广告
加载中

高德CEO郭宁:AI时代不缺答案,稀缺的是理解真实世界

田雨 2026-09-10 15:45
田雨 2026/09/10 15:45

邦小白快读

EN
全文速览

这篇内容是高德CEO在2026高德扫街榜发布会上的核心分享,和大众日常找店消费、出行做攻略的需求直接相关,能帮大家避开消费决策的常见坑。

1. 针对大家吐槽已久的虚假评分、刷出来的假评论问题,高德扫街榜不靠用户写的评价排名,而是以真实导航到店的用户行为作为计算依据,刷榜难度极高,榜单可信度更强。目前该榜单已覆盖8.8亿用户、260万商家,从美食拓展到全品类吃喝玩乐、文旅场景,还能查看飞行街景、动态时令榜单、好友动态,帮大家找到真正靠谱的消费目的地。

2. 针对AI做攻略常见的“看着合理实际踩坑”问题,高德正在升级空间智能能力,不光给推荐结果,还能通过3D街景让大家提前获得身临其境的在场感,行程规划会综合考虑场馆闭馆时间、道路限行、点位排队时长、交通衔接等上下文信息,避免行程安排脱节,帮大家获得顺畅的出行消费体验。

本次发布会披露的高德扫街榜运营现状、AI时代消费决策趋势,对品牌开展本地化营销、搭建用户触达渠道、优化消费体验有重要参考价值。

1. 当前消费决策领域的核心痛点是用户对传统刷分、虚假评论的信任度持续下降,商家投入大量营销预算却很难获得用户认可。而高德扫街榜以用户真实到店行为为排名依据,作弊门槛高、用户信任度强,上线一年已覆盖8.8亿用户,吸引260万商家入驻、全球50多个文旅单位合作,覆盖吃喝玩乐到跨境文旅全场景,平台还为入驻商家免费提供飞行街景服务,是品牌低成本触达用户、靠真实口碑建立信任的新渠道。

2. AI时代用户的消费决策需求已经升级,从“找得到推荐”转向对信任、独特品味、场景在场感、行程顺畅度的要求,品牌可配合平台的空间智能能力,优化门店的场景化展示、到店衔接服务,匹配用户新的决策需求。

本次发布会披露的高德扫街榜运营进展与平台未来升级方向,为本地生活、文旅类卖家指明了新的流量增长机会与运营思路。

1. 当前传统本地生活渠道的评价体系刷量内卷严重,卖家投入高但用户信任度低。高德扫街榜以用户真实导航到店行为作为核心排名依据,刷榜作弊难度极高,目前已有260万商家主动入驻,平台免费为商家提供飞行街景展示服务,没有额外入驻成本,是卖家摆脱无效刷量投入、靠真实服务质量积累口碑的新流量渠道。

2. AI时代用户的消费需求正在发生变化,通用AI生成的攻略无法满足用户对可信度、独特体验、场景感知、行程顺畅度的要求,平台未来会把单一榜单升级为覆盖发现、决策、出发、行程、抵达的全链路体验,卖家可提前完善门店的3D场景信息、到店时间提示、周边行程衔接指引,同时抓住平台覆盖境内外50多个文旅目的地的场景延伸机会,获取跨城、跨境的新增量。

本次高德披露的空间智能建设布局、本地生活场景升级方向,能为相关生产类工厂带来明确的商业机会,也为工厂推进数字化转型提供了可参考的思路。

1. 从商业机会看,高德正在搭建覆盖二维道路拓扑、三维空间结构、时间动态变化、人流车流实时流动的四层真实世界数字体系,推出全模态可预测3D原生城市世界模型,相关建设需要大量街景采集设备、三维建模内容、动态感知硬件的支撑,生产相关产品的工厂可对接这类新型B端采购需求,开拓新的业务增量。

2. 从数字化转型启示看,工厂在拓展线上渠道、推进电商布局时,可参考扫街榜的信任构建逻辑,放弃虚假刷评类的无效营销,通过真实生产场景、产品应用场景的可视化呈现建立客户信任,同时在对接客户需求时关注全链路的衔接体验,提升合作效率。

本次发布会披露的AI时代本地生活行业痛点、高德的技术布局与生态合作方向,能为服务商把握行业趋势、挖掘客户需求、设计适配解决方案提供清晰的参考。

1. 从行业趋势与客户痛点看,当前本地生活领域存在两类明确痛点:一是传统评价体系虚假评分、评论泛滥,商家营销投入高但用户信任度低;二是通用AI生成的消费决策内容可信度不足、同质化严重,缺乏场景代入感,无法适配真实场景下的动态时空衔接需求,用户决策体验差。

