Google (GOOG) 2024 Q2 法說會逐字稿

Welcome, everyone. Thank you for standing by for the Alphabet second quarter 2024 earnings conference call. (Operator Instructions)

歡迎大家。感謝您參加 Alphabet 2024 年第二季財報電話會議。 （操作員指示）

I would now like to hand the conference over to your speaker today, Jim Friedland, Director of Investor Relations. Please go ahead.

現在，我想將會議交給今天的發言人、投資者關係總監吉姆‧弗里德蘭 (Jim Friedland)。請繼續。

Thank you. Good afternoon, everyone, and welcome to Alphabet's second quarter 2024 earnings conference call. With us today are Sundar Pichai, Philipp Schindler and Ruth Porat. Now I'll quickly cover the safe harbor. Some of the statements that we make today regarding our business, operations and financial performance may be considered forward-looking. Such statements are based on current expectations and assumptions that are subject to a number of risks and uncertainties. Actual results could differ materially. Please refer to our forms 10-K and 10-Q, including the risk factors. We undertake no obligation to update any forward-looking statement.

謝謝。大家下午好，歡迎參加 Alphabet 2024 年第二季財報電話會議。今天和我們在一起的有 Sundar Pichai、Philipp Schindler 和 Ruth Porat。現在我將快速介紹安全港。我們今天就我們的業務、營運和財務表現所做的一些聲明可能被視為前瞻性的。此類聲明基於當前的預期和假設，受多種風險和不確定性的影響。實際結果可能存在重大差異。請參閱我們的 10-K 和 10-Q 表格，其中包括風險因素。我們不承擔更新任何前瞻性聲明的義務。

During this call, we will present both GAAP and non-GAAP financial measures. A reconciliation of non-GAAP to GAAP measures is included in today's earnings press release, which is distributed and available to the public through our Investor Relations website located at abc.xyz/investor. Our comments will be on year-over-year comparisons unless we state otherwise.

在本次電話會議中，我們將介紹 GAAP 和非 GAAP 財務指標。今天的收益新聞稿中包含了非 GAAP 與 GAAP 指標的對賬，該新聞稿透過我們的投資者關係網站 abc.xyz/investor 發布並向公眾開放。除非另有說明，否則我們的評論將基於同比比較。

And now I'll turn the call over to Sundar.

現在我將把電話轉給 Sundar。

Thank you, Jim, and hello, everyone. I'm really pleased with our results this quarter. They showed tremendous ongoing momentum in Search and great progress in Cloud with our AI initiatives driving new growth. Search had another excellent quarter. And in terms of product innovation, we are seeing great progress with AI Overviews. In Q2, cloud reached some major milestones. Quarterly revenues crossed the $10 billion mark for the first time at the same time passed the $1 billion mark in quarterly operating profit.

謝謝你，吉姆，大家好。我對本季的業績非常滿意。他們在搜尋領域表現出強勁的持續發展勢頭，在雲端運算領域取得了巨大進步，而我們的人工智慧計畫則推動了新的成長。搜尋業務又度過了一個出色的季度。在產品創新方面，我們看到 AI Overviews 取得了巨大進步。在第二季度，雲端運算達到了一些重要的里程碑。季度營收首度突破100億美元大關，同時季營業利潤突破10億美元大關。

Year-to-date, our AI infrastructure and generative AI solutions for cloud customers have already generated billions in revenues and are being used by more than 2 million developers. As I spoke about last quarter, we are uniquely well positioned for the AI opportunity ahead. Our research and infrastructure leadership means we can pursue an in-house strategy that enables our product teams to move quickly. Combined with our model building expertise, we are in a strong position to control our destiny as the technology continues to evolve.

今年迄今為止，我們為雲端客戶提供的人工智慧基礎設施和生成式人工智慧解決方案已經創造了數十億美元的收入，並被超過 200 萬名開發人員使用。正如我上個季度所說的那樣，我們在未來的人工智慧機會方面具有獨特的優勢。我們在研究和基礎設施方面的領先地位意味著我們可以推行內部策略，使我們的產品團隊能夠快速行動。結合我們的模型建構專業知識，隨著技術的不斷發展，我們有能力掌控自己的命運。

Importantly, we are innovating at every layer of the AI stack from chips to agents and beyond, a huge strength. We are committed to this leadership long term. This was underscored by the announcements we made at I/O, Cloud Next and Google Marketing Live, and we'll touch on many of them here. Today, I'll start with Search, then move to our AI momentum more generally followed by cloud, YouTube and some closing thoughts.

重要的是，我們正在從晶片到代理等人工智慧堆疊的每一層進行創新，這是一個巨大的優勢。我們致力於長期保持這一領導地位。我們在 I/O、Cloud Next 和 Google Marketing Live 上發布的公告強調了這一點，我們將在這裡談論及其中的許多內容。今天，我將從搜尋開始，然後更廣泛地討論我們的人工智慧發展勢頭，接著是雲端、YouTube 和一些結束語。

Let's dive in. Over the past 25 years, we have continued to reimagine and expand Google Search across many technological shifts. With AI, we are delivering better responses on more types of search queries and introducing new ways to search. We are pleased to see the positive trends from our testing continue as we roll out AI Overviews, including increases in search usage and increased user satisfaction with the results.

讓我們深入了解一下。在過去的 25 年裡，我們在許多技術變革中不斷重新構想並擴展 Google 搜尋。借助人工智慧，我們可以對更多類型的搜尋查詢提供更好的回應，並引入新的搜尋方式。我們很高興看到，隨著我們推出 AI 概覽，測試中的正面趨勢仍在繼續，包括搜尋使用量的增加和用戶對結果的滿意度的提高。

People who are looking for help with complex topics are engaging more and keep coming back for AI Overviews. And we see even higher engagement from younger users aged 18 to 24 when they use Search with AI Overviews. As we have said, we are continuing to prioritize approaches that send traffic to sites across the web and we are seeing that ads appearing either above or below AI Overviews continue to provide valuable options for people to take action and connect with businesses.

那些尋求複雜主題幫助的人們參與度更高，並不斷回來尋求 AI 概述。當 18 至 24 歲的年輕用戶使用 AI 概覽搜尋時，我們發現他們的參與度更高。正如我們所說，我們將繼續優先考慮將流量發送到網路上的網站的方法，並且我們看到出現在 AI 概覽上方或下方的廣告繼續為人們採取行動並與企業建立聯繫提供有價值的選擇。

Beyond AI Overviews, AI expands the types of queries we are able to address and opens up powerful new ways to search. Visual search via Lens is one. Soon, you'll be able to ask questions by taking a video with Lens. And already we've seen that AI Overviews and Lens leads to an overall visual search usage.

除了人工智慧概述之外，人工智慧還擴展了我們能夠解決的查詢類型，並開啟了強大的新搜尋方式。透過 Lens 進行視覺搜尋就是其中之一。很快，您就可以透過使用 Lens 拍攝影片來提出問題。我們已經看到 AI Overviews 和 Lens 帶來了整體視覺搜尋的使用。

Another example is Circle to Search, which is available today on more than 100 million Android devices. We are seeing tremendous momentum from our AI investments. More than 1.5 million developers are now using Gemini across our developer tools, and we recently unveiled new models that are more capable and efficient than ever.

另一個例子是 Circle to Search，目前已在超過 1 億台 Android 裝置上提供。我們看到人工智慧投資正呈現巨大動能。目前，已有超過 150 萬名開發人員在我們的開發工具中使用 Gemini，我們最近推出了比以往更強大、更有效率的新模型。

Gemini now comes in four sizes with each model designed for its own set of use cases. It's a versatile model family that runs efficiently on everything from data centers to devices. At 2 million tokens, we offer the longest context window of any large-scale foundation model to date, which powers developer use cases that no other model can handle. Gemini is making Google's own products better. All six of our products with more than 2 billion monthly users now use Gemini. This means that Google is the company that's truly bringing AI to everyone. Gemini is powering incredibly helpful features in Search, Workspace, Google Messages and more.

Gemini 目前有四種尺寸，每種型號都針對自己的用例而設計。它是一個多功能模型系列，可以在從資料中心到設備的所有設備上有效運作。我們擁有 200 萬個令牌，提供了迄今為止所有大型基礎模型中最長的上下文窗口，為其他模型無法處理的開發人員用例提供支援。Gemini 正在讓 Google 自己的產品變得更好。我們所有六款產品每月用戶超過 20 億，現在都使用 Gemini。這意味著谷歌是一家真正將人工智慧帶給每個人的公司。Gemini 為搜尋、工作區、Google 訊息等提供了極為有用的功能。

At our I/O, we showed new features coming soon to Gmail and to Google Photos. Soon, you'll be able to ask for those questions like what did I eat at that restaurant in Paris last year. For a glimpse of the future, I hope you saw project Astra at I/O. It shows multimodal understanding and natural conversational capabilities. We have always wanted to build a universal agent, and it's an early look at how they can be helpful in daily life.

在我們的 I/O 大會上，我們展示了即將在 Gmail 和 Google Photos 上推出的新功能。很快，您就可以詢問諸如“去年我在巴黎那家餐廳吃了什麼”之類的問題。為了一窺未來，我希望您在 I/O 上看到了 Astra 專案。它表現出多模式理解和自然對話能力。我們一直想打造一個通用代理，這是對它們如何在日常生活中提供幫助的初步觀察。

Our AI product advances come from our long standing foundation of research leadership as well as our global network of infrastructure. In Q2, we announced our first data center and cloud region in Malaysia and expansion projects in Iowa, Virginia and Ohio. Our TPUs are a key bet here too. Trillium is the sixth generation of our custom AI accelerator, and it's our best performing and most energy efficient TPU today.

我們的人工智慧產品的進步源於我們長期以來的研究領導基礎以及我們的全球基礎設施網路。在第二季度，我們宣布了在馬來西亞的第一個資料中心和雲端區域以及在愛荷華州、維吉尼亞州和俄亥俄州的擴展項目。我們的 TPU 在這裡也是一個關鍵賭注。Trillium 是我們客製化 AI 加速器的第六代，也是我們目前表現最好、能源效率最高的 TPU。

It achieves a near 5x increase in peak compute performance per chip and a 67% more energy efficient compared to TPU v5e. And the latest NVIDIA Blackwell platform will be coming to Google Cloud in early 2025. We continue to invest in designing and building robust and efficient infrastructure to support our efforts in AI, given the many opportunities we see ahead.

