奥运赞助的范式转移：从 ATS 到数据巨头
长达 35 年，每一次奥运会的屏幕呈现、每一次时间分割、每一次奖牌统计、每一次得分，都曾通过一家名为 ATS 的法国 IT 巨头的服务器运行。这家公司自 1992 年巴塞罗那奥运会起，便从未间断地为奥运会提供技术支持，处理了数亿条赛事结果，并抵御了数亿次网络攻击。然而，在平昌冬奥会前夕，一款名为“Olympic Destroyer”的恶意软件窃取了 ATS 员工的凭证，在开幕式期间瘫痪了官方奥运网站，并导致场馆 Wi-Fi 和票务销售中断数小时。尽管 ATS 修复了漏洞并继续运作，但其内部已然分崩离析。到 2024 年巴黎奥运会，这家公司股价暴跌 98%，负债数十亿美元，甚至需要法国政府的紧急贷款才能维持生存。巴黎奥运会后，ATS 结束了长达 35 年的奥运顶级赞助合作。取而代之的是 Google、Snowflake、Cisco 和 Deloitte 等科技公司，它们正在重新定义奥运赞助的意义。奥运会正逐渐演变为一个为期 17 天、全球规模的 AI 概念验证（Proof of Concept）平台，这彻底改变了奥运会的经济模式。旧模式依赖于转播权和品牌曝光，而新模式则通过数据、算法和基础设施的价值进行变现，这种价值会随着奥运周期的推进而不断增长。对赞助商而言，这是一个现场演示；对国际奥委会（IOC）而言，这是一条永不终结的收入流；对粉丝、运动员和主办城市而言，这意味着参与到能够产生数十亿美元洞察的系统中。
Original English
For 35 years, every Olympic result you ever saw on screen, every split time, every medal count, every score ran through the servers of a single company, a French IT giant called ATS. They started at Barcelona in 1992 and never stopped. By the time they reached Paris 2024, ATS has processed over 300 million results across a dozen Olympics and defended the games against hundreds of millions of cyber attacks. And then just before the Winter Olympics in Pyongqen, a piece of malware called Olympic Destroyer collected a TOSS employee credentials, crashed the official Olympics website during the opening ceremony, knocked out Wi-Fi across the venue, and killed the ticket sales for hours. The company that was supposed to be the tech fortress of the games had been breached since December, and nobody caught it until the lights went on. ATAS patched the hole and carried on. But in the background, the company was falling apart. And by 2024, as it delivered the final Olympics in Paris, its stock has fallen 98%. It was drowning in billions of debt and begging the French government for an emergency loan just to survive. After Paris 2024, ATS exited the Olympics top sponsorship program after 35 years of Olympic partnership. And into the void stepped Google, Snowflake, Cisco, and Deote. The company that knew more about Olympic operations than anyone on Earth almost collapsed. But the companies stepping in are redefining what Olympic sponsorship is supposed to be about. The Olympic is becoming a live 17-day proof of concept for AI at global scale. That changes the economics of the Olympic Games. The old model monetized broadcast rights and visibility. The new model monetizes data, algorithms, and infrastructure that compounds in value across Olympic cycles. For sponsors, it's a live demo. For the International Olympic Committee, it's a revenue stream that doesn't end when the games do. And for fans, athletes, and host cities, it means participating in systems designed to generate insights worth billions of dollars. Today is about that shift. how the Olympics is evolving from a broadcast event showcasing the extremes of human abilities into a data platform where athletic competition runs alongside industrialcale AI experimentation. Let's dive in.体育科技的演进：从广告位到数据基础设施
当我们审视赛道旁的标识时，会发现奥运赞助的本质已发生根本性变化。过去，公司赞助奥运会主要是为了广告空间，例如可口可乐的标志出现在赛场边，但这并不直接促使观众立刻购买。如今，这些标识代表着购买对奥运会特定方面的“访问权”。从 ATS 的时代到如今的科技巨头，这种转变揭示了奥运会正在成为什么。2026 年米兰-科尔蒂纳丹佩佐冬奥会是 ATS 退出后的首届奥运会，其赞助商阵容已截然不同：Juniper Networks（网络安全与网络运营）、Alibaba Cloud（奥运会 GBT 聊天机器人和徽章交易站提供商）以及官方技术合作伙伴 Deloitte（负责奥运粉丝数据平台，该平台已从巴黎 2024 扩展至米兰）。然而，值得注意的是，在核心合作伙伴阵容中，并没有一个纯粹的“数据平台”公司。米兰科尔蒂纳丹佩佐奥运会开始试水 AI，例如实时检测关键时刻、奥运会 GBT 聊天机器人、为运动员提供第一视角画面的无人机，以及 Deloitte 的粉丝数据平台。这些都是“锦上添花”的功能，虽然有趣，但并非核心。
即将到来的洛杉矶 2028 年奥运会（LA28）将截然不同。LA28 不仅仅是为奥运会增加数据基础设施，它本身就是数据基础设施。LA28 将成为首届将 AI、云基础设施和数据平台作为基础赞助类别奥运会。其技术栈堪比典型的企业级 SaaS（软件即服务）规划会议，包括 Google Cloud、Snowflake 的 AI 数据云、Cisco 的 AI 网络、Deloitte 的技术咨询以及 T-Mobile 的 5G。因此，关键问题不再是 LA28 是否会收集数据，而是它究竟要用这些数据构建什么。
