本文第一作者为张桂彬，新加坡国立大学25Fall计算机科学博士生；本文在南洋理工大学的王琨博士、上海人工智能实验室的白磊老师、和中国科学技术大学的王 翔教授指导下完成。 LLM 智能体的时代，单个 Agent 的能力已到瓶颈，组建像 "智能体天团" 一样的多智能体系统已经见证了广泛的成功。但 "天团" 不是人越多越好，手动设计既费 力又不讨好，现有的智能体自动化方法又只会 "一招鲜"，拿一套复杂阵容应对所有问题，导致 "杀鸡用牛刀"，成本高昂。 现在，一篇来自新加坡国立大学、上海 AI Lab、同济大学等机构并被 ICML 2025 接收为 Oral Presentation 的论文，为我们带来了全新的解题思路。 他们将神经网络架构搜索（NAS）的超网络（Supernet）思想引入 Agent 领域，首创了一个名为 "智能体超网"（Agentic Supernet）的概念。它不再寻找一个固定的 最佳 "阵容"，而是根据任务难度，动态 "剪" 出一个量身定制的智能体团队。结果有多惊艳？性能超越现有方法最高 11.82%，推理成本却只有它们的 45%！ Agentic Supernet： 论文地址：https:/ ...

机器之心报道 编辑：泽南、杨文 一个真实世界模拟器。 当世界模型高度进化后，里面的「人」都在做些什么？ 有人会进行街头演说，吸引到了不少听众，小孩会和机器狗玩： 有人会当街作案，警察前去抓捕，又有人会在大庭广众之下求婚： 本周五，来自马萨诸塞大学阿默斯特分校（UMass Amherst）、约翰霍普金斯大学、卡耐基梅隆大学的研究者们提出了一个神奇的研究：虚拟社区 （Virtual Community）。 虚拟社区将真实世界的地理空间数据与生成模型相结合，为多种不同类型的智能体创建了一个具有社会根基的交互式、可扩展开放世界场景。 该工作昨晚提交，立即吸引了一些 AI 圈大佬的关注，纽约大学助理教授谢赛宁表示，这对于智能体研究来说意义重大。 虚拟社区提供了一个统一的框架，用于模拟社区中人类和机器人丰富的社交和物理互动。它建立在通用物理引擎之上，并以现实世界的 3D 场景作为基础。 作者为人类智能体实现了一个虚拟角色模拟框架，而其中的机器人模拟则主要继承自 Genesis。 虚拟社区通过在环境中填充配置机器人、人类角色配置文件和社会关系网络的智能体（由 LLM 提供支持）来支持基于 3D 场景的智能体社区生成。 这一 ...

图片来源：Michael Nagle/Bloomberg via Getty Images 亚马逊（Amazon.com Inc.）CEO安迪·贾西表示，随着这家零售和云计算巨头利用AI技术处理更多任 务，他预计公司员工数量在未来几年将会减少。 在周二发给员工的一封邮件中，贾西阐述了他对这项新兴技术将如何改变工作场所的看法。他表示，生 成式AI和AI驱动的软件智能体"必将改变我们的工作方式"。 贾西写道："未来，我们完成某些现有工作所需的人手将会减少，而其他类型工作需要的人手将会增 加。虽然很难确切预知长期最终的净效应如何，但在未来几年，随着我们在全公司范围内广泛应用AI 并从中获得效率提升，我们预计这将导致公司员工总数下降。" 自AI热潮兴起之初，行业内外的担忧之声就不绝于耳，人们担心AI可能会取代人类员工。随着科技公 司推出越来越先进、能够编写代码并代表用户处理在线任务的AI系统，这种担忧更是与日俱增。 Shopify Inc.已告知员工，申请新增岗位需要说明为何AI无法胜任该工作。多邻国（Duolingo Inc.）表示 将"逐步停止"使用承包商来完成AI能够处理的工作。而微软（Microsoft Co ...

多家通信行业头部企业都展示了其应对这一变革的系统性思考。6月19日，中兴通讯副总裁、终端事业 部产品中心主任罗炜接受南都湾财社记者专访时指出，面对激烈的市场竞争，终端业务的破局点在于精 准的差异化定位和对新兴增长点的把握。 他向记者强调，拥抱年轻化、通过"科技+文化"的融合与Z世代建立情感共鸣，是当前中兴终端实现"破 圈"和价值增长的关键一招。 从技术迭代到生态共生：AI浪潮重塑基础设施价值 "自2022年ChatGPT发布以来，AI大模型经历了从技术迭代到生态共生的'光速'发展，" 6月18日，中兴 通讯首席发展官崔丽在演讲中指出，"无论未来AI技术发展的方向和节奏如何变化，高效基础设施、高 效算法以及应用落地依然是主旋律。" 这一判断的背后，是AI发展对底层资源的巨大消耗。Scaling Law（扩展定律）从训练延伸至推理阶 段，意味着对算力和能源的需求持续攀升。如何为千行百业的智能化提供可持续、可负担的基础设施， 成为摆在全行业面前的课题。 中兴通讯对此提出的解决方案是，以系统性创新打造"网络、算力、能源"三位一体的智能基座，发 力"算力筑基"。 在网络层面，推动架构向"AI原生"演进。 其核心是推动网 ...