與 TPU v5e 相比，它使每塊晶片的峰值計算性能提高了近 5 倍，能源效率提高了 67%。最新的 NVIDIA Blackwell 平台將於 2025 年初登陸 Google Cloud。鑑於我們看到的眾多機遇，我們將繼續投資設計和建立強大而高效的基礎設施，以支持我們在人工智慧領域的努力。

Of course, as we do this, we'll continue create capacity by allocating resources towards our highest priorities. We are relentlessly driving efficiencies in our AI models. For example, over the past quarter, we have made quality improvements that include doubling the core model size for AI Overviews, while at the same time improving latency and keeping cost per AI Overview sell flat. And we are focused on matching the right model size to the complexity of the query in order to minimize impact on costs and latency. Separately on our real estate investments, we are taking a measured approach to match the current and future needs of our hybrid workforce as well as our communities.

當然，當我們這樣做時，我們將繼續透過將資源分配給我們的最高優先事項來創造能力。我們正在不懈地提高人工智慧模型的效率。例如，在過去的一個季度中，我們進行了品質改進，包括將 AI 概覽的核心模型大小增加一倍，同時改善延遲並保持每個 AI 概覽銷售的成本持平。我們專注於將正確的模型大小與查詢的複雜性相匹配，以最大限度地減少對成本和延遲的影響。另外，在我們的房地產投資方面，我們正在採取審慎的方法來滿足我們的混合勞動力以及社區的當前和未來的需求。

Next, Google Cloud. We continue to see strong customer interest, winning leading brands like Hitachi, Motorola Mobility and KPMG. Our deep partnership with Oracle significantly expanded our joint offerings to their large customer base. Our momentum begins with our AI infrastructure, which provides AI startups like Essential AI with leading cost performance for models and high-performance computing applications. We continue to drive fundamental differentiation with new advances since Cloud Next. This includes Trillium, which I mentioned earlier, and A3 Mega powered by NVIDIA H1100 GPUs, which doubles the networking bandwidth of A3.

接下來是 Google Cloud。我們繼續看到客戶的強烈興趣，贏得了日立、摩托羅拉移動和畢馬威等領先品牌。我們與 Oracle 的深度合作大大擴展了我們向其龐大客戶群提供的聯合產品。我們的動力始於我們的人工智慧基礎設施，它為 Essential AI 等人工智慧新創公司提供領先的模型和高效能運算應用性價比。自 Cloud Next 以來，我們繼續透過新的進步推動根本性的差異化。其中包括我之前提到的 Trillium，以及搭載 NVIDIA H1100 GPU 的 A3 Mega，後者的網路頻寬是 A3 的兩倍。

Our enterprise AI platform, Vertex, helps customers such as Deutsche Bank, Kingfisher, and the US Air Force build powerful AI agents. Last month, we announced a number of new advances. Uber and WPP are using Gemini Pro 1.5 and Gemini Flash 1.5 in areas like customer experience and marketing.

我們的企業 AI 平台 Vertex 幫助德意志銀行、翠鳥和美國空軍等客戶建立強大的 AI 代理。上個月，我們宣布了多項新進展。Uber 和 WPP 正在客戶體驗和行銷等領域使用 Gemini Pro 1.5 和 Gemini Flash 1.5。

We broadened support for third-party models, including Anthropic's Claude 3.5 Sonnet and Open Source models like Gemma 2, Llama and Mistral. We are the only cloud provider to offer grounding with Google search, and we are expanding grounding capabilities with Moody's, MSCI, ZoomInfo and more. Our AI powered applications portfolio is helping us win new customers and drive upsell.

我們擴大了對第三方模型的支持，包括 Anthropic 的 Claude 3.5 Sonnet 和 Gemma 2、Llama 和 Mistral 等開源模型。我們是唯一一家提供 Google 搜尋接地的雲端供應商，我們正在與 Moody's、MSCI、ZoomInfo 等公司擴展接地功能。我們的人工智慧應用程式組合正在幫助我們贏得新客戶並推動追加銷售。

For example, our conversational AI platform is helping customers like Best Buy and Gordon Food Service. Gemini for Workspace helps Click Therapeutics analyze patient feedback as they build targeted digital treatments. Our AI-powered agents are also helping customers develop better quality software, fine insights from their data and protect their organization against cybersecurity threats using Gemini.

例如，我們的對話式人工智慧平台正在幫助百思買和戈登食品服務等客戶。Gemini for Workspace 幫助 Click Therapeutics 在製定有針對性的數位治療方案時分析病患回饋。我們的人工智慧代理商還幫助客戶開發品質更好的軟體，從他們的數據中獲得精細的洞察力，並使用 Gemini 保護他們的組織免受網路安全威脅。

Software engineers at Wipro are using Gemini Code Assist to develop and document software faster. And data analysts at Mercado Libre are using BigQuery and Looker to optimize capacity planning and fulfill shipments faster.

Wipro 的軟體工程師正在使用 Gemini Code Assist 來更快地開發和記錄軟體。Mercado Libre 的數據分析師正在使用 BigQuery 和 Looker 來優化容量規劃並更快地完成出貨。

Cyber security risks continue to accelerate, and the number of breaches continue to grow. Something we all see in the news every day and that our Mandiant teams help manage. Our strong track record of uptime, quality control and reliability make Google Cloud the trusted security choice for organizations like Fiserv and Marriott International. In Q2, we infused AI throughout our security portfolio, helping TELUS strengthened its proactive security posture.

網路安全風險持續加劇，違規行為持續增加。我們每天都會在新聞中看到一些事情，而我們的 Mandiant 團隊會幫忙管理這些事情。我們在正常運作時間、品質控制和可靠性方面的良好記錄使 Google Cloud 成為 Fiserv 和萬豪國際等組織值得信賴的安全選擇。在第二季度，我們在整個安全產品組合中融入了人工智慧，幫助 TELUS 加強了其主動安全態勢。

Turning next to YouTube. YouTube is focused on a clear strategy, connecting creators with a massive audience and enabling them to build successful businesses through ads and subscriptions while helping advertisers reach their desired audience. We had a great Brandcast this quarter and Philipp will say more. I'm pleased with the progress there. YouTube has remained number one in US streaming watch time according to Nielsen. Views of YouTube Shorts and Connected TVs more than doubled last year, and we are making it easier for creators to add captions and turn regular videos into Shorts.

接下來轉到 YouTube。YouTube 專注於明確的策略，將創作者與大量受眾聯繫起來，並使他們能夠透過廣告和訂閱建立成功的業務，同時幫助廣告商接觸到他們想要的受眾。本季我們舉辦了一場精彩的品牌宣傳活動，Philipp 將會分享更多內容。我對那裡的進展感到滿意。根據尼爾森的數據，YouTube 在美國串流觀看時間上仍然位居第一。去年，YouTube Shorts 和連網電視的觀看次數增加了一倍多，我們正在讓創作者更輕鬆地添加字幕並將普通影片轉換為 Shorts。

Next on Android and Pixel. We joined Samsung for their Galaxy Unpacked event a few weeks ago, and we shared that Samsung's new devices will include the latest AI-powered Google updates on Android. It was a great event. I'm looking forward to our Made by Google event happening in August. We'll have lots to share around Android and the Pixel portfolio of devices. Our pixel line is doing well. We recently introduced a new Pixel 8a powered by our latest Google Tensor G3 chip. It provides beautiful AI experiences like Circle to Search, Best Take and a Gemini-powered AI assistant.

接下來是 Android 和 Pixel。幾週前，我們參加了三星的 Galaxy Unpacked 活動，並透露三星的新設備將包含 Android 上最新的人工智慧谷歌更新。這是一次偉大的活動。我期待著八月舉行的“谷歌製造”活動。我們將分享很多關於 Android 和 Pixel 裝置組合的資訊。我們的像素線表現良好。我們最近推出了搭載最新 Google Tensor G3 晶片的新款 Pixel 8a。它提供了美妙的人工智慧體驗，如「圓圈搜尋」、「最佳選擇」和由 Gemini 驅動的人工智慧助理。

In Other Bets, I'm really pleased with the progress Waymo's making. A real leader in the space and getting rave reviews from users. Waymo served more than 2 million trips to date and driven more than 20 million fully autonomous miles on public roads. Waymo's now delivering well over 50,000 weekly paid public rides primarily in San Francisco and Phoenix. And in June, we removed the waitlist in San Francisco so anyone can take a ride. Fully autonomous testing is underway in other Bay Area locations without a human in the driver's seat.

在其他賭注方面，我對 Waymo 的進展感到非常高興。該領域真正的領導者，並獲得用戶的熱烈評價。迄今為止，Waymo 已完成 200 多萬次出行，並在公共道路上完全自動駕駛了 2000 多萬英里。Waymo 目前每週提供超過 50,000 次付費大眾運輸服務，主要在舊金山和鳳凰城。今年 6 月，我們取消了舊金山的候補名單，以便任何人都可以搭乘。灣區其他地區正在進行無人駕駛的全自動測試。

Before I close, I want to acknowledge that today is Ruth's final earnings call. Let me take a moment to thank her for all she has done for Google and Alphabet as our longest-serving CFO. I'm excited to continue to work with her in her new role. And I look forward to welcoming our newly appointed CFO, Anat Ashkenazi. She starts next week, and you'll hear from her on our call next quarter. Thanks, as always, to our employees and partners everywhere for a great Q2.

在結束之前，我想承認今天是露絲的最後一次財報電話會議。請允許我花點時間感謝她作為我們任職時間最長的財務長為谷歌和 Alphabet 所做的一切。我很高興能繼續與她在新職位上合作。我期待著歡迎我們新任命的財務長 Anat Ashkenazi。她將於下週開始工作，您將在下個季度的電話會議中聽到她的消息。一如既往地感謝我們各地的員工和合作夥伴，讓我們度過了一個美好的第二季。

With that, over to you, Philip.

