Meta Platforms Inc (META) 2025 Q1 法說會逐字稿

Good afternoon. My name is Christa, and I will be your conference operator today. At this time, I would like to welcome everyone to the Meta first-quarter earnings conference call. (Operator Instructions) This call will be recorded.

午安.我叫克里斯塔，今天我將擔任您的會議主持人。現在，我歡迎大家參加 Meta 第一季財報電話會議。（接線員指示）此通話將會被錄音。

Thank you very much. Ken Dorell, Meta's Director of Investor Relations. You may begin.

非常感謝。Ken Dorell，Meta 投資者關係總監。你可以開始了。

Thank you. Good afternoon, and welcome to Meta's first-quarter 2025 earnings conference call. Joining me today to discuss our results are Mark Zuckerberg, CEO; and Susan Li, CFO. Our remarks today will include forward-looking statements, which are based on assumptions as of today. Actual results may differ materially as a result of various factors, including those set forth in today's earnings press release and in our annual report on Form 10-K filed with the SEC.

謝謝。下午好，歡迎參加 Meta 2025 年第一季財報電話會議。今天與我一起討論我們業績的有執行長馬克·祖克柏和財務長蘇珊·李。我們今天的評論將包括前瞻性陳述，這些陳述基於今天的假設。由於各種因素，實際結果可能存在重大差異，包括今天的收益新聞稿和我們向美國證券交易委員會提交的 10-K 表年度報告中所述的因素。

We undertake no obligation to update any forward-looking statement. During this call, we will present both GAAP and certain non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in today's earnings press release. The earnings press release and an accompanying investor presentation are available on our website at investor.atmeta.com.

我們不承擔更新任何前瞻性聲明的義務。在本次電話會議中，我們將介紹 GAAP 和某些非 GAAP 財務指標。今天的收益新聞稿中包含了 GAAP 與非 GAAP 指標的對帳。收益新聞稿和隨附的投資者介紹可在我們的網站 investor.atmeta.com 上查閱。

And now I'd like to turn the call over to Mark.

現在我想把電話轉給馬克。

All right. Thanks, Ken. Thanks, everyone, for joining today. We've had a strong start to the year. Our community keeps growing with more than 3.4 billion people now using at least one of our apps each day.

好的。謝謝，肯。感謝大家今天的參與。今年我們有一個好的開始。我們的社群不斷壯大，目前每天有超過 34 億人使用我們的至少一個應用程式。

Our business is also performing very well, and I think we're well positioned to navigate the macroeconomic uncertainty. The major theme right now of course, is how AI is transforming everything we do. And as we continue to increase our investments and focus more of our resources on AI, I thought it would be useful today to lay out the five major opportunities that we are focused on. Those are improved advertising, more engaging experiences, business messaging Meta AI and AI devices.

我們的業務表現也很好，我認為我們已準備好應對宏觀經濟的不確定性。當然，現在的主要主題是人工智慧如何改變我們所做的一切。隨著我們繼續增加投資並將更多資源集中在人工智慧上，我認為今天列出我們關注的五大機會將會很有用。這些是改進的廣告、更具吸引力的體驗、商業訊息傳遞元人工智慧和人工智慧設備。

And these are each long-term investments that are downstream from us building general intelligence and leading AI models and infrastructure. Even with our significant investments, we don't need to succeed in all of these areas to have a good ROI. But if we do, then I think that we will be wildly happy with the investments that we are making.

這些都是我們建構通用智慧和領先的人工智慧模型和基礎設施的下游長期投資。即使我們投入了大量資金，我們也不需要在所有這些領域都取得成功才能獲得良好的投資報酬率。但如果我們這樣做了，那麼我認為我們會對我們所做的投資感到非常滿意。

The first opportunity is improved advertising. Our goal is to make it that any business can basically tell us what objective they're trying to achieve, like selling something or getting a new customer and how much they're willing to pay for each result and then we just do the rest. Businesses used to have to generate their own ad creative and define what audiences they wanted to reach, but AI has already made us better at targeting and finding the audiences that will be interested in their products than many businesses are themselves and that keeps improving. And now AI is generating better creative options for many businesses as well. I think that this is really redefining what advertising is into an AI agent that delivers measurable business results at scale.