Original English
Look at whose logos are on the sides of tracks. When a company sponsors the Olympics, this isn't really about advertising space. I mean, okay, you see the Coca-Cola logo on the side. So what? It's not like you're going to run and get a lot of coke all of a sudden. It is about these logos buying access to certain aspects of the games. And the shift from a toas to what comes next tells you everything about what the Olympics is becoming. Milano Cartina 2026 is the first Olympics after ATS's exit. And the sponsor roster reads very differently. In Milan, the partners are Juniper Networks, cyber security and network ops. It's their first sponsorship of the Olympics, by the way. Alibaba Cloud, the provider of the Olympic GBT and a pin trading station. And Deote, the official tech partner. They're running the Olympic fan data platform. They did it back in Paris 24 and scaled to Milan. But do you notice anything peculiar? There isn't a single data platform in the core lineup of partners. Milano Cortina flirts with AI. It's testing the waters, detecting key moments in seconds, Olympic GBT chatbot, drones moving alongside athletes for point of view footage. Deoid's fan data platform. All of this is useful stuff. It's fun to play with, but all of these things are nice to have add-ons on top of the games. LA28 will be drastically different. LA28 doesn't just add data infrastructure to the Olympics. LA28 is the data infrastructure. LA28 is going to be the first Olympics where AI cloud infra and data platforms are foundational sponsor categories. If you listen to the tech stack behind the Olympics, it's giving a typical enterprise SAS feature planning meeting. Google Cloud, Snowflakes AI data cloud, Cisco AI networking, Deoid tech consulting, and T-Mobile's 5G. So the question isn't whether LA28 will collect data. The question is what exactly are they building with it?数据协作新范式：Snowflake 的清洁室与隐私保护
这引出了奥运会历史上最引人注目的赞助商之一：Snowflake。其新闻稿称，Snowflake 将整合运动员训练数据、健康记录和营养信息。但 Snowflake 本身并非健康或健身应用，它是一个数据平台，核心产品允许公司在云端存储、处理和分析海量数据。因此，LA28 和 Snowflake 正在构建的是一个覆盖整个赛事的数据系统。它们将创建一个平台层，使数十个体育组织、美国奥委会、各单项体育联合会、美国奥运代表团、LA28 运营方、赞助商和转播商都能接入同一个数据环境。
体育领域的数据变现正在爆炸式增长：联盟将运动员追踪数据出售给博彩公司，转播商利用其进行 AR 叠加，动态票价设定能大幅提升收入。与 Snowflake 的合作正是加速了这一进程。你可能会问，谁来决定数据的共享方式？数据隐私法如何执行？这些公司又将如何访问那些本应由奥委会控制的数据？答案是：数据清洁室（Data Clean Rooms）。Snowflake 的清洁室解决了体育赞助中的核心摩擦点——它们允许权利持有者和赞助商在不违反任何隐私法律的前提下，寻找受众重叠。
以粉丝数据为例：在金字塔顶端是拥有粉丝数据的权利持有者，在 LA28 的语境下，LA28 就是权利持有者。LA28 收集粉丝数据，如购票信息、App 登录、奥运 ID 偏好、商品购买、关注的运动员和运动类别等。这些数据被匿名化并载入 Snowflake，然后转化为用户画像和细分群体，例如“高价值洛杉矶排球迷”或“休闲越野滑雪爱好者”。这本质上就像在超市扫描会员卡，只不过你被标记为“购买了 50 美元商品的粉丝”。
第二层是赞助商，例如达美航空。达美航空也拥有大量客户数据，但因隐私法限制，无法合法共享个人数据。Snowflake 引入的清洁室概念，允许任何人将匿名化数据上传到 Snowflake，而无需共享个人身份信息。达美航空也可以分享其数据，例如来自西海岸枢纽的常旅客数据。在此基础上，LA28 和达美航空可以测试其受众是否重叠。LA28 输入：“这是购买了排球门票且消费超过 50 美元的粉丝。”达美航空输入：“这是预订飞往西海岸城市的经济舱的常旅客。”清洁室运行查询，结果显示 32% 的高价值奥运粉丝已经是达美航空的客户，并且在参加奥运会后，他们预订航班的可能性是普通人的 2.3 倍。
Original English