2. 从解决方案方向看,服务商可围绕高德的空间智能能力,为入驻商家提供飞行街景素材拍摄、3D门店场景搭建、行程衔接提示优化等运营服务,也可对接各地文旅单位的需求,帮其在高德平台落地场景化的文旅服务内容,匹配平台倡导的信任、品味、在场感、上下文的价值导向,挖掘新的服务机会。

高德在本次发布会上分享的扫街榜运营经验、AI时代的平台升级逻辑,为各类生活服务、内容服务平台提供了可借鉴的运营方法与风险规避方向。

1. 从可参考的平台运营做法看,高德扫街榜上线一年就积累8.8亿用户、260万入驻商家,核心是跳出传统评价体系的用户生成内容逻辑,以用户真实导航到店的行为作为排名依据,大幅提高作弊门槛,直击用户信任缺失的核心痛点,同时为入驻商家免费提供飞行街景服务,联动全球50多个文旅单位拓展场景边界,快速扩大平台生态规模。

2. 从平台发展风向与风险规避看,AI时代平台不能仅靠大模型生成通用内容留住用户,要围绕用户对信任、品味、在场感、上下文的核心需求,搭建对真实世界时空动态的理解能力,把单一的工具功能升级为覆盖用户全链路需求的完整体验,同时主动规避虚假内容、AI生成内容脱离实际场景的运营风险,筑牢用户信任基础。

本次高德发布会披露的扫街榜运营数据、对AI时代消费决策特征的判断、空间智能的布局方向,为数字经济、本地生活、AI应用领域的研究者提供了鲜活的产业研究样本。

1. 从产业新动向与新问题看,当前AI大模型在消费决策领域的应用已经出现明显的供需错配:AI生成通用内容的门槛大幅降低,用户并不缺少标准化答案,但内容信任度不足、同质化严重、缺乏场景代入感、无法适配动态时空衔接等问题持续凸显,对真实世界的理解能力成为AI落地实体消费场景的核心稀缺能力。

2. 从商业模式创新看,高德跳出传统本地生活平台的评价、推荐逻辑,以真实用户行为数据构建信任基础,依托四层真实世界数字化能力打造空间智能,把单一的榜单工具升级为覆盖消费全链路的真实世界体验入口,同时联动商家、文旅单位搭建开放生态,这种“大模型+空间智能”的落地路径,为AI在实体场景的商业化提供了新的研究方向。

返回默认

声明:快读内容全程由AI生成,请注意甄别信息。如您发现问题,请发送邮件至 run@ebrun.com 。

我是 品牌商 卖家 工厂 服务商 平台商 研究者 帮我再读一遍。

Quick Summary

This article features key takeaways from the speech by Amap’s CEO at the 2026 Amap Street Sweep List launch event, closely tied to people’s daily needs for finding shops, making consumption decisions, and planning travel itineraries, to help audiences avoid common pitfalls in decision-making.

1. Addressing longstanding user frustration over fake ratings and paid fraudulent reviews, the Amap Street Sweep List does not rely on user-generated reviews for rankings. Instead, it uses verified real-world user navigation and store visit data as the core calculation metric, making ranking manipulation extremely difficult and the list far more credible. To date, the list covers 880 million users and 2.6 million merchants, expanding from dining recommendations to all local lifestyle categories including dining, entertainment, and cultural tourism. It also integrates aerial street view, dynamic seasonal rankings, and friend activity updates, helping users find truly reliable destinations.

2. To solve the common problem of AI-generated travel itineraries that look reasonable on paper but lead to bad real-world experiences, Amap is upgrading its spatial intelligence capabilities. Beyond basic recommendations, it uses 3D street view to give users an immersive, on-site preview of destinations. Its itinerary planning also accounts for contextual factors including venue closing times, road restrictions, queue lengths at locations, and transit connections, to prevent disjointed plans and deliver smooth travel and consumption experiences.

The operating status of the Amap Street Sweep List and the AI-era consumption decision trends disclosed at this launch event provide valuable reference for brands to carry out localized marketing, build user reach channels, and optimize consumer experiences.

1. The core pain point in the current consumption decision landscape is eroding user trust in traditional paid ratings and fake reviews: brands invest large sums in marketing budgets yet struggle to win user recognition. By contrast, the Amap Street Sweep List ranks venues based on real user store visit behavior, with high barriers to manipulation and strong user trust. One year after launch, it covers 880 million users, hosts 2.6 million onboarded merchants, and has established partnerships with over 50 cultural and tourism authorities worldwide, covering scenarios from local dining and entertainment to cross-border tourism. The platform also provides free aerial street view services for onboarded merchants, creating a new channel for brands to reach users at low cost and build trust through authentic word of mouth.