第一個機會是改進廣告。我們的目標是讓任何企業基本上都能告訴我們他們想要實現的目標，例如銷售某種產品或獲得一個新客戶，以及他們願意為每個結果支付多少錢，然後我們只需完成剩下的工作。企業過去必須自己製作廣告創意並確定他們想要接觸的受眾，但人工智慧已經讓我們比許多企業自己更善於定位和尋找對其產品感興趣的受眾，而這種能力還在不斷提高。現在，人工智慧也為許多企業創造了更好的創意選擇。我認為這實際上重新定義了廣告，使其成為能夠大規模提供可衡量業務成果的人工智慧代理商。

And if we deliver on this vision, then over the coming years, I think that the increased productivity from AI will make advertising a meaningfully larger share of global GDP than it is today. In just the last quarter, we are testing a new ads recommendation model for reels, which has already increased conversion rates by 5%. And we're seeing 30% more advertisers are using AI creative tools in the last quarter as well.

如果我們實現了這個願景，那麼在未來幾年裡，我認為人工智慧帶來的生產力提升將使廣告在全球 GDP 中所佔的份額比現在更大。就在上個季度，我們正在測試新的捲軸廣告推薦模型，該模型已經將轉換率提高了 5%。我們還發現上個季度使用 AI 創意工具的廣告商增加了 30%。

The second opportunity is more engaging experiences. This will come in two forms, better recommendations for existing content types and better new types of content. In the last six months, improvements to our recommendation systems have led to a 7% increase in time spent on Facebook, a 6% increase on Instagram and 35% on Threads. Threads now also has more than 350 million monthly actives and continues to be on track to become our next major social app. In addition to better recommendations for existing content types, AI is also enabling the creation of better content as well.

第二個機會是更具吸引力的體驗。這將以兩種形式出現：對現有內容類型的更好推薦和更好的新內容類型。在過去的六個月中，我們推薦系統的改進使得用戶在 Facebook 上的花費時間增加了 7%，在 Instagram 上的花費時間增加了 6%，在 Threads 上的花費時間增加了 35%。Threads 目前每月活躍用戶超過 3.5 億，預計將成為我們的下一個主要社交應用程式。除了對現有內容類型提供更好的推薦之外，人工智慧還可以創造更好的內容。

Some of this will be helping people produce better content to share themselves. Some of this will be AI generating content directly for people that is personalized for them. Some of this will be in existing formats like photos and videos, and some of it will be increasingly interactive.

其中一些將幫助人們創造更好的內容來分享。其中一些將是人工智慧直接為人們產生個人化的內容。其中一些將採用照片和影片等現有格式，而另一些將越來越具有互動性。

I've often talked about this long-term trend of content becoming richer over time. Our feeds started mostly with text and then became mostly photos when we all got mobile phones with cameras and then became mostly video when mobile networks became fast enough to handle that well. We are now in the video era, but I don't think that this is the end of the line. In the near future, I think that we're going to have content in our feeds that you can interact with and that it will interact back with you rather than you just watching it. Over the long term, as AI unlocks more productivity in the economy, I also expect that people will spend more of their time on entertainment and culture, which will create an even larger opportunity to create more engaging experiences across all of these apps.

我經常談論內容隨著時間的推移而變得更加豐富的長期趨勢。一開始，我們的資訊流主要以文字為主，當我們都擁有帶相機的手機時，資訊流主要以照片為主，而當行動網路速度夠快時，資訊流主要以影片為主。我們現在處於視訊時代，但我不認為這是時代的終結。在不久的將來，我認為我們將在我們的資訊流中添加您可以與之互動的內容，並且它也會與您互動，而不是您只是觀看它。從長遠來看，隨著人工智慧釋放出更多的經濟生產力，我還預計人們將在娛樂和文化上花費更多的時間，這將為在所有這些應用程式中創造更具吸引力的體驗創造更大的機會。

The third opportunity is business messaging. Right now, the vast majority of our business is advertising and feeds on Facebook and Instagram. But WhatsApp now has more than 3 billion monthly actives, with more than 100 million people in the US and growing quickly there. Messenger is also used by more than 1 billion people each month, and there are now as many messages sent each day on Instagram as they are on Messenger.