現在，就交給你了，菲利普。

Thanks, Sundar, and hello, everyone. Starting with performance, Google Services delivered revenues of $73.9 billion for the quarter, up 12% year-on-year. Search and other revenues grew 40% year-on-year, led by growth in the retail vertical followed by financial services. YouTube ads revenues were up 13% year-on-year driven by growth in brand as well as direct response. Network revenues declined 5% year-on-year. In subscriptions, platforms and devices, year-on-year revenues increased 14%, driven again by strong growth in YouTube subscriptions. For the rest of my remarks, I want to double-click on two topics: first, how we are applying AI across the marketing process to deliver an even stronger ads experience; second, YouTube's position as the leading multi-format platform.

謝謝，Sundar，大家好。從業績來看，Google服務本季實現營收739億美元，年增12%。搜尋和其他收入年增 40%，其中零售垂直領域的成長最快，其次是金融服務。YouTube 廣告收入較去年同期成長 13%，這得益於品牌成長和直接反應。網路收入較去年同期下降5%。在訂閱、平台和裝置方面，營收年增 14%，這再次得益於 YouTube 訂閱量的強勁成長。在接下來的發言中，我想重點談談兩個主題：首先，我們如何在整個行銷過程中應用人工智慧來提供更強大的廣告體驗；第二，YouTube 作為領先的多格式平台的地位。

So let me start by sharing some of the ways we're applying AI to bring more performance benefits to even more businesses. Q2 brought several major opportunities to meet and learn from users, developers, creators and customers. From I/O to Brandcast, Google Marketing Live and Cannes, a growing number of our customers and partners are looking to understand how to successfully incorporate AI into their businesses.

因此，首先讓我分享一些我們應用人工智慧為更多企業帶來更多績效優勢的方法。第二季帶來了幾次與使用者、開發人員、創作者和客戶會面和學習的重要機會。從 I/O 到 Brandcast、Google Marketing Live 和坎城，越來越多的客戶和合作夥伴正在尋求了解如何將 AI 成功地融入他們的業務中。

This quarter, we announced over 30 new ads features and products to help advertisers leverage AI and keep pace with the evolving expectations of customers and users. Across Search, PMX, demand gen and retail, we are plying AI to streamline workflows, enhance creative asset production and provide more engaging experiences for consumers.

本季度，我們宣布了 30 多種新的廣告功能和產品，以幫助廣告商利用人工智慧並跟上客戶和用戶不斷變化的期望。在搜尋、PMX、需求生成和零售領域，我們正在利用人工智慧來簡化工作流程、增強創意資產生產並為消費者提供更具吸引力的體驗。

Listening to our customers, retailers, in particular, have welcomed AI-powered features to help scale the depth and breadth of their assets. For example, as part of the new and easier to use Merchant Center, we've expanded product studio with tools that bring the power of Google AI to every business owner. You can upload a product image from the AI with something like feature this product with Paris skyline in the background and product studio will generate campaign ready assets.

傾聽客戶的意見，尤其是零售商，他們歡迎人工智慧功能來幫助擴大其資產的深度和廣度。例如，作為全新且更易於使用的 Merchant Center 的一部分，我們擴展了產品工作室，並提供了可將 Google AI 的強大功能帶給每位企業主的工具。您可以從 AI 上傳產品圖片，例如以巴黎天際線為背景展示產品，產品工作室將產生可用於活動的資產。

I also hear great feedback from our customers on many of our other new AI-powered features. We're beta testing virtual try-on and shopping ads and plan to roll it on YT later this year. Feedback shows this feature gets 60% more high-quality use than other images and a higher click out to retailer sites. Retailers love it because it drives purchasing decisions and fewer returns. Our AI-driven profit optimization tools have been expanded to Performance Max and standard shopping campaigns. Advertisers use profit optimization and smart bidding, see a 15% uplift in profit on average compared to revenue-only bidding.

我還聽到了客戶對我們許多其他新的人工智慧功能的正面回饋。我們正在對虛擬試穿和購物廣告進行 Beta 測試，並計劃於今年稍後在 YT 上推出。回饋顯示，此功能比其他圖像的高品質使用率高出 60%，零售商網站的點擊率也更高。零售商喜歡它，因為它可以推動購買決策並減少退貨。我們的人工智慧驅動的利潤優化工具已擴展到 Performance Max 和標準購物廣告系列。廣告主使用利潤優化和智慧競價，與僅基於收入的競價相比，利潤平均可提高 15%。

Lastly, demand gen is rolling out to display Video 360 and Search Ads 360 in the coming months. With new generative image tools that create stunning, high-quality image assets for social marketers. As we said at GML, when paired with Search or PMX, demand gen delivers an average of 14% more conversions. The use cases we're seeing across the industry show the incredible potential of these AI-enabled products to improve performance.

最後，需求產生將在未來幾個月內推出影片 360 和搜尋廣告 360 展示。利用新的生成圖像工具，為社交行銷人員創建令人驚嘆的高品質圖像資產。正如我們在 GML 所說的那樣，當與搜尋或 PMX 配對時，需求產生平均可帶來 14% 以上的轉換率。我們在整個行業中看到的用例顯示了這些支援人工智慧的產品在提高效能方面的巨大潛力。

Let me briefly share two examples with you. Luxury jewelry retailer, Tiffany, leveraged demand gen during the holiday season and saw a 2.5% brand lift in consideration and actions such as adding items to carts and booking appointments. The campaign drove a 5.6 times more efficient cost per click compared to social media benchmarks. Our own Google marketing team used demand gen to create nearly 4,500 ad variations for Pixel 8 campaign shown across YouTube, Discover and Gmail, delivering twice the clicks per rate at nearly a quarter of the cost.

我先簡單跟大家分享兩個例子。奢侈珠寶零售商蒂芙尼 (Tiffany) 在假日季利用了需求生成，在考慮度以及將商品添加到購物車和預約等操作方面實現了 2.5% 的品牌提升。與社群媒體基準相比，此次活動的每次點擊成本效率提高了 5.6 倍。我們自己的 Google 行銷團隊使用需求生成為 Pixel 8 廣告系列創建了近 4,500 種廣告變體，這些廣告變體在 YouTube、Discover 和 Gmail 上展示，以近四分之一的成本實現了兩倍的點擊量。

In addition to strengthening our ads products for customers, we continue to evolve our existing systems and products with improved models delivering further performance gains. In just six months, AI-driven improvements to quality, relevance and language understanding have improved broad match's performance by 10% for advertisers using smart bidding.

除了為客戶加強我們的廣告產品之外，我們還繼續改進現有的系統和產品，透過改進的模型來進一步提高效能。在短短六個月內，人工智慧推動的品質、相關性和語言理解方面的改進使使用智慧競價的廣告商的廣泛匹配效果提高了 10%。

Also, advertisers who adopt P Max to Broad Match and Smart Bidding in their search campaigns, see an average increase of over 25% more conversions or value at a similar cost. We'll continue to listen to our customers and use their feedback to drive innovation across our products. As you can hear, I continue to be excited about the AI era for ads.

此外，在搜尋廣告活動中採用 P Max 廣泛匹配和智慧出價的廣告客戶，在成本相似的情況下，轉換次數或價值平均增加 25% 以上。我們將繼續傾聽客戶的意見並利用他們的回饋來推動我們產品的創新。正如你所聽到的，我繼續對廣告的人工智慧時代感到興奮。

Now let's turn to YouTube. I've talked before about our approach to making YouTube the best place to create, watch and monetize. First, the best place to create. What sets YouTube apart from every other platform are the creators and the connection they have with their fans. Audience is tuning in to watch their favorite creators continue to grow. For example, two weeks ago, Mr. Beast's Channel hit more than 300 million subscribers.

現在讓我們轉到 YouTube。我之前曾談到我們如何讓 YouTube 成為創作、觀看和獲利的最佳場所。第一，最佳創作地點。YouTube 與其他平台的不同之處在於其創作者以及他們與粉絲之間的聯繫。觀眾正在觀看他們最喜歡的創作者的不斷成長。例如，兩週前，Mr. Beast 頻道的訂閱用戶數突破 3 億。

Next, the best place to watch. Our long-term investment in CTV continues to deliver. Views on CTV have increased more than 130% in the last three years. According to Nielsen, YouTube is the number one most watched streaming platform on TV screens in the US for the 17th consecutive month. Zooming out when you look not just at streaming, but at all media companies and their combined TV viewership, YouTube is the second most watched after Disney. And this growth is happening in multiple verticals, including sports, which has seen CTV watch time on YouTube grow 30% year-over-year.

接下來是最佳觀賞地點。我們對 CTV 的長期投資持續取得成果。在過去三年裡，CTV 的收視率增加了 130% 以上。根據尼爾森的調查，YouTube 連續 17 個月成為美國電視螢幕上觀看次數最多的串流平台。如果你不僅關注串流媒體，還關注所有媒體公司及其綜合電視收視率，你會發現 YouTube 的收視率僅次於迪士尼，為第二大收視率最高的平台。這種成長發生在多個垂直領域，包括運動領域，YouTube 上的 CTV 觀看時間比去年同期成長了 30%。

Lastly, the best place to monetize. CTV on YouTube is continuing to benefit from a combination of strong watch time growth, viewer and advertiser innovation and a shift in brand advertising budgets from linear TV to YouTube. Our largest advertisers across verticals, including retail, entertainment, telco and home and personal care are partnering with creators on ads and organic integrations. Verizon, for example, worked with a YouTube creator and Verizon customer to show the many ways that plans and offerings can be customized to fit people's lives.

最後，最好的貨幣化地點。YouTube 上的 CTV 繼續受益於觀看時間的強勁增長、觀眾和廣告商的創新以及品牌廣告預算從線性電視轉向 YouTube 等因素。我們最大的廣告商遍布零售、娛樂、電信以及家庭和個人護理等垂直領域，正在與創作者合作進行廣告和有機整合。例如，Verizon 與 YouTube 創作者和 Verizon 客戶合作，展示了可以自訂計劃和服務以適應人們生活的多種方式。

Using AI-powered formats, they created sketches in multiple lengths and orientations to serve the right creative to the right viewer and drive people to their site. Verizon's creator ads had a 15% lower CPA and a 38% higher conversion rate versus other ads.