Which brings us to the most notable sponsor in Olympic history. Snowflake. The PR says Snowflake will centralize athlete training data, health records, and nutrition. But Snowflake isn't a health or a fitness app. Snowflake is a data platform. Its core product lets companies store, process, and analyze massive amounts of data in the cloud. So what LA28 and Snowflake are building is a data system for the entire games. They're building a layer where dozens of sports organizations, US Olympic Committee, individual sport federations, Team USA, LA28 operations, sponsors, broadcasters can all plug into one data environment. Data monetization in sports is exploding. Leagues sell player tracking to firms who bet on those players. Broadcasters use it for AR overlays. Dynamic pricing on tickets skyrockets the revenue. And the partnership with Snowflake accelerates that. You may ask, but who decides how this data gets shared? What about data privacy laws? How is it that all of these companies are going to access something that only the Olympic Committee gets to control? And the answer is data. Clean rooms. Snowflakes clean rooms solve the core friction in sports sponsorship. Clean rooms allow rights holders and sponsors to look for overlapping audiences without violating any privacy laws. Let's use fan data as an example. At the top of the pyramid, there is a rights holder who owns fan data. In the context of the next summer games, LA28 is the rights holder. LA28 collects data on their fans, ticket sales, app loginins, Olympic ID preferences, merch, athletes they follow, sports categories of choice. This data gets anonymized and loaded into Snowflake and then it gets turned into profiles and segments. For example, a high value LA volleyball fan or a casual crosscountry ski enthusiast. It is basically like scanning your loyalty card at a grocery store, but instead you're being labeled as this fan buys $50 merch. Then there is the second layer underneath the rights holder, the sponsors. For example, Delta Airlines. Delta also has a lot of data on their customers, but they can't legally share personal data because that would violate privacy laws. So what Snowflake is doing is it introduces the concept of clean rooms. and clean rooms let anyone upload anonymized data into Snowflake without sharing personal identifiable information. So Delta can also share their data. For example, frequent flyers from West Coast hubs. And this is where LA28 and Delta test if their audiences overlap. LA28 puts in, "Here are the fans who bought volleyball tickets and spent $50 plus on merch." Delta puts in, "Here are our frequent flyers who book economy to West Coast cities?" The clean room runs a query. How much overlap is there. Do Olympic fans match Delta customers? And the results look something like this. 32% of high-v valueue Olympic fans are already Delta customers and they're 2.3 times more likely to book flights after attending the events.AI 驱动的个性化体验与多维度收入增长
Snowflake 进一步将粉丝划分为不同类别，如“加州高价值排球迷”或“购买高端商品的游泳爱好者”。这些群体成为细分市场（segments），赞助商可以订阅这些细分市场。例如，达美航空愿意支付每月 10,000 美元的费用，以获取这些可供其精准定位的粉丝画像。凭借这些信息，他们可以开展“为 2032 年布里斯班奥运会的所有奥运粉丝提供 20% 的机票折扣”等活动。这种模式的优势在于，他们能够精准定位那些根据数据分析极有可能转化的用户。
在此过程中，LA28 通过每一次交易获利。每当有用户通过此类活动预订机票，LA28 都会从中抽取一定比例的收益。将此乘以订阅这些细分市场的赞助商数量——达美航空、Visa、可口可乐、酒店、租车公司等——所有这些参与者都在交换行为数据。这些洞察被反馈到 AI 模型中，使得下一次预测更加精准。粉丝画像变得更丰富，定位也更精确，原本价值 20 美元的粉丝画像，现在可能因为赞助商愿意为更准确的数据支付更高费用，而价值达到 30、40、甚至 50 美元。
国际奥委会（IOC）已发布《奥运会 AI 议程》，旨在将 AI 嵌入未来每一届奥运会的商业 DNA 中。AI 将被整合到运动员项目、公平竞赛、赛事运营、粉丝互动和治理等各个领域，每个领域都对应着一个潜在的收入网络。与以往不同，例如东京奥运会与巴黎 2024 奥运会相比，IOC 虽掌握粉丝信息，但这些信息并未跨平台共享。粉丝可能收到不感兴趣的体育项目的邮件，赛事结束后则进入长达数月的“寂静期”，而寂静期并不能产生收入。
在 LA28 奥运会中，生成式 AI（GenAI）将被用于以 12 种不同语言、针对不同粉丝群体，以不同的语气、通过不同的渠道和行动号召来定制信息。例如，对于一位痴迷花样滑冰的粉丝，系统会检测到她观看了所有 Kori Sakamoto 的比赛录像，反复观看她的跳跃分解动作，并投入大量时间研究技术分析内容。GenAI 会为她量身定制一条消息：“Sakamoto 在自由滑中获得 147 分，创下自世锦赛以来的赛季最佳。点击查看她的技术分解图。”粉丝的每一次互动，如打开 App、登录、点击特定比赛或技术分析内容，都会被记录为会话事件，并流入奥运粉丝数据平台。该平台会进行行为评分，将该粉丝归类为“表现动机型”（performance motivation cohort），因为她对技术图表的停留时间是情感亮点内容的四倍。当 Sakamoto 的自由滑比赛结束且官方分数公布时，她将收到一条专门为其定制的通知，由 GenAI 根据其群体模板和分数进行个性化撰写。