第三個機會是商業訊息傳遞。目前，我們的絕大部分業務是在 Facebook 和 Instagram 上投放廣告和訊息。但 WhatsApp 目前每月活躍用戶已超過 30 億，其中美國用戶超過 1 億，且成長迅速。Messenger 每月的用戶數超過 10 億，現在 Instagram 上每天發送的訊息數量與 Messenger 上一樣多。

So business messaging should be the next pillar of our business. In countries like Thailand and Vietnam, where there is a low cost of labor, we see many businesses conduct commerce through our messaging apps. There's actually so much business through messaging that those countries are both in our top 10 or 11 by revenue, even though they're ranked in the 30s in global GDP. This phenomenon hasn't yet spread to developed countries because the cost of labor is too high to make this a profitable model before AI, but AI should solve this.

因此，商業資訊傳遞應該成為我們業務的下一個支柱。在泰國和越南等勞動力成本較低的國家，我們看到許多企業透過我們的訊息應用程式進行商業活動。事實上，透過簡訊開展的業務非常多，儘管這些國家的 GDP 排名在全球 30 多位，但它們的收入卻都排在我們的前 10 名或前 11 名。這種現象還沒有蔓延到已開發國家，因為在人工智慧出現之前，勞動成本太高，無法使其成為一種獲利模式，但人工智慧應該可以解決這個問題。

So in the next few years, I expect that just like every business today has an e-mail address, social media account and website. They'll also have an AI business agent that can do customer support and sales, and they should be able to set that up very easily given all the context that they've already put into our business platforms. And we're going to have more to share on upcoming calls about our progress in this area.

因此，在未來幾年，我希望就像今天的每家企業一樣，擁有一個電子郵件地址、社群媒體帳號和網站。他們還將擁有一個可以進行客戶支援和銷售的人工智慧業務代理，並且考慮到他們已經放入我們業務平台的所有背景，他們應該能夠非常輕鬆地進行設定。我們將在即將召開的電話會議上分享更多有關我們在這一領域取得的進展的資訊。

The fourth opportunity is Meta AI across our apps, there are now almost 1 billion monthly actives using Meta AI. Our focus for this year is deepening the experience in making AI the leading personal AI with an emphasis on personalization, voice conversations and entertainment. I think that we're all going to have an AI that we talk to throughout the day, while we're browsing content on our phones, and eventually, as we're going through our days with glasses. And I think that this is going to be one of the most important and valuable services that has ever been created. In addition to building Meta AI into our apps, we just released our first Meta AI stand-alone app.

第四個機會是我們的應用程式中的 Meta AI，現在每月有近 10 億活躍用戶使用 Meta AI。我們今年的重點是深化體驗，使AI成為領先的個人AI，重點是個人化、語音對話和娛樂。我認為，我們每個人都會擁有一個人工智慧，我們可以整天與它交談，也可以在瀏覽手機內容時與它交談，最終，我們可以戴著眼鏡度過每一天。我認為這將是有史以來最重要的和最有價值的服務之一。除了將 Meta AI 建置到我們的應用程式之外，我們也剛剛發布了我們的第一個 Meta AI 獨立應用程式。

It is personalized. So you can talk to it about interests that you've shown, while browsing reels or different content across our apps. And we built a social feed into it. So you can discover entertaining ways that others are using Meta AI. And initial feedback on the app has been good so far.

它是個性化的。因此，您可以在瀏覽我們應用程式中的捲軸或不同內容時與它談論您所表現出的興趣。我們在其中建立了社交資訊流。因此，您可以發現其他人使用 Meta AI 的有趣方式。到目前為止，該應用程式的初步反饋良好。

Over time, I expect that the business opportunity for Meta AI to follow our normal product development playbook. First, we build and scale the product. And then once it is at scale, then we focus on revenue. In this case, I think that there will be a large opportunity to show product recommendations or ads as well as a premium service for people who want to unlock more compute for additional functionality or intelligence. But I expect that we're going to be largely focused on scaling and deepening engagement for at least the next year before we'll really be ready to start building out the business here.