他們使用人工智慧格式創建了多種長度和方向的草圖，為合適的觀眾提供合適的創意，並吸引人們造訪他們的網站。與其他廣告相比，Verizon 的創作者廣告的每次轉換費用 (CPA) 低 15%，轉換率高 38%。

Turning to Shorts. Last quarter, I shared that in the US, the monetization rate of YouTube Shorts relative to in-stream viewing is showing a healthy rate of growth. Again, this quarter, we continue to see an improvement in Shorts monetization, particularly in the US. We're also seeing a very encouraging contribution from brand advertising on Shorts, which we launched on the product in Q4 last year.

轉向短褲。上個季度，我曾分享過，在美國，YouTube Shorts 相對於串流觀看的貨幣化率呈現出健康的成長率。同樣，本季我們繼續看到 Shorts 貨幣化的改善，尤其是在美國。我們也看到 Shorts 上的品牌廣告做出了非常令人鼓舞的貢獻，我們在去年第四季在該產品上推出了 Shorts 品牌廣告。

Lastly, a few words on shopping. Last year, viewers watched 30 billion hours of shopping-related videos, and we saw a 25% increase in watch time for videos that help people shop. While it's early days, shopping remains a key area of investment. At GML, we rolled out several product updates to YouTube shopping, helping creators sell products to their viewers. These updates included product tagging where creators can tag products in their videos for viewers to discover and purchase, product collections and a new affiliate hub, a one-stop shop for creators to find deals and promotional offers from brands and track their affiliate earnings.

最後，再說幾句關於購物的事。去年，觀眾觀看了 300 億小時的購物相關視頻，我們發現幫助人們購物的影片的觀看時間增加了 25%。儘管尚處於早期階段，但購物仍然是一個重要的投資領域。在 GML，我們為 YouTube 購物推出了多項產品更新，幫助創作者向觀眾銷售產品。這些更新包括產品標記，創作者可以在影片中標記產品以供觀眾發現和購買，產品集合和新的聯盟中心，創作者可以一站式找到品牌的交易和促銷優惠並追蹤他們的聯盟收入。

With that, I'll finish by saying a huge thank you to Googlers everywhere for their extraordinary commitment and to our customers and partners for their continued collaboration and trust. And Ruth, thanks for your amazing leadership and partnership over all these years.

最後，我要向世界各地的Google員工表達衷心的感謝，感謝他們的非凡奉獻，也感謝我們的客戶和合作夥伴的持續合作和信任。露絲，感謝您這些年來出色的領導和合作。

Now for one last time, it's over to you.

現在最後一次，一切都交給你了。

Thank you, Philipp, and thanks, Sundar, for those kind words. We had another strong quarter, driven in particular by performance in Search and Cloud as well as the ongoing efforts to durably reengineer our cost base. My comments will be on year-over-year comparisons for the second quarter unless I state otherwise. I will start with results at the Alphabet level followed by segment results and conclude with our outlook.

謝謝菲利普和桑達爾的善意言辭。我們又度過了一個強勁的季度，這主要得益於搜尋和雲端運算領域的表現以及持續不斷的成本基礎重塑努力。除非我另有說明，否則我的評論將基於第二季的年比比較。我將從字母層級的結果開始，然後是細分結果，最後是我們的展望。

For the second quarter, our consolidated revenues were $84.7 billion, up 14% or up 15% in constant currency. Search remained the largest contributor to revenue growth. In terms of expenses, total cost of revenues was $35.5 billion, up 11%. Other cost of revenues was $22.1 billion, up 14%, with the increase driven primarily by content acquisition costs, followed by depreciation as well as the impact of the Canadian digital services tax, which was applied retroactively.

第二季度，我們的綜合營收為 847 億美元，成長 14%，以固定匯率計算成長 15%。搜尋仍然是收入成長的最大貢獻者。在支出方面，總收入成本為 355 億美元，成長 11%。其他收入成本為 221 億美元，成長 14%，成長主要源自於內容獲取成本，其次是折舊以及追溯實施的加拿大數位服務稅的影響。

Operating expenses were $21.8 billion, up 5%, primarily reflecting an increase in R&D partially offset by a decline in G&A with sales and marketing essentially flat to the second quarter last year. The increase in R&D was driven primarily by compensation, which was affected by lapping a reduction in valuation-based compensation liabilities in certain other bets in the second quarter last year followed by depreciation. The largest single factor in the year-on-year decline in G&A expenses was lower charges related to legal matters. Operating income was $27.4 billion, up 26% and our operating margin was 32%. Net income was $23.6 billion and EPS was $1.89.

營運費用為 218 億美元，成長 5%，主要反映了研發費用的成長，但被一般及行政費用的下降部分抵消，銷售和行銷費用與去年第二季度基本持平。研發費用的成長主要由薪酬推動，薪酬受到去年第二季度某些其他投資中基於估值的薪酬負債減少以及隨後折舊的影響。一般及行政費用較去年同期下降的最大單一因素是與法律事務相關的費用降低。營業收入為 274 億美元，成長 26%，營業利益率為 32%。淨收入為 236 億美元，每股收益為 1.89 美元。

We delivered free cash flow of $13.5 billion in the second quarter and $60.8 billion for the trailing 12 months. As a reminder, last year, we had a timing benefit in the second and third quarters from a $10.5 billion deferred cash tax payment made in the fourth quarter, which depressed reported free cash flow growth this quarter, and will do so again next quarter. We ended the quarter with $101 billion in cash and marketable securities.

我們在第二季度實現了 135 億美元的自由現金流，過去 12 個月實現了 608 億美元的自由現金流。提醒一下，去年，我們在第二季度和第三季度獲得了時間效益，因為第四季度支付了 105 億美元的遞延現金稅，這抑制了本季度報告的自由現金流增長，並且下個季度還會再次出現這種情況。本季末，我們擁有 1,010 億美元的現金和有價證券。

Turning to segment results. Within Google services, revenues were $73.9 billion, up 12%. Google Search and other advertising revenues of $48.5 billion in the quarter were up 14%, led again by growth in retail, followed by the financial services vertical. YouTube advertising revenues of $8.7 billion were up 13%, driven by brand followed by direct response advertising.

轉向細分結果。谷歌服務業務營收為 739 億美元，成長 12%。本季度，Google搜尋和其他廣告收入為 485 億美元，成長 14%，這再次得益於零售業的成長，其次是金融服務垂直產業。YouTube 廣告收入達 87 億美元，成長 13%，主要得益於品牌廣告和直接反應廣告。

Network advertising revenues of $7.4 billion were down 5%. Subscription platforms and devices revenues were $9.3 billion, up 14%, primarily reflecting growth in YouTube subscription revenues. TAC was $13.4 billion, up 7%. Google Services operating income was $29.7 billion, up 27% and the operating margin was 40%.

網路廣告收入74億美元，下降5%。訂閱平台和設備收入為 93 億美元，成長 14%，主要反映了 YouTube 訂閱收入的成長。TAC 為 134 億美元，成長 7%。谷歌服務營運收入為297億美元，成長27%，營運利潤率為40%。

Turning to the Google Cloud segment. Revenues were $10.3 billion for the quarter, up 29%, reflecting first significant growth in GCP, which was above growth for cloud overall and includes an increasing contribution from AI; and second, strong Google Workspace growth, primarily driven by increases in average revenue per seat. Google Cloud delivered operating income of $1.2 billion and an operating margin of 11%. As to our Other Bets for the second quarter, revenues were $365 million and the operating loss was $1.1 billion.

轉向 Google Cloud 部分。本季營收為 103 億美元，成長 29%，反映了 GCP 的首次顯著成長，其成長高於雲端運算整體的成長，其中包括人工智慧的貢獻不斷增加；其次，Google Workspace 成長強勁，主要得益於每座位平均營收的成長。Google Cloud 的營業收入為 12 億美元，營業利潤率為 11%。至於我們第二季的其他業務，營收為 3.65 億美元，營運虧損為 11 億美元。

Turning to our outlook for the business. With respect to Google services, first, within advertising. The strong performance of search was broad-based across verticals. In YouTube, we are pleased with the growth in the quarter. We had healthy watch time growth, continued to close the monetization gap in Shorts and had continued momentum in Connected TV with brand benefiting in part from an ongoing shift in budgets from linear television to digital. As we look forward to the third quarter, we will be lapping the increasing strength in advertising revenues in the second half of 2023 in part from APAC-based retailers.

談談我們對業務的展望。關於Google服務，首先是在廣告方面。搜尋的強勁表現遍佈各個垂直領域。在 YouTube 方面，我們對本季的成長感到滿意。我們的觀看時間成長穩健，持續縮小 Shorts 的貨幣化差距，並在連網電視領域保持強勁發展勢頭，品牌部分受益於預算從線性電視向數位電視的持續轉變。展望第三季度，我們將在 2023 年下半年享受來自亞太地區零售商的部分廣告收入的強勁成長。

Turning to subscriptions, platforms and devices. First, we continue to have significant growth in our subscriptions business which drives the majority of revenue growth in this line. However, there was a sequential decline in the year-on-year growth rate as we anniversaried the impact of a price increase for YouTube TV in the second quarter last year. The impact will persist through the balance of the year.

轉向訂閱、平台和設備。首先，我們的訂閱業務繼續保持顯著成長，這推動了該業務的大部分收入成長。然而，由於去年第二季 YouTube TV 價格上漲的影響，年成長率出現季減。這種影響將持續到今年底。

Second, with regard to platforms. We are pleased with the performance in play driven by an increase in buyers. Finally, with respect to devices. The most important point as we look forward is that our Made by Google launches have been pulled forward into the third quarter from the fourth quarter last year, benefiting revenues in Q3 this year.