2. In the AI era, user demands for consumption decisions have upgraded from “finding recommendations” to requiring trust, unique taste, immersive on-site experience, and smooth itinerary flow. Brands can leverage the platform’s spatial intelligence capabilities to optimize scenario-based store presentation and last-mile arrival services, to align with users’ updated decision-making needs.

The operational progress of the Amap Street Sweep List and the platform’s future upgrade direction disclosed at the event point to new traffic growth opportunities and operational strategies for local lifestyle and cultural tourism sellers.

1. Traditional local lifestyle channels currently suffer from severe rating manipulation and cutthroat fake review competition, leaving sellers with high input costs but low user trust. The Amap Street Sweep List uses real user navigation and store visit behavior as its core ranking metric, making manipulation extremely difficult. To date, 2.6 million merchants have voluntarily joined the platform, which provides free aerial street view display services for merchants with no extra onboarding costs, creating a new traffic channel for sellers to escape wasteful spending on fake traffic and build reputation through actual service quality.

2. User consumption demands are shifting in the AI era: generic AI-generated itineraries fail to meet user requirements for credibility, unique experiences, scenario perception, and smooth itinerary flow. The platform will upgrade its standalone ranking function into an end-to-end experience covering discovery, decision-making, departure, en-route travel, and arrival. Sellers can proactively improve their stores’ 3D scenario information, arrival time prompts, and surrounding itinerary connection guidance, while seizing opportunities from the platform’s scenario expansion covering over 50 domestic and cross-border cultural tourism destinations to capture incremental cross-city and cross-border traffic.

Amap’s disclosed spatial intelligence layout and local lifestyle scenario upgrade direction offer clear business opportunities for relevant manufacturing factories, as well as actionable references for factories advancing digital transformation.

1. In terms of business opportunities, Amap is building a four-layer digital system of the real world covering 2D road topology, 3D spatial structure, dynamic temporal changes, and real-time flows of people and vehicles, and launching a fully modal, predictive 3D native city world model. This infrastructure buildout requires large volumes of street view collection equipment, 3D modeling content, and dynamic sensing hardware. Factories producing related products can tap into these new B-end procurement demands to open up new business growth streams.

2. In terms of digital transformation takeaways, when expanding online channels and developing e-commerce layouts, factories can reference the Street Sweep List’s trust-building logic: abandon ineffective marketing tactics such as fake review brushing, build customer trust through visual presentation of real production scenarios and product application scenarios, and focus on end-to-end connection experiences when responding to customer needs to improve cooperation efficiency.

The AI-era local life industry pain points, Amap’s technology layout, and ecosystem cooperation directions disclosed at the event provide clear references for service providers to grasp industry trends, identify customer needs, and design adapted solutions.

1. In terms of industry trends and customer pain points, two clear pain points exist in the current local life sector: first, traditional review systems are flooded with fake ratings and reviews, leading to high merchant marketing investment but low user trust; second, generic AI-generated consumption decision content lacks credibility, is highly homogeneous, fails to deliver immersive scenario experience, cannot adapt to dynamic spatiotemporal connection needs in real scenarios, and results in poor user decision-making experiences.

2. In terms of solution directions, service providers can build operational services for onboarded merchants around Amap’s spatial intelligence capabilities, including aerial street view material shooting, 3D store scenario construction, and itinerary connection prompt optimization. They can also respond to the needs of local cultural and tourism authorities to help them launch scenario-based cultural tourism services on the Amap platform, aligning with the platform’s advocated value orientation of trust, taste, immersive presence, and contextual awareness, to tap into new service opportunities.

Amap’s sharing of Street Sweep List operational experience and AI-era platform upgrade logic at this event provides replicable operational methods and risk avoidance guidance for various lifestyle service and content service platforms.

1. In terms of replicable platform operation practices, the Amap Street Sweep List accumulated 880 million users and 2.6 million onboarded merchants within one year of launch. Its core success lies in breaking away from the user-generated content logic of traditional review systems, using real user navigation and store visit behavior as the ranking basis to drastically raise manipulation barriers and directly address the core pain point of eroding user trust. Meanwhile, it provides free aerial street view services for onboarded merchants and partners with over 50 global cultural and tourism authorities to expand scenario boundaries, rapidly scaling up the platform ecosystem.