隨著時間的推移，我預計 Meta AI 的商業機會將遵循我們正常的產品開發劇本。首先，我們建構並擴展產品。一旦規模擴大，我們就會關注收入。在這種情況下，我認為對於想要解鎖更多計算以獲得額外功能或智慧的人來說，將有很大的機會展示產品推薦或廣告以及高級服務。但我預計，至少在明年，我們將主要專注於擴大規模和深化參與，然後我們將真正準備好開始在這裡開展業務。

The fifth opportunity is AI devices, which is increasingly how we are thinking about our work on the next generation of computing platforms. Glasses are the ideal form factor for both AI and the Metaverse. They enable you to let an AI see what you see, hear what you hear and talk to you throughout the day, and they let you blend the physical and digital worlds together with holograms.

第五個機會是人工智慧設備，這越來越成為我們思考下一代運算平台工作的方式。眼鏡是人工智慧和元宇宙的理想外形。它們使你能夠讓人工智慧一整天看到你所看到的內容、聽到你所聽到的內容並與你交談，並讓你將物理世界和數位世界與全像圖融合在一起。

More than 1 billion people worldwide wear glasses today, and it seems highly likely that these will become AI glasses over the next 5 to 10 years, building the devices that people use to experience our services lets us deliver the highest quality AI and social experiences. And this will serve as an amplifier on all of the opportunities I've mentioned so far as well as unlocking some new opportunities as well.

如今，全球有超過 10 億人佩戴眼鏡，而且很有可能在未來 5 到 10 年內，這些眼鏡將成為人工智慧眼鏡，建構人們用來體驗我們服務的設備使我們能夠提供最高品質的人工智慧和社交體驗。這將增強我迄今為止提到的所有機會，同時也帶來一些新的機會。

Ray-Ban Meta AI glasses have tripled in sales in the last year. The people who have them are using them a lot. We've got some exciting new launches with our partner, Essilor Luxottica later this year as well that should expand that category and add some new technological capabilities to the glasses.

雷朋 Meta AI 眼鏡的銷量去年增長了兩倍。擁有它們的人正在大量使用它們。今年晚些時候，我們還將與合作夥伴 Essilor Luxottica 一起推出一些令人興奮的新產品，這將擴大該類別並為眼鏡增加一些新的技術功能。

On Quest, we are also seeing deeper engagement as Quest 3S makes VR accessible to more people and more people are creating experiences in horizon with AI tools. Now everything that I've talked about today is built on top of our AI models and our infrastructure. We released the first Llama 4 models earlier this month. They are some of the most intelligent, best multimodal, lowest latency and most efficient models that anyone has built. We have more models on the way, including the massive Llama 4 Behemoth model.

在 Quest 上，我們也看到了更深層的參與，因為 Quest 3S 讓更多人可以使用 VR，更多人正在使用 AI 工具在 Horizo​​n 中創造體驗。我今天談論的一切都建立在我們的人工智慧模型和基礎設施之上。我們本月初發布了第一款 Llama 4 車型。它們是迄今為止人們所建構的最聰明、最佳多模式、最低延遲和最高效的模型。我們還有更多的模型即將推出，包括巨大的 Llama 4 Behemoth 模型。

Overall, we are focused on building full general intelligence. All of the opportunities that I've discussed today are downstream of delivering general intelligence and doing so efficiently. The pace of progress across the industry and the opportunities ahead for us are staggering. I want to make sure that we're working aggressively and efficiently.

總的來說，我們專注於建立完整的通用智能。我今天討論的所有機會都與提供通用情報以及高效提供通用情報有關。整個產業進步的速度和我們面臨的機會是驚人的。我想確保我們積極、有效率地工作。

And I also want to make sure that we are building out the leading infrastructure and teams that we need to achieve our goals. So to that end, we are accelerating some of our efforts to bring capacity online more quickly this year as well as some longer-term projects that will give us the flexibility to add capacity in the coming years as well. And that has increased our planned investment for this year. More broadly, this has been a good start to what I expect will continue to be an intense year. We've got a lot more exciting work in the pipeline that I'm looking forward to sharing soon.

