运营商财经 康钊/文 最近，高德扫街榜开了一个发布会，公布了上线百日以来的成绩，以及官宣进行了全面升级，推出了一 些新功能。 其中最引起我关注的是高德扫街榜基于世界模型规模化落地应用的"飞行街景"功能。 第一，早在一两天前，就有媒体报道称，高德已布局世界模型，并且基于该模型还会推出新的产品应 用。 万万没想到，速度会这么快，基于世界模型而打造的新东西就已经面世了，不得不说，高德扫街榜能够 快速崛起并不是没有道理的，这种科技落地速度就让人惊叹。 第二，科技落地出发点。有一个细节大家发现没有，高德扫街榜基于世界模型推出的"飞行街景"功能， 是真的在解决老百姓痛点问题的。 现在很多人之所以不相信网上的榜单，不愿意到线下消费，就是因为很多店铺线下和线上完全对不上， 失真很严重。 但是现在有了"飞行街景"这个新功能就不一样了，通过这个功能，用户在出门之前就可以直接以第一视 角飞越城市街区，从高空俯瞰门店周边街道情况，看看有没有停车位设施之类的。 同时还能实地探访门店包厢环境、店里座位分布情况等等，将门店最真实的情况，在出发之前就掌握在 手里，真正实现所见即所得。这一点是非常赞的。 我之所以特别关注到它，主要原因有二: ...

这意味着，扫街榜这一产品在体量上，已经有了庞大的用户和商家基础。不仅激发线下经济，也在助推高德突破流量天花板：根据QuestMobile数据， 2025年10月，高德APP月活跃用户数（MAU）逼近10亿大关，达到9.96亿，短短一个月内新增了近4600万月活用户。 在"种草"可能被异化为"割草"，好评也可以被量产的时代，高德扫街榜选择了一条最笨、也是最难的道路：摒弃流量思维，以真实行为为核心逻辑，做现 实世界的忠实映射。事实证明，在当下，真实恰恰是扫街榜最大魔力，近半数中国人"用脚投票"出来的选择，打开了通往烟火经济的入口。 今天，随着扫街榜2026正式发布，高德以真实数据为基础，利用海量时空数据积累和"世界模型"，围绕数字、信任、科技全面进化，继续向传统本地生活 市场发起一场变革。 真实重构信任 老北京胡同里的"隐形"小店，曾是城市商业难以触达的末梢。 道路狭窄、地形复杂、停车困难等物理空间的限制，让许多坚守传统手艺的"苍蝇馆子"难以被发现。在店里忙活了十多年的店长刘师傅，是"锅儿挑"面馆 的掌勺大厨，这家藏在北京雍和宫附近五道营胡同深处的小店，至今还保留着老北京人"面条出锅不过水"的地道吃法。 刘师傅守 ...

Core Insights - The launch of Gaode's "Street Ranking" achieved over 40 million users on its first day and surpassed 660 million users within 100 days, indicating a significant natural growth in user engagement [1] - The initiative has attracted over 86,000 merchants, with their revenue increasing by over 270% in the same period, showcasing the product's impact on local economies [1][9] - Gaode's approach focuses on real user behavior rather than traditional traffic acquisition methods, emphasizing authenticity and trust in the local service market [2][10] User Growth and Engagement - By October 2025, Gaode's monthly active users (MAU) reached 996 million, with an increase of nearly 46 million users in just one month [1][9] - The "Street Ranking" has become a significant driver for Gaode's user base expansion, moving beyond its traditional transportation services [9] Merchant Impact - The "Street Ranking" allows small, lesser-known shops to gain visibility through a "foot voting" mechanism, which aggregates real user choices rather than relying on paid promotions [4][7] - Merchants like Liu's noodle shop experienced a 50% increase in daily customer flow after being featured on the ranking, demonstrating the effectiveness of this new exposure method [6][9] Market Dynamics - The traditional local service market has been challenged by the rise of fake reviews and high marketing costs, making Gaode's real behavior data a valuable asset for small businesses [3][10] - The "Street Ranking" model is seen as a blue ocean opportunity for merchants, providing a more efficient way to attract customers without the burden of high marketing expenses [6][9] Technological Advancements - Gaode's "World Model" technology enhances the accuracy of its digital mapping and user experience, allowing for immersive features like "flying street views" [12] - The model has been recognized as a leader in global assessments, indicating its high fidelity in representing the physical world [12] Strategic Vision - Gaode aims to redefine the local service landscape by leveraging its extensive data assets and technological innovations, creating a bridge between the digital and physical worlds [10][12] - The focus on authenticity and real user experiences positions Gaode as a transformative player in the local service market, fostering growth for small businesses and enhancing user trust [2][10]