第二，關於平台。我們對因買家增加而帶來的業績表現感到滿意。最後，關於設備。展望未來，最重要的一點是，我們的「Google製造」發布會從去年第四季度提前到了第三季度，這將有利於今年第三季度的收入。

Turning to Cloud, which continued to deliver very strong results. For the first time, Cloud crossed $10 billion in quarterly revenues and $1 billion in quarterly operating profit. As Sundar noted, year-to-date, our AI infrastructure and generative AI solutions for cloud customers have already generated billions in revenues and are being used by more than 2 million developers. We're particularly encouraged that the majority of our top 100 customers are already using our generative AI solutions. We continue to invest aggressively in the business.

轉向雲端運算，它繼續帶來非常強勁的業績。雲端運算季度營收首度突破100億美元，季度營業利潤首度突破10億美元。正如 Sundar 所指出的，今年迄今為止，我們為雲端客戶提供的人工智慧基礎設施和生成式人工智慧解決方案已經創造了數十億美元的收入，並被超過 200 萬開發人員使用。我們尤其感到鼓舞的是，我們的前 100 名客戶中的大多數已經在使用我們的生成式 AI 解決方案。我們將繼續大力投資該業務。

Turning to margins. The margin expansion in Q2 versus last year reflects our ongoing efforts to durably reengineer our cost base as well as revenue strength. Our leadership team remains focused on our efforts to moderate the pace of expense growth in order to create capacity for the increases in depreciation and expenses associated with the higher levels of investment in our technical infrastructure. Once again, headcount declined quarter-on-quarter, which reflects both actions we have taken in the first half of the year and a much slower pace of hiring.

轉向利潤。與去年同期相比，第二季的利潤率有所擴大，這反映了我們持續努力重塑成本基礎和收入實力。我們的領導團隊將繼續致力於減緩費用成長速度，以便為因技術基礎設施投資增加而導致的折舊和費用增加創造能力。員工人數再次環比下降，這反映了我們上半年採取的行動以及招募速度的大幅放緩。

Looking ahead, we expect a slight increase in headcount in the third quarter as we bring on new graduates. As we have discussed previously, we are continuing to invest in top engineering and technical talent, particularly in cloud and technical infrastructure. Looking forward, we continue to expect to deliver full year 2024 Alphabet operating margin expansion relative to 2023.

展望未來，隨著新畢業生的加入，我們預計第三季員工人數將略有增加。正如我們之前討論過的，我們將繼續投資頂尖工程和技術人才，特別是在雲端和技術基礎設施方面。展望未來，我們繼續預期 Alphabet 2024 年全年營業利潤率將較 2023 年擴大。

However, in the third quarter, operating margins will reflect the impact of both the increases in depreciation and expenses associated with the higher levels of investment in our technical infrastructure as well as the increase in cost of revenues due to the pull forward of hardware launches into Q3.

然而，在第三季度，營業利潤率將反映出與我們技術基礎設施投資水準提高相關的折舊和費用增加的影響，以及由於硬體發布提前到第三季度而導致的收入成本增加的影響。

With respect to CapEx, our reported CapEx in the second quarter was $13 billion, once again, driven overwhelmingly by investment in our technical infrastructure with the largest component for servers followed by data centers. Looking ahead, we continue to expect quarterly CapEx throughout the year to be roughly at or above the Q1 CapEx of $12 billion, keeping in mind that the timing of cash payments can cause variability in quarterly reported CapEx.

關於資本支出，我們報告的第二季度資本支出為 130 億美元，這再次主要受到對技術基礎設施的投資推動，其中最大的組成部分是伺服器，其次是資料中心。展望未來，我們繼續預期全年季度資本支出將大致等於或高於第一季的 120 億美元，同時記住現金支付的時間可能會導致季度報告的資本支出發生變化。

With regard to Other Bets, we continue to focus on improving overall efficiencies as we invest for long-term returns. Waymo is an important example of this with its technical leadership coupled with progress on operational performance. As you will see in the 10-Q, we have chosen to commit to a new multiyear investment of $5 billion. This new round of funding, which is consistent with recent annual investment levels will enable Waymo to continue to build the world's leading autonomous driving technology company.

對於其他投資，我們將繼續致力於提高整體效率，因為我們正在進行長期回報的投資。Waymo 就是一個重要的例子，它不僅擁有技術領先優勢，而且在營運績效方面也取得了進步。正如您在 10-Q 中看到的，我們已選擇承諾進行 50 億美元的新多年期投資。本輪融資與近幾年的年度投資水準一致，將使 Waymo 繼續打造全球領先的自動駕駛技術公司。

To close, this is my 56th and last earnings call, 37 of them at Alphabet. So I have a few closing thoughts of gratitude. I've been so proud to be at Google and Alphabet as CFO and to work with some of the smartest people in the world every day. I think we have accomplished a lot in the last nine-plus years, and I am confident that progress will continue.

最後，這是我第 56 次也是最後一次財報電話會議，其中 37 次是在 Alphabet。最後，我表達一些感激。我很自豪能夠擔任 Google 和 Alphabet 的首席財務官，並每天與世界上最聰明的人一起工作。我認為我們在過去九年多的時間裡取得了很多成就，我相信我們將繼續取得進步。

Of course, I'm not going far and I'm honored to have my new role, which I've been slowly working my way into during the past 11 months and I look forward to continuing to work with Sundar, and our great team. Being CFO of one of the most important companies in the world has been the opportunity and responsibility of a lifetime. Google's mission of advancing technology and bringing information to people throughout the world is as relevant today as it was when I worked on its IPO.

當然，我不會走遠，我很榮幸能擔任新職務，在過去的 11 個月裡，我一直在慢慢努力適應這個新職務，我期待著繼續與 Sundar 和我們優秀的團隊合作。擔任世界上最重要的公司之一的財務長是我一生中最大的機會和責任。谷歌的使命是推動技術進步，為全世界人民提供信息，這一使命在今​​天仍然與我參與其首次公開募股時一樣重要。

Technology has been a catalyst for economic growth throughout human history. The people on this call know that if technological advancement is not the focus of every business and government, it will be left behind. Underpinning this is the need for sound and responsible investment. That has never been more important than today and certainly, that is Google and Alphabet's focus.

在整個人類歷史中，科技一直是經濟成長的催化劑。參加這次電話會議的人們都知道，如果技術進步不是每個企業和政府關注的重點，那麼它就會落後。支撐這一目標的根本是需要合理且負責任的投資。這一點在今天比以往任何時候都更重要，當然，這也是Google和 Alphabet 的重點。

I want to end by thanking Googlers around the world for the innovation and commitment that has enabled us to deliver such extraordinary products and services globally. I also want to thank our investors and analysts for your long-term support and your feedback. Thank you. Sundar, Philipp and I will now take your questions.

最後，我要感謝世界各地的Google員工，感謝他們的創新和奉獻，使我們能夠在全球範圍內提供如此卓越的產品和服務。我還要感謝我們的投資者和分析師的長期支持和回饋。謝謝。桑達爾、菲利普和我現在將回答你們的問題。

(Operator Instructions) Brian Nowak, Morgan Stanley.

（操作員指示）摩根士丹利的 Brian Nowak。

First, thank you, Ruth, for all the help and significant impact over the past decade. The first one, it's a little bit of a jump, I guess, for Sundar, Philipp or Ruth. I guess we're sort of 18 months this fever pitch around Gen AI focus in the world. Maybe from any of your perspective, can you just sort of talk to us about areas where you've seen faster-than-expected traction or testing adoption of some of the AI, generative AI capabilities versus slower-than-expected traction and testing from a Google perspective? And then, Ruth, I appreciate all the comments about structurally reengineering the OpEx base. Are there any more tangible examples of areas you can talk to us about where you still see further ways to drive more efficiency across the company?

首先，感謝露絲在過去十年中提供的所有幫助和重大影響。我想，對於 Sundar、Philipp 或 Ruth 來說，第一個有點跳躍。我想我們已經在全世界範圍內對通用人工智慧的關注熱潮中度過了 18 個月。也許從您的任何角度來看，您能否與我們談談您看到的哪些領域對某些人工智慧、生成人工智慧功能的牽引力或測試採用速度比預期更快，而從谷歌的角度來看，哪些領域的牽引力和測試速度比預期更慢？然後，露絲，我很感謝所有關於結構性重組營運支出基礎的評論。您能否與我們談談更具體的例子，說明您認為還有什麼方法可以提高整個公司的效率？

Brian, I'll take the first part. I think it's a good question. Obviously, I think there is a time curve in terms of taking the underlying technology and translating it into meaningful solutions across the board, both on the consumer and the enterprise side. Definitely, on the consumer side, I'm pleased, as I said in my comments earlier, in terms of how for a product like Search, which is used at that scale over many decades, how we've been able to introduce it in a way that it's additive and enhances to -- enhances overall experience and this -- positively contributing there. I think across our consumer products, we've been able -- I think we are seeing progress on the organic side.

布萊恩，我將擔任第一部分。我認為這是個好問題。顯然，我認為在採用底層技術並將其轉化為消費者和企業方面全面有意義的解決方案方面存在一個時間曲線。毫無疑問，從消費者角度來看，正如我之前的評論中所說的那樣，對於像搜尋這樣已經大規模使用幾十年的產品，我們能夠以一種附加的方式引入它，並增強它——增強整體體驗，並在這方面做出積極貢獻。我認為，在我們的消費產品中，我們已經能夠——我認為我們在有機方面看到了進展。

Obviously, monetization is something that we would have to earn on top of it. The enterprise side, I think we are at a stage where definitely there are a lot of models. I think roughly, the models are all kind of converging towards sort of base capabilities, but I think where the next wave is working to build solutions on top of it. And I think there are pockets, be it coding, be it in customer service, et cetera, where we are seeing some of those use cases seeing traction, but I still think there is hard work there to completely unlock those.

顯然，貨幣化是我們必須在此基礎上努力實現的。在企業方面，我認為我們正處於一個肯定有很多模型的階段。我認為，粗略地講，這些模型都朝著某種基本能力的方向發展，但我認為下一波浪潮將致力於在其基礎上建立解決方案。我認為，無論是在編碼、客戶服務等方面，我們都看到了一些用例正在發揮作用，但我仍然認為，要完全解鎖這些用例，還需要做很多工作。

And on your second question first, thank you for your comments. Look, the reason we've consistently used the term, the phrase that we're focused on durably reengineering our cost base is because these are deep work streams. They're not tactical fixes, and we continue to build on them. And so the main areas that we've talked about are around product and process prioritization around organizational efficiency and structure.