Original English
Snowflake creates categories of fans, high value volleyball fans in California, swimming enthusiasts who buy premium merch. Each of these groups is a segment, a group of customers with similar behaviors and preferences. and sponsors can subscribe to the segments. Delta Airlines is willing to pay a monthly fee of, let's say, $10,000 just to have access to these profiles that they can target. With this information, they can run campaigns like 20% off flights to Brisbane for the 2032 games for all Olympic fans. And the beauty of this is that they would be targeting users that Snowflake knows are likely to convert because the data says so. While this is happening, LA28 makes money on every transaction. Every time someone books a flight through that campaign, LA28 takes a cut. And now multiply this by the number of sponsors who subscribe. Delta, Visa, Coca-Cola, hotels, car rentals. All of these players exchanging behavioral data. These insights are fed back into AI models, making the next prediction even better. The profiles get richer, the targeting gets sharper, and suddenly a fan profile that was worth $20 is now worth $30, $40, $50 because sponsors are willing to pay more for access to data that is this accurate. The logos that you see in the sponsorship roster redefine what the Olympic Games are. The International Olympic Committee published an Olympic AI agenda, the purpose of which is to embed AI into the commercial DNA of every future games. AI will be embedded in athlete sport, equal access, game operations, fan engagement, and governance. Each of these areas maps to a revenue web. Let me show you how. In previous games, let's say comparing Tokyo to Paris 2024, the Olympic Committee had information about their fans, but that information was not shared across platforms. You, as a fan, could get an email from them about a sport that you would not necessarily care about. And then after the event, they would enter a period of complete silence between the summer and winter games, which is a natural thing, one might say. Sure. Except silence doesn't generate revenue. In LA28, Genai will be used to craft messages in 12 different languages, targeting fans in different tones, through different channels, and with different calls to action. For example, I'm a huge fan of figure skating. In my case, the system would detect that I have watched all of Kori Sakamoto's routines. I replayed her jump breakdowns. I spent a lot of data studying her technical analysis content. For me, Jenna would create a message. Sakamoto just scored 147 points in her freekate, her season's best since worlds. Tap for her technical breakdown chart. Everything I just did adds up like pearls on a necklace. Every time I open the app, I fire a session event. Olympics ID login timestamp device. I tap on figure skating women's freekate. I tap on her program components or jump technique breakdown. Each action streams into the Olympic fan data platform. The platform runs behavioral scoring classifying me as performance motivation cohort because my dwell time is linked to technical charts four times higher than emotional highlights. When Sakamoto's free skate ends and official scores get published, there's going to be a notification that will be sent specifically to me. Genai will compose it through the cohort template and personalize it with the score.监控、数据所有权与技术的常态化