Core Insights - The AI industry is shifting focus from large model parameter competition to practical technology implementation, marking a strategic pivot towards physical hardware [1][3] - OpenAI's acquisition of the hardware company io for $6.5 billion is seen as a significant move towards integrating AI with physical devices [1] - The development stages of AGI proposed by OpenAI align with this shift, particularly emphasizing the transition from passive thinking to active execution in AI capabilities [1] Industry Transformations - AI development is undergoing three major transformations: 1. The competitive focus is shifting from model performance to practical implementation capabilities [5][7] 2. The value logic is transitioning from tool efficiency to result-oriented closed loops [5][7] 3. The product form is evolving from cloud computing to user-centric hardware [5][7] User-Centric AI - Companies are now prioritizing how AI can be integrated into daily life, providing practical value rather than just showcasing advanced algorithms [7][10] - AI is expected to deliver seamless, one-stop services in various scenarios, enhancing user experience and efficiency [7][10] - The concept of AI as a personal assistant is gaining traction, with devices like smartwatches and smart speakers becoming integral to daily routines [7][10] Honor's Strategic Framework - Honor's CEO, Li Jian, presented a systematic framework for AI technology implementation at the MWC Shanghai, emphasizing the importance of AI's practical application [3][9] - The upcoming Honor Magic V5 smartphone is positioned as a key product in this strategy, designed to address industry challenges through integrated AI capabilities [3][9] - Honor's "Alpha Strategy" aims to transform the company into a global AI terminal ecosystem provider, aligning with the multi-terminal AI demand [8][9] Closing the Three Loops - To effectively integrate AI into users' lives, three critical loops must be closed: scene, trust, and performance [10][12] - The scene loop focuses on providing seamless service across multiple devices, ensuring that AI understands user intent and connects tasks fluidly [12][13] - The trust loop aims to build user confidence in AI systems by addressing challenges such as data privacy and ethical concerns [13][14] - The performance loop emphasizes the need for efficient resource allocation between devices and cloud services to ensure smooth user experiences [14][15] Collaborative Ecosystem - Honor advocates for the establishment of an open AI terminal ecosystem alliance to facilitate collaboration among various stakeholders, including AI model companies and telecom operators [15][17] - This initiative aims to drive innovation and accelerate AI's practical application across devices, enhancing user experience [17][18] - Honor's proactive approach in the AI sector, from strategic proposals to product showcases, positions the company as a leader in the ongoing industry transformation [17][18]

机器之心报道 最近，很多 AI 大佬一反常态，对未来做出了超出预期的乐观预测。 诺贝尔奖获得者、AI 先驱 Geoffrey Hinton 表示，人工智能将在多个领域「取代所有人」，只有顶尖技能人才能够找到 AI 无法处理的工作。 特斯拉前负责人 Andrej Karpathy 在演讲中也认为，我们正在进入「软件 3.0」时代，自然语言在成为新的编程接口，大模型会完成剩下的工作。 作者：泽南、陈陈 Code is cheap, show me the talk. 正在让 AI 能力大幅提升的技术被称为智能体（Agentic AI），它能够长时间独立运行、感知环境，自主使用各种工具来完成复杂任务。 最近有研究甚至证明，智 能体也遵循大语言模型的测试时扩展（Test-Time Scaling）规律，能够通过强推理不断提升解题能力。 事实上，智能化发展的过程比我们想象得还要快。正在上海举行的亚马逊云科技中国峰会上，我们看到了一系列基于大模型、Agentic AI 的创新和案例，让我们眼 花缭乱。 「通俗说来，Agentic AI 就是让基于大模型的 AI 从『我问 AI 答』、『我说 AI 写』发展到『我说 AI ...