首先關於您的第二個問題，感謝您的評論。你看，我們之所以一直使用這個術語，之所以說我們專注於持久地重新設計我們的成本基礎，是因為這些都是深度工作流程。它們不是戰術上的解決辦法，我們會繼續在此基礎上繼續改進。因此，我們討論的主要領域是圍繞組織效率和結構的產品和流程優先順序。

Both of those are reflected in the headcount and the fact that headcount is down year-on-year. And across all of those -- as I said, across our entire leadership team, we remain focused on continuing to execute against them.

這兩點都反映在員工人數上，而且員工人數較去年同期下降。正如我所說，在我們整個領導團隊中，我們將繼續專注於執行這些目標。

So in terms of the most recent examples, as we talked about last quarter with the combination of the devices and services product area with the platforms and ecosystems product area, we announced that back in April. And what we discussed last quarter and what we're seeing is that unifying teams across these organizations helps with product execution and what we're really focused on is really adding to velocity and efficiency. So kind of the gift that keeps giving.

就最近的例子而言，正如我們上個季度討論的，將設備和服務產品領域與平台和生態系統產品領域結合起來，我們在四月份就宣布了這一點。我們上個季度討論的內容以及我們看到的是，統一這些組織中的團隊有助於產品執行，而我們真正關注的是真正提高速度和效率。這是一種持續給予的禮物。

And then more broadly, all of the work streams that we've talked to you about before, we continue to remain focused on. A big one, very important one, is all of the efficiency efforts, the work streams around technical infrastructure and improving efficiency there. We're also working on the use of AI across Alphabet. We're working on continuing to build on what we've done with our centralized procurement organization. We're continuing to optimize our real estate portfolio. And so again, this is across our leadership team. These are efforts that all build to this phrase, durably reengineering our cost base.

更廣泛地說，我們之前與您討論過的所有工作流程，我們都會繼續關注。其中一個重要且非常重要的方面是所有的效率努力、圍繞技術基礎設施的工作流程以及提高其效率。我們也致力於在 Alphabet 中使用人工智慧。我們正在努力繼續鞏固我們集中採購組織所取得的成就。我們正在繼續優化我們的房地產投資組合。再次強調，這涉及我們的領導團隊。所有這些努力都是為了實現這句話，並持久地重塑我們的成本基礎。

Doug Anmuth, JPMorgan.

摩根大通的道格安穆斯 (Doug Anmuth)。

One for Ruth and one for Sundar. Ruth, you've now had Google Services operating margins roughly 40% for the past two quarters. Just as you create more capacity to help offset the future investments, is it reasonable to think that you could sustain at those kind of levels going forward?

一個給露絲 (Ruth)，一個給桑達爾 (Sundar)。露絲，過去兩個季度，Google服務的營運利潤率約為 40%。正如您創造更多產能來幫助抵消未來的投資一樣，您是否合理地認為您可以在未來維持這種水平？

And then, Sundar, just as it relates to AI Overviews, you talked about the positive trends there. Can you just help us understand where you are, how far along in rolling out AI Overviews and then any more color around kind of click-through rates and monetization levels relative to your traditional searches?

然後，Sundar，正如它與 AI 概述相關一樣，您談到了那裡的積極趨勢。您能否幫助我們了解一下你們的現狀，你們推出 AI 概覽的進展如何，以及相對於傳統搜尋的點擊率和貨幣化程度方面的更多資訊？

So in terms of the Google services operating margin, it did reflect all the work that we're doing on durably reengineering the cost base. It also reflected the benefit of strong revenue performance in Search. And so what I tried to lay out in the comments, as we look forward to the third quarter is operating margins will reflect the increases in depreciation and expenses associated with higher levels of our investment in technical infrastructure.

因此，就Google服務營運利潤率而言，它確實反映了我們為持久地重新設計成本基礎所做的所有工作。這也反映出搜尋業務強勁的營收表現。因此，我在評論中試圖闡明的是，當我們展望第三季度時，營業利潤率將反映出與我們對技術基礎設施的投資水準提高相關的折舊和費用的增加。

It will also reflect higher expenses associated with the Pixel launch due to the pull forward. So those are important factors. I would say, overall, company-wide, it's important to note that we do expect to continue to deliver full year '24 Alphabet operating margin expansion relative to 2023, but I did want to highlight those important points as we look forward to Q3.

由於提前發布 Pixel，這也將反映與發布相關的更高費用。這些都是重要因素。我想說，總體而言，就全公司而言，值得注意的是，我們確實預計 2024 年全年 Alphabet 營業利潤率將繼續相對於 2023 年擴大，但在展望第三季度時，我確實想強調這些要點。

And Doug, thanks on the AI overview, we are -- we have rolled it out in the US and we are -- will be through the course of the year, definitely scaling it up, both to more countries and also we have taken a conservative start focused on quality, making sure the metrics are healthy and so on. But you will see us expand the use cases around it, and we'll touch definitely more queries. All the feedbacks we have seen are positive. And on the monetization side, I think Philipp has touched upon it. Maybe Philipp, anything more you want to add there?

道格，感謝您對人工智慧的概述，我們已經在美國推出了這項服務，並將在今年內擴大其規模，覆蓋更多國家。同時，我們採取了保守的開端，專注於質量，確保指標健康等等。但您會看到我們擴展了它的用例，並且我們肯定會觸及更多的查詢。我們看到的所有回饋都是正面的。在貨幣化方面，我認為菲利普已經談到了這一點。也許 Philipp，您還有什麼要補充的嗎？

Yeah. Yeah, look, innovation and improvements to the user experience on Search have historically opened up new opportunities for advertisers. We talked about this before we saw this when we navigated from desktop to mobile, for example. And we can see Gen AI obviously expand the types of questions we can help people with, as Sundar mentioned. And as we said before, people are finding ads, either above or below AI Overviews, helpful.

是的。是的，從歷史上看，搜尋用戶體驗的創新和改進為廣告商開闢了新的機會。例如，當我們從桌面導航到行動裝置時，我們在看到這個問題之前就討論過這個問題。正如 Sundar 所提到的，我們可以看到 Gen AI 明顯擴展了我們可以幫助人們解決的問題類型。正如我們之前所說，人們發現 AI 概覽上方或下方的廣告都很有幫助。

We have a solid baseline here from which we can innovate and as you have probably noticed at GML, we announced that soon we'll actually start testing search and shopping ads in AI Overviews for users in the US and they will have the opportunity to actually appear within the AI Overview in a section clearly labeled as Sponsored when they're relevant to both the query and the information in the AI Overview, really giving us the ability to innovate here and take this to the next level.

我們擁有堅實的基礎，可以從中創新，正如您可能在 GML 注意到的那樣，我們宣布很快將開始在 AI Overview 中為美國用戶測試搜索和購物廣告，當它們與查詢和 AI Overview 中的信息相關時，它們將有機會出現在 AI Overview 中明確標記為贊助的部分中，這真正使我們有能力在一個新的水平上提升

Michael Nathanson, Moffett.

邁克爾·納桑森，莫菲特。

I have two, one for Sundar and one for Philipp. Sundar, on decision this week to not deprecate cookies. I know it's been a long journey, can you talk a bit about what we should expect in terms of new experience in Chrome and why the company makes a decision not to go down the path on deprecating cookies?

我有兩個，一個給 Sundar，一個給 Philipp。 Sundar 本週決定不棄用 cookies。我知道這是一個漫長的旅程，您能否談談我們對 Chrome 的新體驗有何期待，以及為什麼公司決定不棄用 Cookie？

And then Philipp, I know it's only one quarter, but it's interesting that Search is growing faster than YouTube, which surprised some of us. But can you talk about what factors you think are behind the difference in growth rates between these markets? And is there anything on the AI front that you could see maybe reaccelerating YouTube growth as you've seen happen with Search?

然後菲利普，我知道這只是一個季度，但有趣的是，搜尋的成長速度比 YouTube 更快，這讓我們中的一些人感到驚訝。但是您能談談您認為造成這些市場成長率差異的因素是什麼嗎？您是否認為人工智慧方面的任何進步可能會重新加速 YouTube 的成長，就像您在搜尋領域看到的那樣？

And Michael, on cookies, look, I think, obviously, we are super committed to improving privacy for users in Chrome and that was the whole focus around privacy. Sandbox and we remind, committed on the journey, but on third-party cookies, given the implications across the ecosystems and considerations and feedback across so many stakeholders, we now believe user choice is the best path forward there. And we'll both improve privacy by giving users choice and we'll continue our investments in privacy-enhancing technologies, but it's obviously an area we will be taking feedback from the players in the ecosystem, and we are committed to being privacy first as well.

邁克爾，關於 Cookie，我認為，顯然，我們非常致力於改善 Chrome 用戶的隱私，而這正是圍繞隱私的重點。沙盒和我們提醒我們，在旅程中做出了承諾，但在第三方 cookie 上，考慮到整個生態系統的影響以及眾多利益相關者的考慮和反饋，我們現在相信用戶的選擇是最好的前進道路。我們將透過為用戶提供選擇來改善隱私，並將繼續在隱私增強技術方面進行投資，但顯然，我們將在此領域聽取生態系統參與者的回饋，我們也致力於將隱私放在首位。

And on the second part of your question, maybe, Ruth, you want to jump in first and then I take the rest, if needed.

關於你問題的第二部分，露絲，也許你想先插話，然後如果需要的話，我再回答剩下的問題。

Absolutely. So look, as we both noted, Search revenues really reflected broad-based growth across verticals. That was led by retail followed by financial services. I think your question really goes to the heart of the year-on-year growth comparison. And as both Philipp and I noted, we're really pleased with YouTube, the YouTube team, all that was done, it was driven by brand followed by direct response, and they have very strong ongoing operating metrics, which Philipp will comment on.