2. In terms of platform development trends and risk avoidance, in the AI era, platforms cannot retain users solely by relying on large models to generate generic content. They need to build capabilities to understand the real world’s spatiotemporal dynamics around users’ core demands for trust, taste, immersive presence, and contextual relevance, upgrade standalone tool functions into complete experiences covering users’ end-to-end needs, and proactively mitigate operational risks from fake content and AI-generated content that is disconnected from real scenarios, to solidify the foundation of user trust.

The Street Sweep List operational data, judgments on AI-era consumption decision characteristics, and spatial intelligence layout direction disclosed at Amap’s launch event provide a vivid industry research sample for researchers in the digital economy, local life, and AI application fields.

1. In terms of new industry trends and emerging issues, the application of large AI models in the consumption decision field currently faces a clear supply-demand mismatch: while the threshold for AI to generate generic content has dropped sharply and users are no longer short of standardized answers, problems including insufficient content trustworthiness, severe homogeneity, lack of immersive scenario experience, and failure to adapt to dynamic spatiotemporal connections remain prominent. Understanding of the real world has become the core scarce capability for AI to land in physical consumption scenarios.

2. In terms of business model innovation, Amap has broken away from the traditional local life platform’s review and recommendation logic, building a trust foundation with real user behavior data, developing spatial intelligence based on four-layer real-world digital capabilities, upgrading the standalone ranking tool into a real-world experience entry covering the full consumption chain, and building an open ecosystem with merchants and cultural tourism authorities. This “large model + spatial intelligence” implementation path provides a new research direction for the commercialization of AI in physical scenarios.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

【亿邦原创】9月10日,在2026高德扫街榜发布会上,高德CEO郭宁分享了过去一年扫街榜的发展,以及AI时代高德对“真实世界”的新思考。

在郭宁看来,当AI让“答案”越来越容易获得,真正稀缺的东西变成了:信任、品味、在场感和上下文。而这些能力的背后,是对于真实世界的理解。

01

扫街榜一年:8.8亿用户260万商家

“越来越多的虚假评分、虚假评论,商家投入越来越多的营销预算,但是用户反而越来越不敢相信。”郭宁表示,这是高德推出扫街榜最初想解决的问题。

与传统评价体系相比,高德扫街榜强调通过真实行为进行计算,让用户“用脚投票”。2025年9月10日,高德正式发布扫街榜时,就将其定位为基于用户行为产生的线下餐饮及生活服务榜单。

在郭宁看来,写一条评论很容易,但真正导航30公里去一家店,是更难被“刷”出来的真实行为。高德希望通过这种更难作弊的计算方式,重新建立消费者与商家之间的信任。

一年过去,扫街榜的影响力超过了高德最初的预期。

郭宁在现场披露,目前已有8.8亿用户在高德评分、发现身边的吃喝玩乐;已有超过260万商家主动入驻高德,高德还免费为商家提供飞行街景服务;同时,全球已有超过50个文旅合作单位与高德展开合作。

事实上,高德此前已经持续将扫街榜从美食扩展到吃喝玩乐等更多生活服务场景。

2026年1月,扫街榜上线100天时,用户规模已突破6.6亿,并推出飞行街景、动态时令榜单、好友动态等功能。

与此同时,扫街榜的应用场景也进一步延伸至文旅领域。据介绍,目前已有包括广东、浙江、上海、北京、香港、澳门以及新加坡、越南、泰国等地在内的文旅机构与高德合作。

在郭宁看来,扫街榜真正想建立的并不只是一个“更真实的榜单”。“真实”最终指向的是“信任”。

而随着AI进入消费决策,另一个问题正在出现:当AI可以快速生成答案,下一个时代真正稀缺的东西又是什么?

02

AI时代,更稀缺的是:信任 品味 在场感 上下文

“AI时代,生成看起来真实的内容正在变得越来越容易。”

郭宁举了个例子。如果把1万条评价交给AI,它可以迅速总结出“本地10家必吃”;告诉大模型时间、预算和同行人数,它甚至可以在几秒钟内生成一套完整攻略。

“未来,我们可能并不缺少答案。”郭宁说道,那么,什么才是真正稀缺的?