Group 1 - The stock of Jinko Environment (688466) has a comprehensive score of 51.09, indicating a strong performance [1] - The main cost analysis shows the following prices: 16.56 CNY for the day's main cost, 16.99 CNY for the 5-day main cost, 16.99 CNY for the 20-day main cost, and 16.38 CNY for the 60-day main cost [1] - Over the past year, the stock has reached its upper limit once and has not hit its lower limit [1] Group 2 - The short-term resistance level is at 17.23 CNY, while the short-term support level is at 16.83 CNY [2] - If the stock price falls below the short-term support level, it is advisable to adopt a wait-and-see approach; if it falls below the mid-term support level, the willingness of main funds to buy is weak, and a wait-and-see approach is also recommended [2] - The K-line patterns indicate potential topping out with formations such as "Three Black Crows" and "Double Flying Crows" [2] Group 3 - On June 20, 2025, the net inflow of main funds was 130,800 CNY, accounting for 2% of the total transaction volume [2] - Large orders showed a net inflow of 130,800 CNY, while retail investors experienced a net outflow of 478,300 CNY [2] - Related sectors such as the environmental protection industry, micro-cap stocks, and AI intelligence saw declines of -0.62%, -0.32%, and -2.17% respectively [2]

家电丨市场观察 本周看点： -奔向AI的厨电战争，该为谁打响？ -阿里、小米们加速"围攻"Meta，AI眼镜打响"持久战"; -空间计算机很好， 但苹果、INAIR、魅族谁更接近未来？ 行业环境 1.奔向AI的厨电战争，该为谁打响？ 关键词 ： 厨房,年轻人,智能化,烹饪,户外经济 概要 ： 在消费格局变化下，厨房对年轻人吸引力减弱，中国厨电企业通过AI技术融合推出智 能化产品吸引用户，如"食神"大模型。但市场对其高溢价和数据安全问题态度不一。行业还面 临房地产下行、海外竞争等挑战，企业尝试多元化发展、优化设计并加速出海。未来，厨电企 业需拓展饮食产业链，借助AI赋能降低下厨门槛，重塑市场格局，把握智能化发展机遇。 2.阿里、小米们加速"围攻"Meta，AI眼镜打响"持久战" 关键词 ： AI眼镜,智能硬件,市场表现,供应链,佩戴体验 概要 ： 扎克伯格认为智能眼镜将成未来手机，2024年一季度全球销量达60万台，同比增长 216%，远超VR和AR。Meta引领行业，苹果等加速布局，但中国产品多停留在"PPT"阶段。AI 眼镜因成本低、佩戴舒适受青睐，预计2025年销量达550万台。然而，技术设计平衡、SKU ...

Core Viewpoint - The article emphasizes the significant updates in HarmonyOS 6, showcasing a comprehensive embrace of AI and intelligent agents, marking a pivotal evolution in the operating system's capabilities [2][60]. Group 1: AI Integration - The new AI features include a video call capability for the Xiao Yi assistant, allowing real-time interaction and explanation of its surroundings [3][26]. - AI will enhance various system applications, including advanced photo editing capabilities, with AI-driven style effects and composition assistance based on over 500,000 images [5][20][22]. - The Xiao Yi assistant has been upgraded to integrate with Huawei's Pangu and DeepSeek models, utilizing a training dataset of 20 trillion tokens [14][15]. Group 2: Ecosystem Expansion - Over 3,000 applications and meta-services are currently in accelerated development to become more integrated with HarmonyOS [9]. - The HarmonyOS 6 developer Beta version introduces a new interconnected architecture, supporting over 660 applications for a smoother and more innovative experience [8]. - The launch of over 50 HarmonyOS smart agents, including popular applications like Weibo and DingTalk, highlights the ecosystem's growth [8][34]. Group 3: Cross-Device Connectivity - The "One Touch Share" feature allows users to share content across devices seamlessly, now supporting over 50 applications and enabling multi-device sharing without data consumption [37][40]. - The system supports reverse transmission, allowing users to transfer edited images back to their mobile devices effortlessly [54]. - The seamless integration across devices is showcased in a promotional video, demonstrating the fluid transfer of media between various Huawei devices [56]. Group 4: Developer Engagement - The event featured numerous developers showcasing their contributions to the HarmonyOS ecosystem, indicating a robust collaborative environment [57][58]. - The article suggests that HarmonyOS is not merely a replacement for Android but is designed for the AI era, emphasizing its unique capabilities and fully domestic development [59][60].

Group 1 - The 2025 World Mobile Communication Conference held in Shanghai highlighted AI intelligent agents as a key focus, showcasing innovations in smart terminals, communication technologies, and urban governance [1][4] - AI smartphones can autonomously interpret information, such as identifying medication and setting reminders, demonstrating advanced user interaction capabilities [1][3] - The concept of intelligent agents extends beyond smartphones to various devices like AI glasses and watches, enhancing personal task management and spatial understanding [3][4] Group 2 - The Wi-Fi sensing integration project in Shanghai aims to automatically detect personnel status and optimize energy consumption, achieving a 15% reduction in air conditioning energy use through AI algorithms [4][6] - The integration of AI with communication networks is leading to the emergence of urban intelligent agents, which enhance city governance by enabling real-time data-driven decision-making [6][8] - The development of city super intelligent agents allows for more complex urban tasks, enhancing the capabilities of city management systems beyond mere data collection and visualization [8]