等模型，这一领域的技术迭代呈现出加速发展的态势。 在基础 T2I 模型能力不断提升的背景下，图像定制化（Image Customization）需求日益凸显。所谓图像定制化，是指在对参考图像中的主体（subject）保持身份特 征和本质属性的前提下，根据文本或视觉条件生成该主体的新表现形式。这一技术对于电子商务（ ...

Core Insights - The article discusses the rapid evolution of generative AI, particularly focusing on Alibaba's Quark "AI Super Box" and its recent upgrade to "Deep Search," marking a significant advancement in AI applications [2][6][8]. Company Developments - Quark has launched the first "Deep Search" product in China, positioning itself as the first AI application to cater to over 200 million active users [3][4]. - The cumulative download of Quark has surpassed 370 million by 2024, indicating its substantial user base and strategic importance within Alibaba's AI To C (AI to Consumer) business [5][4]. - The new "AI Super Box" integrates various AI capabilities, transforming Quark into a comprehensive "AI Super Assistant" for users [10][20]. Product Features - The "Deep Search" feature allows for intelligent recognition of user intent, utilizing various models and agent modules to provide quick and accurate information [8][9]. - Quark's "High Search Intelligence" (高搜商) enables the AI to understand what to search for, where to search, and how to search, moving beyond traditional search paradigms [8][10]. - The implementation of a "transparent reasoning chain" technology allows users to trace the sources of information, enhancing trust and reliability, especially in sensitive fields like healthcare and law [18][20]. Industry Implications - The launch of Quark's "Deep Search" signifies the onset of the "AI Super Application" era, with Alibaba positioning itself as a leader in this space [6][21]. - The integration of AI into everyday tasks suggests a shift where users will delegate tasks directly to AI, rather than using traditional search tools [20][21]. - Alibaba's commitment to AI-native applications as a core investment direction indicates a strategic focus on enhancing its AI capabilities and ecosystem [21].

Group 1: Revenue and Business Strategy - The company's overseas business revenue accounted for a significant portion of total revenue in 2024, with plans to deepen cooperation with existing overseas clients and expand new customer resources [2] - In 2025, the company aims to actively expand its business, particularly in overseas markets, leveraging opportunities from the digital transformation of banks and focusing on new customer development and product innovation [3] - The company has a strong technical capability in digital currency and cross-border payment, with plans to capitalize on the Hong Kong government's support for fintech development [3] Group 2: Technology and Innovation - The company has been involved in the mBridge project since 2020, providing technical consulting and support, and plans to continue its involvement in the next phases of the project [4] - The company has launched and upgraded the FINNOSafe platform in 2023, which supports various functions related to virtual assets, enhancing its technical capabilities [4] - The company is committed to supporting multiple Central Bank Digital Currency (CBDC) projects, including mBridge and e-HKD, and aims to leverage its first-mover advantage in these areas [4] Group 3: Cost Management and Efficiency - The management plans to continue fine-tuning project management and strictly control operating costs to achieve cost reduction in 2025 [4] - The company emphasizes the importance of financial expense management and labor cost control as part of its strategy to maintain cost advantages [4]

Core Insights - The financial technology (FinTech) industry is experiencing significant growth, driven by technological innovation and policy support, with an expected compound annual growth rate (CAGR) of approximately 13.3%, reaching over 650 billion yuan by 2028 [1] Group 1: Industry Trends - The focus is on building a technology-driven financial value enhancement system, emphasizing the importance of data elements in diverse and compliant financial applications [1] - The integration of technology and finance is becoming increasingly important, with financial institutions prioritizing the contribution of technology to business value [33] - The domestic financial technology regulatory framework is becoming clearer, leading to increased attention on compliance technology [33] Group 2: Key Players and Innovations - iResearch has identified the top 50 outstanding financial technology service providers in China, categorized into four types: comprehensive technology service providers, FinTech IT service providers, core technology application service providers, and digital innovation service providers [5] - Softstone Technology is leveraging generative AI to drive innovation in financial technology applications, focusing on areas such as intelligent customer service and risk control [11][13] - CarX Technology specializes in credit risk control, offering a comprehensive suite of platforms for automotive finance, including AI-driven risk management solutions [15][16] Group 3: Technological Advancements - The integration of financial cloud and core business middle platform systems is providing smart banking solutions for consumer credit [19] - China Mobile Financial Technology is developing a comprehensive enterprise consumption management platform, addressing various corporate spending scenarios [22] - The development of large-scale financial risk control models is enhancing capabilities in customer identification and credit assessment, achieving over 99% accuracy in production-level services [24] Group 4: Future Directions - The industry is moving towards a dual development phase of technological deepening and value release, with a focus on customized and scenario-based product capabilities [33] - The application of large models in financial scenarios is expected to drive a shift from basic digitization to deep intelligence transformation, addressing challenges in data privacy and algorithm transparency [36]

Core Viewpoint - The article discusses the transformative impact of generative AI on technology, business, and society, emphasizing the shift from an information society to an intelligent society, and the need to explore new opportunities and challenges brought by AI [1]. Group 1: Insights from Experts - The article features insights from Yu Jingyi, a prominent professor in computer science, who highlights the current bottlenecks in large model technology and the potential of generative AI in spatial intelligence [5][6]. - Yu emphasizes that the understanding of spatial intelligence is evolving, moving from simple digital reconstructions to more complex intelligent interpretations of space, aided by advancements in generative AI [12][13]. Group 2: Technological Breakthroughs - The development of generative AI technologies, such as DALL-E 3 and GPT-4o, showcases the potential for significant advancements in image and video generation, indicating that the capabilities of language models in visual generation are far from being fully realized [10][11]. - The introduction of the CAST project, which incorporates actor-network theory and physical rules, aims to enhance the understanding of spatial relationships among objects, marking a significant step in the evolution of spatial intelligence [16][18]. Group 3: Challenges and Opportunities - A major challenge in the field is the lack of sufficient 3D scene data, particularly real-world data, which hampers the development of robust AI models for spatial understanding [18][19]. - The article discusses the potential of cross-modal methods to address data scarcity in 3D environments, leveraging advancements in text-to-image technologies to infer spatial relationships [19][20]. Group 4: Future Applications - The short-term applications of spatial intelligence are expected to be in the fields of art creation, gaming, and film production, where generative AI can significantly enhance efficiency and creativity [42][43]. - In the medium to long term, spatial intelligence is anticipated to become a core component of embodied intelligence, potentially transforming industries such as smart devices and robotics [43][44]. Group 5: Ethical Considerations - The rise of AI companionship raises ethical questions regarding emotional dependency and the implications of human-robot interactions, necessitating ongoing discussions about ethical frameworks in technology development [50][51].