絕對地。所以，正如我們都注意到的，搜尋收入確實反映了各個垂直領域的廣泛成長。其中零售業是首屈一指的，其次是金融服務業。我認為您的問題確實觸及了同比增長率比較的核心。正如菲利普和我所指出的，我們對 YouTube 和 YouTube 團隊感到非常滿意，他們所做的一切都是由品牌驅動的，其次是直接反應，他們擁有非常強大的持續營運指標，菲利普將對此發表評論。

I think the important point to note, and I tried to tease out in opening comments, was that the deceleration in this year-on-year revenue growth for the second quarter versus the first quarter, primarily reflected the tougher comparison versus the first quarter because at that point, as you probably recall, YouTube was lapping negative year-on-year growth in Q1 last year.

我認為需要注意的重點是，我在開場白中試圖闡明的是，第二季度收入同比增長與第一季相比有所放緩，這主要反映了與第一季相比更加嚴格的比較，因為那時，正如你可能還記得的那樣，YouTube 在去年第一季度出現了同比負增長。

And then also Q1 benefited from the extra day from leap year. And so what you're also seeing here is with YouTube, we were anniversarying the ramp in APAC-based retailers that began in the second quarter last year and foreign exchange headwinds as well that we noted. And so it's -- there are some timing issues going on and what we're trying to highlight is the underlying operating strength. Back to you, Philipp.

而且第一季也受惠於閏年的額外一天。因此，您在這裡還可以看到，對於 YouTube 來說，我們正在慶祝去年第二季度開始的亞太地區零售商的成長，以及我們注意到的外匯逆風。因此，存在一些時間問題，而我們試圖強調的是潛在的營運實力。回到你這裡，菲利普。

Yeah, that was very comprehensive. Nothing really to add from my side here.

是的，非常全面。我這邊確實沒有什麼好補充的。

The only thing I would say, I think, to Brian's first question on areas where things are maybe taking longer. I think, look, we are all building multimodal models. At least Gemini has been natively multimodal from the ground up. But most of the use cases today is -- that have been unlocked have been around the tech side. So in terms of getting real generative audio, video experience is working well, I think there is still -- it's going to take some time. But over time, obviously, it will be deeply relevant to YouTube and so it's an area I'm excited about in the future.

我認為，對於布萊恩的第一個問題，我唯一想說的是，哪些領域的工作可能需要更長的時間。我認為，我們都在建立多模式模型。至少 Gemini 從一開始就是原生多模式的。但如今大多數已解鎖的用例都與技術方面有關。因此，就獲得真正的生成音訊、視訊體驗而言，我認為仍然需要一些時間。但隨著時間的推移，顯然它將與 YouTube 密切相關，因此這是我對未來感到興奮的領域。

Eric Sheridan, Goldman Sachs.

高盛的艾瑞克·謝裡丹。

And I'll echo the thanks for Ruth for all the insights and partnership over the years on these earnings calls. Sundar, maybe first for you, in terms of Cloud and bringing AI to the enterprise, I wanted to know if you go a little bit deeper in terms of how you're seeing AI actually get adopted, implemented, what it potentially could mean for the strategic positioning of the Cloud business and the potential for AI workloads to be a stimulant to revenue growth for Cloud First?

我還要對露絲多年來在收益電話會議上提供的所有見解和合作表示感謝。 Sundar，首先，關於雲端運算並將人工智慧引入企業，我想知道您是否可以更深入地了解您如何看待人工智慧的實際採用和實施，這對雲端運算業務的策略定位意味著什麼，以及人工智慧工作負載是否有可能刺激 Cloud First 的營收成長？

And then following up the last set of questions on YouTube or really about the macro or the ad environment, what do you guys, as a team, continue to learn about the subscription side of YouTube and the appetite for consumers to engage with a broader array of media products at the subscription layer?

然後，繼續回答關於 YouTube 的最後一組問題，或者實際上是關於宏觀或廣告環境的問題，作為一個團隊，你們對 YouTube 的訂閱方面以及消費者在訂閱層面參與更廣泛的媒體產品的興趣有何了解？

Look, on the Cloud and AI stuff, obviously, it's something which I think will end up being a big driver over time. I mentioned in my opening remarks already if you take a look at our AI infrastructure and generative AI solutions for cloud, across everything we do be it compute on the AI side, the products we have through Vertex AI, Gemini for Workspace and Gemini for Google Cloud, et cetera, we definitely are seeing traction. People are deeply engaging with Gemini models. Across Vertex and AI studio, we now have over 2 million developers playing around with these things. And you are definitely seeing early use cases.

你看，關於雲端運算和人工智慧，顯然，我認為隨著時間的推移，它最終會成為一個巨大的推動力。我在開場白中已經提到過，如果你看一下我們的人工智慧基礎設施和用於雲端的生成人工智慧解決方案，我們所做的一切，無論是人工智慧方面的運算，還是我們透過 Vertex AI、Gemini for Workspace 和 Gemini for Google Cloud 等提供的產品，我們肯定看到了發展勢頭。人們對雙子座模型的熱情十分濃厚。在 Vertex 和 AI 工作室中，我們現在有超過 200 萬名開發人員正在使用這些東西。而且您肯定看到了早期的用例。

But I think we are in this phase where we have to deeply work and make sure on these use cases, on these workflows, we are driving deeper progress on unlocking value, which I'm very bullish will happen. But these things take time. So -- but if I were to take a longer-term outlook, I definitely see a big opportunity here. And I think particularly for us, given the extent to which we are investing in AI, our research infrastructure leadership, all of that translates directly. And so I'm pretty excited about the opportunity space ahead.

但我認為我們正處於這個階段，我們必須深入工作並確保在這些用例、這些工作流程上，我們在釋放價值方面取得更深層的進展，我非常樂觀地認為這將會發生。但這些事情需要時間。所以——但如果我從更長遠的角度來看，我肯定會看到一個巨大的機會。我認為，特別是對我們來說，考慮到我們在人工智慧方面的投資程度，以及我們在研究基礎設施方面的領導地位，所有這些都直接轉化為現實。因此，我對未來的機會空間感到非常興奮。

And then on your second question with respect to subscriptions, I am [impressed] in your question, how strong is it. As I noted in opening comments that overall line subscriptions platforms and devices delivered healthy growth and that was led by subscriptions. And as we've said on many calls here in a row, the subscription revenue growth continued to be quite strong. It was driven by subscriber growth in both YouTube TV and YouTube Music premium.

然後關於您關於訂閱的第二個問題，我對您的問題印象深刻，它有多強大。正如我在開場評論中指出的那樣，整體訂閱平台和設備實現了健康成長，而這主要由訂閱業務推動。正如我們連續在多次電話會議上所說的那樣，訂閱收入成長持續強勁。這是由 YouTube TV 和 YouTube Music Premium 的用戶成長所推動的。

And then the other component within that line is Google One that also delivered strong subscriber and revenue growth. But I think the heart of your question is really around YouTube and that is the heart of the revenues in that line. So it continues to be very strong. We see a lot of take up in it, strong subscriber growth, really pleased with it. We did note that growth on that line decelerated due to anniversarying this YouTube TV price increase. But at the heart of it, our people interested in the subscription offerings and is to take up significant, we're really pleased with it.

該產品線中的另一個組成部分是 Google One，它也實現了強勁的用戶和收入成長。但我認為你問題的核心是 YouTube，它是這個產業收入的核心。因此它仍然非常強勁。我們看到它受到廣泛歡迎，用戶成長強勁，對此感到非常高興。我們確實注意到，由於 YouTube 電視價格上漲週年紀念，該系列的成長有所放緩。但從本質上講，我們的員工對訂閱服務很感興趣，並且將承擔重要的責任，我們對此感到非常高興。

Ross Sandler, Barclays.

巴克萊銀行的羅斯桑德勒。

Just two questions on the AI CapEx. So it looks like from the outside at least, the hyperscaler industry is going from kind of an under bill situation this time last year to better meeting the demand with capacity right now to potentially being overbuilt next year if these CapEx growth rates keep up. So do you think that's a fair characterization? And how are we thinking about the return on invested capital with this AI CapEx cycle. And then related to that, do you think that the AI industry is close to or far away from hitting some kind of wall on foundation model improvement in AI training based on like lack of availability of new data to train on? Just your thoughts on that would be great.

關於 AI CapEx 僅有兩個問題。因此，至少從外部來看，超大規模產業正在從去年這個時候的帳單不足狀況轉變為現在可以透過產能更好地滿足需求，如果這些資本支出成長率保持下去，明年可能會出現產能過剩的情況。那麼您認為這樣的描述公正嗎？我們如何看待這個 AI CapEx 週期的投資資本報酬。與此相關，您是否認為人工智慧產業即將或即將遇到某種障礙，例如由於缺乏可用於訓練的新數據而導致人工智慧訓練的基礎模型改進？只要你對此有想法就很好了。

I think great questions. Look, I -- obviously, we are at a early stage of what I view as a very transformative area and in technology when you're going through these transitions, aggressively investing upfront in a defining category, particularly in an area in -- which in a leveraged way cuts across all our core areas or products, including Search, YouTube and other services as well as fuels growth in Cloud and supports the innovative long-term bets and Other Bets. It's definitely something for us makes sense to lean in.

我認為這些問題很棒。看，我——顯然，我們正處於一個我認為非常具有變革性的領域的早期階段，在技術領域，當你經歷這些轉變時，我們會在一個決定性的類別上進行積極的前期投資，特別是在一個領域——它以槓桿的方式貫穿我們所有的核心領域或產品，包括搜尋、YouTube 和其他服務，並推動雲端運算的成長，並支持創新的長期投資和其他投資。對我們來說這絕對是一件有意義的事。

I think the one way I think about it is when we go through a curve like this, the risk of under-investing is dramatically greater than the risk of over-investing for us here, even in scenarios where, if it turns out that we are over-investing, we clearly -- these are infrastructure which are widely useful for us. They have long useful lives and we can apply it across, and we can work through that.

我認為，當我們經歷這樣的曲線時，投資不足的風險遠大於投資過度的風險，即使在我們投資過度的情況下，這些基礎設施對我們來說非常有用。它們的使用壽命很長，我們可以充分利用它們，並努力實現它們。

But I think not investing to be at the front here, I think, definitely has much more significant downsides. Having said that, we obsess around every dollar we put in. Our teams are -- work super hard, and I'm proud of the efficiency work, be it optimization of hardware, software, model deployment across our fleet. All of that is something we spend a lot of time on, and that's how we think about it.