每一次数据点的记录，每一次个性化消息的推送，都为多方创造了价值。首先，推送通知作为“钩子”，极有可能吸引粉丝点击。粉丝在技术分解页面停留 90 秒，该页面可能包含来自顶级赞助商的广告，如“获取日本队周边商品”。如果粉丝购买了 35 美元的帽子，LA28 的电商系统会捕获销售并归因于此次推送。将此乘以数百万粉丝在 17 天内收到的数百条个性化推送，仅个性化引擎带来的商品收入就可能达到数千万美元。其次，粉丝访问的页面会展示“由 Omega Timing 提供”的品牌集成和 Visa 支付小部件。这些展示是基于粉丝画像进行的。例如，一位被归类为“高意向表现型粉丝”，历史上点击体育分析内容，且其奥运账户关联了 Visa 卡。最后，与粉丝相关的每一次查询、每一次触发事件、每一次 GenAI 生成的消息，都运行在 Snowflake 的基础设施上。Snowflake 乐于此合作，因为奥运会产生了巨大的查询量，而它们能够处理这些请求，这相当于向所有观看奥运会的企业首席技术官（CTO）进行了一场长达 17 天的现场演示，远胜于 30 秒的超级碗广告。此外，Google 及其 AI 模型也从中受益。用户搜索某位运动员的最终得分，Google 的 AI 模型会返回来自 Olympics.com 的答案，并结合 YouTube 精彩集锦视频和 NBC 的赞助流媒体提示。Google 捕获搜索数据，投放广告，并将流量归因于奥运合作伙伴关系。而粉丝所获得的，仅仅是一个更相关的通知。他们可能不会意识到，自己在页面上停留的 90 秒，同时为六家不同的公司创造了价值。
为了理解 LA28 将成为什么样子，可以参考巴黎 2024 已经实现的情况。2023 年 5 月，法国议会通过了《奥运会法案》，在实验基础上合法化了算法视频监控。巴黎地区升级了 200 个摄像头，并在 46 个地铁站部署了约 300 个摄像头。LA28 继承了这一模式，但面临着截然不同的政治和法律环境。例如，加州拥有严格的消费者隐私保护法。然而，美国宪法第四修正案仅限制政府行为，而不限制私营公司。LA28 是一个私营组委会，与私营科技赞助商合作，后者并不被禁止部署 AI 监控。
那么，LA28 将收集哪些数据？包括粉丝移动模式（从入口到座位、到餐饮区、到商品店）、热力图和停留时间数据、无现金支付、身份识别（通过生物识别或 Tap-to-Pay）、二维码扫描、增强现实滤镜激活、App 内签到、CCTV 摄像头、无人机、车牌识别器以及通信元数据。
数据所有权问题也浮出水面：粉丝互动数据归国际奥委会和 LA28 所有；安全监控数据由政府机构收集；场馆运营数据由 Cisco、T-Mobile 和 Snowflake 等私营承包商收集；生物识别数据目前尚无明确政策，但它是最具法律争议的数据类别。
巴黎奥运会安全技术的经验将被推广。为巴黎部署 AI 监控系统的四家法国公司，现在拥有了一个全球可见的案例研究，可以向世界各地的城市、体育场馆、运营商和活动组织者销售。生物识别基础设施正在常态化并扩展。47% 的全球体育场馆在 2025 年将优先考虑生物识别技术，它们正在关注 LA28 作为示范点。如果奥运场馆的面部认证能够顺利进行，我们应该会看到 NFL、NBA、MLB 和音乐会等领域迅速采纳。粉丝数据将成为永久性的商业支柱，被整合进用户画像，并为赛后赞助商提供增值服务。一旦 LA28 结束，这些定位策略预计将被用于 2032 年布里斯班奥运会。
每一次奥运会都为下一次奥运会提供了验证。在巴黎测试的监控系统现已部署到法国的交通网络直至 2027 年。为 LA28 构建的数据架构将延续到 2032 年布里斯班奥运会。在奥运会期间常态化的生物识别数据，将迁移到 NFL 体育场、NBA 球馆和音乐厅。收集到的每一个数据点都具有复利价值，因为它能改进 AI 模型，训练下一代系统，并提高大型活动中常规数据收集的基准预期。
Original English
Okay, every data point gets recorded. You get a personalized message. So, who profits from this? Everyone does. Number one, that push notification is the hook. The fan taps it with a high probability based on their profile. I land on the jump to jump comparison and I spend 90 seconds on the technical breakdown page. That page contains an ad from a top sponsor, get team Japan's merch. If the fan converts on the $35 cap, the LA28 e-commerce system captures the sale and attributes it to the push notification. Multiply that across millions of fans receiving hundreds of personalized pushes over 17 days and the merch revenue from the personalization engine alone could turn into tens of millions. Number two, the page that I land on has a powered by Omega timing brand integration and a Visa checkout widget. Those placements are random. They're shown to me because the platform knows my fans profile. Let's say I'm classified as a high intent performance fan who historically clicks on sports analytics content and in addition to that has a Visa card linked to my Olympics account. Number three, every query associated with the fan, every trigger that fires, every Janaya call that composes a message, all of it runs on Snowflake's infrastructure. Snowflake wants this partnership because the Olympics generate massive query volume that Snowflake is able to handle. And the fact that they're able to handle it is a live demo to every enterprise chief technology officer watching the Olympics becomes a Super Bowl ad for Snowflake. except instead of a 30- secondond slot, it's a 17day live demo processing billions of events. And finally, Google and its AI mode. I might simply Google Sakamoto's final score and Google's AI mode will return an answer from Olympics.com. And that would be combined with a YouTube highlight video and a sponsored NBC streaming prompt. Google captures the search data, serves ads