在最新的 CES 2026 上，英伟达（NVIDIA）首席执行官黄仁勋发表了主题演讲。这次发布会被业界视为从"生成式AI"转向" 物理AI "与" 推理型AI "的历 史性节点。 在本地发布会上，黄仁勋 宣布 Vera Rubin平台全面量产： 作为Blackwell 的继任者，Rubin架构正式亮相并宣布已进入全面投产阶段，预计2026年下半年 交付。其推理性能是前一代Blackwell的5倍，训练性能是Blackwell的3.5倍， 生成AI Token的成本直接降低至原来的1/10。 最关键的是，黄仁勋在演讲中强调，AI的下个阶段是物理AI，不仅要懂语言，也要懂物理世界，并发布了三款平台或应用模型： 1. Cosmos物理AI平台：英伟达发布了Cosmos系列基础模型（包括 Reason2、Predict2.5等）。它能通过视频和遥测数据学习物理规律，在虚拟的 Omniverse 空间中生成大规模合成数据。这意味着 AI 不再只是"猜下一个词"，而是能理解重力、碰撞和物体的物理属性。 2. Alpamayo推理型智驾模型：这是首个具备"思考"能力的自动驾驶模型。它不再仅仅依赖预设规则，而是能像人类一样对 ...

最近在复盘各家如何使用世界模型的，今天和大家盘一下理想在这方面的工作。理想对世界模型的定义在 重建+生成 ，利用重建技术（3DGS）建模自动驾驶场 景，再利用生成方法实现闭环仿真或者场景生成。 这里面核心的技术是3DGS和生成。 ICCV2025中稿的Hierarchy UGP，自动驾驶场景重建。 具有时空一致性的多风格自动驾驶场景生成算法StyledStreets。 整合多模态驾驶意图与潜在世界模型实现合理规划的World4Drive，中稿ICCV 2025。 自动驾驶的视频生成扩散世界模型GeoDrive。 统一生成视觉与lidar的自驾世界模型框架OmniGen，中稿ACMMM2025。 结合强化学习的自动驾驶视频生成世界模型RLGF，中稿NeurIPS 2025。 利用稀疏注意力实现4D OCC世界模型预测算法 SparseWorld-TC。 世界模型端到端闭环强化学习框架AD-R1。 整体上来看，世界模型是围绕视频为核心搭建的时空认知系统 ，这一点也和蔚来任少卿的观点一致。通过跨模态的互相预测和重建，促使系统学习时空和物理规 律，让机器能像人一样理解环境。 通过重建+生成，既可以做云端的数据生成，也 ...

高德在本地生活服务领域的进一步渗透。 上线百日，高德地图旗下的生活服务榜单产品"扫街榜"再迭代。1月7日，高德宣布"扫街榜"完成升级， 依托自研世界模型技术，推出"飞行街景"功能，并将榜单覆盖范围从美食扩展至更多场景。 此次升级均指向高德在本地生活服务领域的进一步渗透，作为阿里巴巴旗下本地生活板块重要抓手，高 德正试图利用其地图底层数据和流量优势，在本地生活市场中寻找差异化路径。 同时，高德尝试降低商户端飞行街景落地门槛。高德介绍称，商家仅需使用手机拍摄并上传视频和照 片，系统即可免费生成专属的飞行街景。高德方面宣布，将投入算力资源，免费为100万家中小商户 （即"烟火小店"）接入该功能，吸引更多长尾中小商家入驻，从而丰富高德的内容生态，促进线上流量 向线下交易的转化。目前，故宫博物院的飞行街景已上线，显示该技术在文旅场景应用潜力。 针对中小微商户，高德发布对应扶持计划，宣布由平台承担算力成本，为100万家小微商户免费接入前 述"飞行街景"功能。此外，高德计划与全国100个城市合作，在降低商户入驻门槛的同时，提供数据分 析、出行补贴等资源。目前，上海、四川、济南等地已先行启动相关合作。 地图类App正加速从" ...

Core Viewpoint - Alibaba's Gaode has launched the upgraded Gaode Street Ranking 2026, introducing the "Flying Street View" feature, which allows users to view real store environments from a street perspective, and is opening this feature to 1 million merchants for free [1][2]. Group 1: User Growth and Engagement - Gaode's user base has surpassed 660 million, with the app gaining an average of 46 million monthly active users over the past three months, bringing the total to 996 million [1]. - The Gaode Street Ranking has attracted 860,000 new merchants within 100 days of its launch, with merchant order volume increasing by over 330% and revenue rising by over 270% [1]. Group 2: New Features and Offerings - The "Flying Street View" feature allows users to preview restaurant environments and parking conditions, enhancing decision-making efficiency and helping merchants showcase their stores [2][3]. - Gaode has introduced seasonal and location-based dynamic service rankings, along with 1,550 category rankings covering various activities such as hiking and camping [3]. - The platform now offers personalized ranking creation, enabling users to create and share their own exclusive lists [3].