但我認為，不投資處於領先地位，肯定會帶來更顯著的負面影響。話雖如此，我們對投入的每一美元都非常重視。我們的團隊非常努力地工作，我為高效的工作感到自豪，無論是硬體、軟體的最佳化，還是整個車隊的模型部署。所有這些都是我們在事情上花費大量時間的事情，我們就是這樣思考的。

To your second question on whether -- how do the scaling loss hold? Are we hitting on some kind of wall or something? Look, I think we are all pushing very hard, and there's going to be a few efforts which will scale up on the compute side and push the boundaries of these models. What I would tell is regardless of how that plays out, I still think there is enough optimizations we are all doing which is driving constant progress in terms of the capabilities of the models. And more importantly, taking them and translating into real use cases across the consumer and enterprise side, I think on that frontier, I think there's still a lot of progress to be had. And so we are pretty focused on that as well.

對於你的第二個問題，縮放損失如何維持？我們是不是碰到了什麼障礙還是別的什麼？看，我認為我們都在努力，並且會做出一些努力來擴大計算方面的規模並突破這些模型的界限。我想說的是，無論結果如何，我仍然認為我們都在進行足夠的最佳化，這將推動模型能力的持續進步。更重要的是，將它們轉化為消費者和企業方面的實際用例，我認為在這一領域，我們還有很大的進步空間。因此我們也非常關注這一點。

Stephen Ju, UBS.

瑞銀集團 Stephen Ju。

Okay. So Sundar, I guess, to ask the AI question a different way. As we talk to some of the model builders out there, it looks like the initial use cases are more on the cost savings or efficiency side. But when do we -- when do you think we'll start thinking about products that can help revenue generation for the Fortune 500, Fortune 1,000 companies, which is probably something that can hopefully create greater value over time versus just cutting costs?

好的。所以我想，Sundar 應該以不同的方式提出人工智慧問題。當我們與一些模型建構者交談時，看起來最初的用例更多是在節省成本或提高效率方面。但是，您認為我們什麼時候才會開始考慮能夠幫助財富 500 強、財富 1,000 強公司創造收入的產品，這些產品可能有望隨著時間的推移創造更大的價值，而不僅僅是削減成本？

And Philipp, listening to what will be, I guess, Ruth's final comments on Q2 on these public calls. And thank you, by the way, Ruth, for all the help. I couldn't help but notice that the bigger factors were brand followed by direct response. And if we continue to think that the one you bring up first is the larger factor and tying this into your prior commentary about shopping being an important consideration, when do you think we'll start talking about direct response being a much bigger contributor to YouTube's growth versus brand?

菲利普，我想，請聽一下露絲關於第二季這些公開電話會議的最後評論。順便說一句，感謝露絲提供的所有幫助。我禁不住注意到，更重要的因素是品牌，其次是直接反應。如果我們繼續認為您首先提到的因素是更重要的因素，並將其與您之前關於購物是一個重要考慮因素的評論聯繫起來，那麼您認為我們什麼時候會開始談論直接反應對 YouTube 增長的貢獻比品牌更大？

On the first part of the question, look, I think the technology's horizontal enough, it can apply on both sides. If you take a use case like improving the customer service experience, it is part of it, which is driving efficiencies, and you can look at it from a cost standpoint. But you could also be overall improving the experience, improving conversion, driving the funnel better and so increasing basket size if you're a retail e-commerce player, et cetera. So we are seeing people experiment across both sides. And so I think you will see it played across both sides.

關於問題的第一部分，我認為這項技術足夠水平，可以應用於兩側。如果您採用改善客戶服務體驗這樣的用例，它就是其中的一部分，可以提高效率，您可以從成本的角度來看待它。但如果您是零售電子商務參與者，您也可以整體改善體驗、提高轉換率、更好地推動通路並增加購物籃規模等等。因此我們看到人們正在兩方面進行實驗。所以我認為你會看到雙方都在上演這一幕。

Philipp, on the second one?

菲利普，第二個呢？

Yeah. On the second one, look, on the direct response side, as you know, it's about driving and converting commercial intent and customers are obviously benefiting from including video in their AI-powered campaigns, it could be P Max, it could be demand gen and obviously using our automated tools to enhance and create video creatives. And we're very, very optimistic about this path. On average, advertisers who run both image and video ads with demand gen campaigns see 6% more conversions per dollar than those running image only ads and discovery. And this is just one little example of how this can obviously boost your performance business. So that's a big part.

是的。第二，就直接回應方面而言，如您所知，它是關於推動和轉換商業意圖，並且客戶顯然受益於將視頻納入他們的人工智能驅動的活動，它可能是 P Max，可能是需求生成，並且顯然使用我們的自動化工具來增強和創建視頻創意。我們對這條路非常非常樂觀。平均而言，同時投放圖片和影片廣告並進行需求產生活動的廣告客戶，其每投入一美元的轉換率比僅投放圖片廣告和發現廣告的廣告客戶高出 6%。這只是一個小例子，它能明顯提升您的績效。這是很重要的一環。

The brand side, as you know, Google AI continues to make it easier for brands to show up next to the content where viewers are obviously, the most engaged, and they're finding it, as you can see from the numbers, a very effective way to drive awareness and consideration. And we're also quite excited about some of the recent launches on the YouTube shopping side, if you want to put that into the direct response bucket.

在品牌方面，正如您所知，Google AI 不斷使品牌更容易出現在觀眾最感興趣的內容旁邊，並且他們發現，正如您從數據中看到的那樣，這是一種提高知名度和考慮度的非常有效的方式。如果您想將其放入直接回應桶中，我們對 YouTube 購物方面最近推出的一些功能也感到非常興奮。

Justin Post, Bank of America.

美國銀行的賈斯汀·波斯特 (Justin Post)。

Great. I'll ask a couple of areas. First, on the cloud acceleration, would you characterize that as new AI demand helping drive that year-to-date? Or is that more of a rebound in just general compute and other demand or is AI really moving this forward and helping drive acceleration? And then I wanted to ask about your internal cost savings, which has been really strong. How are you using AI internally to help cut costs? Are you seeing better efficiencies with your engineers? And just would love to hear about how you're applying AI to cut your own costs?

偉大的。我會問幾個方面。首先，關於雲端運算的加速，您是否認為這是新的人工智慧需求推動今年迄今雲端運算的發展？或者這只是一般運算和其他需求的反彈，還是人工智慧真的在推動這一進程並有助於推動加速？然後我想問一下你們的內部成本節約情況，這確實非常強勁。您如何在內部使用人工智慧來幫助削減成本？您是否發現工程師的效率提高了？我很想聽聽您是如何應用人工智慧來削減成本的？

Great. Thank you for that. So overall, we're -- as both Sundar and I said, we're very pleased with the results in Cloud. And there is clearly a benefit as the Cloud team is engaging broadly with customers around the globe with AI-related solutions, AI infrastructure solutions and the generative AI solutions. I think we noted that we're particularly encouraged that the majority of our top 100 customers are already using our generative AI solution.

偉大的。謝謝你。所以總的來說，正如 Sundar 和我所說，我們對雲端運算的成果非常滿意。這顯然是有好處的，因為雲端團隊正在透過與人工智慧相關的解決方案、人工智慧基礎設施解決方案和產生人工智慧解決方案與全球客戶廣泛合作。我想我們注意到，我們特別受鼓舞的是，我們的前 100 名客戶中的大多數已經在使用我們的生成式 AI 解決方案。

So it is clearly adding to strength of the business on top of all that they're doing. And just to be really clear, the results for GCP, the growth rate for GCP is above the growth for cloud overall. And then I'll turn it to Sundar on the cost saving point. But just one point, we're really pleased as well that Cloud's margin improved as it did and in part, that reflects the revenue strength that they delivered and all of the efficiency efforts that I've already spoken about. But looking ahead in Q3, we do expect the same seasonal pattern that you saw last year with respect to margin, and we are continuing to invest in the business.

因此，這顯然在他們所做的一切之上增強了業務實力。需要明確的是，GCP 的結果顯示，GCP 的成長率高於雲端運算的整體成長率。然後我將把這個問題交給 Sundar 來討論成本節約問題。但有一點，我們也非常高興雲端的利潤率有所提高，這在一定程度上反映了他們提供的收入實力以及我已經談到的所有效率努力。但展望第三季度，我們確實預期利潤率將呈現與去年相同的季節性模式，我們將繼續對該業務進行投資。

Look, I think specifically, if the question is about engineers and coding, et cetera, we definitely want to be on the cutting edge there. I think we are making these tools available to some of the most productive engineers and demanding engineers out there, and they are definitely kicking the tires hard. And -- but I would say, it's still all in very early stages. I think particularly when it comes to writing high-quality secure code. But I think all the learnings what we are gaining here will translate into our models and products, and that's the virtuous cycle, which I'm excited by. So there's a lot more to come.

看，我認為具體來說，如果問題是關於工程師和編碼等等，我們肯定希望處於領先地位。我認為我們正在向一些最有效率的工程師和要求最高的工程師提供這些工具，他們肯定會努力嘗試。但我想說，一切仍處於非常早期的階段。我認為特別是在編寫高品質安全程式碼時。但我認為我們在這裡獲得的所有經驗都將轉化為我們的模型和產品，這是一個良性循環，我對此感到興奮。因此，未來還會有更多。

Thank you. And that concludes our question-and-answer session for today. I'd like to turn the conference back over to Jim Friedland for any further remarks.

謝謝。今天的問答環節到此結束。我想將會議交還給吉姆‧弗里德蘭 (Jim Friedland)，請他發表進一步的評論。

Thanks, everyone, for joining us today. We look forward to speaking with you again on our third quarter 2024 call. Thank you and have a good evening.

感謝大家今天加入我們。我們期待在 2024 年第三季電話會議上再次與您交談。謝謝您，祝您晚上愉快。

Thank you, everyone. This concludes today's conference call. Thank you for participating. You may now disconnect.

謝謝大家。今天的電話會議到此結束。感謝您的參與。您現在可以斷開連線。