郭宁给出的答案是四个词:信任、品味、在场感、上下文。

首先是信任。AI可以帮人汇总大量评价,也可以告诉用户一家店为什么受欢迎,但它无法保证这个答案一定经得起真实世界的验证。

第二个是品味。大模型能拟合大众口味的最大公约数,但是不能给出一个人独特的视角和近乎直觉的选择。

第三个是在场感。模型能列出一份目的地清单,但是不能让人身临其境。“同一个目的地,看网上的一张照片和身临其境的感受有巨大区别。”

第四个是上下文。郭宁举了一个例子:如果把太平山缆车、晚餐预约和最后一班渡轮安排在同一个晚上,看起来每一个节点都没有问题。但如果缆车排队比预计多了30分钟,后面的晚餐和渡轮就可能全部受到影响。

“行程不是清单上一个又一个孤立的点,而是一条牵一发而动全身的时空链条。”他说道。

也就是说,AI给出的每一个答案单独看可能都是正确的。但真正进入现实世界后,它们之间还有大量关系需要被理解。

郭宁表示,信任的前提,是理解真实世界中的行为;品味的前提,是理解不同人的视角;在场感的前提,是理解真实世界的空间结构;上下文的前提,是理解真实世界的时间序列。

最终,它们指向理解真实世界的能力。

03

用“空间智能”把榜单变成一场真实世界体验

郭宁认为,真实世界并不只是存在于语言里,它还存在于真实的网络拓扑、时间空间、车流人流,以及人与世界关系的推演预测中,存在于对真实世界的时空动态理解。

高德希望把这些信息统一起来。

在他的描述中,真实世界可以被拆成四层:第一层是二维网络拓扑,即道路、节点以及彼此之间的连接关系;第二层是三维空间结构,包括地库的深度、高架桥的高度、建筑、道路、山川等;第三层是时间变化,例如博物馆周一闭馆、一条道路在特定时间限行;第四层是真实世界的流动,包括人流、车流以及各种动态变化。

当二维网络、三维空间、时间变化与真实世界的流动被统一起来,高德认为,这将形成一种不同于传统搜索和推荐的新型智能:空间智能。

“它不是更强的搜索引擎,不是更好的推荐算法,而是一个真正理解物理真实世界的空间智能。”郭宁说。

他进一步表示:“如果说大模型理解语言,那我希望高德空间智能能够理解世界。”

这也解释了为什么此次高德强调“扫街,不止于榜”。榜单只是用户发现真实世界的起点。

在消费决策阶段,高德希望通过真实行为帮助用户找到真正值得信赖、有品味的地方;在出发前,通过空间信息让用户提前获得“在场感”;在行程规划中,理解不同地点之间的时间、空间关系,提前规避风险;抵达之后,则让空间变得连续、可读,用户可以边走边聊、实时互动。

高德正在尝试把扫街从一个“找什么”的榜单,扩展成覆盖发现、决策、出发、行程、抵达的完整体验。

发布会上,郭宁还介绍了基于世界模型的ABot-Earth。“我们基于全球首个全模态可预测3D原生的城市世界模型,还原了地球的每一个角落,把全球的空间、时间、现实信息组织成一个可以连续进入、自由探索、可交互、可模拟推演的三维时空世界。”

归根结底,高德现在试图回答的是另一个问题:当AI越来越会生成答案,地图如何进一步理解人们生活的真实世界?

“基于对真实世界的理解,我们希望能够重塑扫街体验,把信任、品味、在场感、上下文这些最稀缺的部分,通过高德空间智能一气呵成。”郭宁补充道,“扫街本来就不只是一个孤立的榜单,而是包括榜单在内的一连串可推演的完整体验。”

亿邦持续追踪报道该情报,如想了解更多与本文相关信息,请扫码关注作者微信。

本文首发于 亿邦动力 官方网站

文章来源:亿邦动力

广告
微信
朋友圈

FAQ回顾

高德扫街榜是什么?

高德扫街榜是高德推出的基于用户真实行为计算的线下餐饮及生活服务榜单,主打难以作弊的用户导航到店等真实行为“用脚投票”的评价机制,旨在解决虚假评分评论问题、重建消费信任,目前已覆盖吃喝玩乐、文旅等场景,拥有8.8亿用户、超260万入驻商家。

AI时代消费决策领域真正稀缺的核心能力是什么?

AI可快速生成评价汇总、出行攻略等内容,但消费决策领域真正稀缺的是信任、品味、在场感、上下文四大核心能力,这些能力的底层支撑是对真实世界的用户行为、多元视角、空间结构、时间序列的深度理解,而非单纯的内容生成能力。

高德提出的空间智能是什么?

高德提出的空间智能是区别于传统搜索、推荐算法的新型智能,核心是整合真实世界的二维道路拓扑、三维空间结构、时间变化规律、人流车流等动态信息,实现对物理真实世界的深度理解,可支撑用户从消费决策到行程落地的全流程服务。

这么好看,分享一下?

朋友圈 分享

APP内打开

赞 +1
+1
微信好友 朋友圈 新浪微博 QQ空间
关闭
收藏成功
发送
/140 0