基于高德自研世界模型，高德扫街榜首推"飞行街景"功能。"飞行街景"通过高拟真数字还原，实现了连 续、动态、与现实几乎一致的实景导航和实景探店，帮助用户到餐厅、景点前即可沉浸式俯瞰内外部实景，"实 地"体验沿途风景、停车位、餐厅包厢环境等。即日起，用户在高德APP搜索"飞行街景"，即可体验。 这项新功能也为商家提供了展示店铺特色的新方式，有助于提高生意转化。商家仅需一部智能手机，根据 指引拍摄并上传几段店内视频和门头照片，即可免费生成专属飞行街景，更加有效地把线上"种草"转化为线下 到店。 1月7日，上线100天的高德扫街榜宣布三大升级：依托世界模型发布全球首个"飞行街景"，实现从街景一路 俯瞰至店内实景；推出全球首个应季、应时、应地的生活服务动态榜单；引入好友关系，新增好友动态和个人 榜单功能。经过100天的发展，高德扫街榜已经从全国最大的美食榜单扩展至"吃喝玩乐"全场景，成为本地生活 超级入口。 真实榜单带来真实增长。当前高德扫街榜用户规模已经突破6.6亿，根据QuestMobile的最新数据，在扫街榜 带动下，高德APP单月新增4600万月活用户，月活用户数增至9.96亿。高德扫街榜上线100天来，累计吸 ...

Core Insights - The article discusses the innovative incubation model of the "Pazhou Mofang" incubator in Guangzhou, which focuses on nurturing startups with real purchase orders, thereby ensuring practical application and commercial viability of their technologies [2][3][4]. Group 1: Incubator Model - The Pazhou Mofang incubator is characterized by its unique approach of "incubating with orders," allowing projects to directly apply their products in real-world scenarios [3][4]. - The incubator was established by Haizhu Chengfa Group in collaboration with Qingzhi Incubator and operates under the guidance of Tsinghua University's Intelligent Industry Research Institute [3][4]. - The selection criteria for startups include original innovation, product maturity, team depth, and market validation, ensuring that only viable projects are supported [4][5]. Group 2: Phased Development - The incubation process is divided into two phases: a 6-month technical incubation followed by a 6-month commercial incubation, with expert guidance from leading tech companies like NVIDIA and Microsoft [5][7]. - During the technical phase, projects receive industry classification scores and targeted technical resources based on identified weaknesses [5][7]. - The commercial phase focuses on planning business paths and connecting startups with downstream industry resources to accelerate commercialization [7][10]. Group 3: Real-World Application - The incubator emphasizes real-world training environments, allowing startups to test their technologies in various high-traffic scenarios, such as hotels and retail settings [10][11]. - As of May 2025, the Pazhou Mofang has successfully incubated 29 intelligent projects, with 70% being original research projects established that year [10][11]. - The outcomes span three levels of the industry chain: technology, hardware, and application, with results already integrated into retail, e-commerce, industrial, and educational sectors [10][11]. Group 4: Regional Context - The Pazhou Mofang aims to leverage Guangzhou's complex industrial landscape, which includes manufacturing, circulation, consumption, and service sectors, to foster AI applications [11][12]. - Unlike other regions, Guangzhou's strength lies in its real-world industrial scenarios rather than algorithmic innovation or capital density, making it a unique incubator environment [11][12]. - The incubator's strategy reflects a shift in Guangzhou's economic focus towards a modernized industrial system, emphasizing the integration of manufacturing and service sectors [12][15]. Group 5: Future Prospects - The long-term goal of the Pazhou Mofang is to cultivate AI companies that can thrive commercially and contribute to the regional industrial ecosystem by 2028 [20]. - The operational model encourages a dynamic flow of resources, allowing successful projects to transition out and make room for new teams, thus maintaining a fresh and innovative environment [16][20]. - The success of this model will depend on its ability to continuously attract technology and demand while minimizing trial-and-error costs in the incubation process [20].

免责声明：本文内容与数据由观点根据公开信息整理，不构成投资建议，使用前请核实。 高德CEO郭宁在发布会上透露，目前高德扫街榜用户规模已突破6.6亿；QuestMobile最新数据显示，受 扫街榜带动，高德APP单月新增月活用户4600万，月活总数升至9.96亿。 观点网讯：1月7日，高德在北京发布全球首个基于自研世界模型的"飞行街景"，并宣布将为100万家"烟 火小店"接入该服务，实现连续、动态、与现实几乎一致的实景导航与探店体验，首次把世界模型应用 于普惠生活服务行业。 ...