against it, and attributes the traffic back to the Olympic partnership. And me as a fan, all I get is a better notification. It's just an alert that is in fact relevant to me. And I don't think about the fact that my 90 seconds on that page generate value for six different companies simultaneously. This is brilliant and beautiful from the business standpoint. To understand what LA28 will become, think about what Paris 24 already normalized. In May 2023, the French Parliament passed the Olympic Games Law, which legalized algorithmic video surveillance on an experimental basis. 200 cameras in the Paris area were upgraded with AI, plus around 300 cameras across 46 metro stations. LA28 inherits this pattern because it was done before. But now put this in perspective of a fundamentally different political and legal environment. California, for example, has very strong consumer privacy protections. But on the other hand, the fourth amendment of the United States only restricts government but not private companies. And LA28 is a private organizing committee working with private tech sponsors who are not forbidden from deploying AI surveillance. So what data will be collected at LE28? Fan movement patterns, entry to seat, seat to concession, concession to merge store. All of this boils down to heat maps and dwelltime data, cashierless checkouts, links, biometric or tap to pay identity to purchases, QR scans, augmented reality filter activations, inapp checkins, CCTV cameras, drones, license plate readers, and communications metadata. Now to the interesting part. Who owns this data? Fan engagement is owned by the International Olympic Committee and LA28. Security surveillance data will be collected by government agencies. Venue operations data is collected by private contractors, Cisco, T-Mobile, and Snowflake. Biometric data does not have a specific policy yet, but it is the most legally contested category. And what happens after security tech goes to market? The four French companies that built Paris's AI surveillance system now have a live, globally visible case study to sell to every city, every stadium, every operator, every event organizer on Earth because it was tested at the Olympics. Biometric infrastructure normalizes and expands. 47% of global stadiums prioritizing biometrics in 2025 are watching LA28 as a proof point. If facial authentication at Olympic venues works smoothly, we should be expecting rapid adoption across NFL, NBA, MLB, and concerts. Fan data becomes the permanent commercial backbone. It will be folded into profiles. It will be enriching them for postgame sponsors as well. And the moment LA28 is done, expect that targeting to be used for Brisbane 2032. So where does this leave us? Each Olympics serves as the validation for the next. The surveillance systems tested in Paris are now deployed across French transit until 2027. The data architecture built for LA28 will carry forward to Brisbane 2032. The biometric data normalized at the Olympics will migrate to NFL stadiums, NBA arenas, and concert halls. Every data point collected has compounding value because it improves AI models, trains the next generation of systems, and increases the baseline expectations for what normal data collection looks like at large scale events. We hope this was helpful. Let us know what you think in the comments. We'll see you next time. Bye.📌 文中提及的人物和组织
公司/组织: ATS, Google, Snowflake, Cisco, Deoitte, International Olympic Committee, LA28
产品/模型: Snowflake clean rooms