我還想確保我們正在建立實現目標所需的領先基礎設施和團隊。為此，我們正在加快一些努力，以便今年更快地實現產能上線，並加快一些長期項目，以便我們能夠在未來幾年靈活地增加產能。這增加了我們今年的計劃投資。更廣泛地說，這是一個好的開端，我預計今年仍將是緊張的一年。我們還有很多令人興奮的工作正在籌備中，我期待很快與大家分享。

I continue to think that this year is going to be a pivotal moment for our industry, and I am grateful for everyone who is working so hard at the company to deliver all of this amazing technology and new experiences. As always, thank you all for being on this journey with us. And now, Susan.

我仍然認為今年將是我們行業的關鍵時刻，我感謝公司裡所有辛勤工作、為我們提供所有這些令人驚嘆的技術和新體驗的人。一如既往，感謝大家與我們一起踏上這段旅程。現在，蘇珊。

Thanks, Mark, and good afternoon, everyone. Let's begin with our consolidated results. All comparisons are on a year-over-year basis, unless otherwise noted. Q1 total revenue was $42.3 billion, up 16% or 19% on a constant currency basis. Q1 total expenses were $24.8 billion, up 9% compared to last year. In terms of the specific line items, cost of revenue increased 14% driven primarily by higher infrastructure costs and payments to partners, partially offset by a benefit from the previously announced extension of server useful lives.

謝謝，馬克，大家下午好。讓我們從綜合結果開始。除非另有說明，所有比較均以同比為基礎。第一季總營收為 423 億美元，成長 16%，以固定匯率計算成長 19%。第一季總支出為 248 億美元，比去年同期成長 9%。就具體項目而言，收入成本增加了 14%，主要原因是基礎設施成本和支付給合作夥伴的費用增加，但部分被先前宣布的延長伺服器使用壽命帶來的好處所抵消。

R&D increased 22%, mostly due to higher employee compensation and infrastructure costs. Marketing and sales increased 8% driven mainly by an increase in professional services related to our ongoing platform integrity efforts. G&A decreased 34%, driven primarily by lower legal-related costs. We ended Q1 with over 76,800 employees, up 4% quarter-over-quarter. First quarter operating income was $17.6 billion, representing a 41% operating margin.

研發費用增加了 22%，主要原因是員工薪資和基礎設施成本增加。行銷和銷售額成長了 8%，這主要得益於與我們持續的平台完整性努力相關的專業服務的成長。一般及行政費用下降 34%，主要因為法律相關成本下降。截至第一季末，我們擁有超過 76,800 名員工，季增 4%。第一季營業收入為 176 億美元，營業利益率為 41%。

Our tax rate for the quarter was 9% as we recognized excess tax benefits from share-based compensation due to the increase in our share price versus prior periods. Net (technical difficulty) expenditures, including principal payments on finance leases were $13.7 billion, driven by investments in servers, data centers and network infrastructure.

本季我們的稅率為 9%，因為我們的股價較前期有所上漲，從而確認了基於股票的薪酬的超額稅收優惠。淨（技術難度）支出（包括融資租賃的本金支付）為 137 億美元，主要由對伺服器、資料中心和網路基礎設施的投資推動。

Free cash flow was $10.3 billion. We repurchased $13.4 billion of our Class A common stock and paid $1.3 billion in dividends to shareholders, ending the quarter with $70.2 billion in cash and marketable securities and $28.8 billion in debt.

自由現金流為103億美元。我們回購了價值 134 億美元的 A 類普通股，並向股東支付了 13 億美元的股息，本季末我們持有 702 億美元的現金和有價證券以及 288 億美元的債務。

Moving now to our segment results. I'll begin with our Family of Apps segment. Our community across the Family of Apps continues to grow and we estimate more than 3.4 billion people used at least one of our Family of Apps on a daily basis in March. Q1 total Family of Apps revenue was $41.9 billion, up 16% year-over-year. Q1 Family of Apps ad revenue was $41.4 billion up 16% or 20% on a constant currency basis.

現在轉到我們的分部業績。我將從我們的應用程式系列部分開始。我們的應用程式系列社群正在不斷壯大，我們估計，3 月每天有超過 34 億人至少使用我們的一款應用程式系列。第一季應用程式系列總營收為 419 億美元，年增 16%。Q1 應用程式系列廣告收入為 414 億美元，以固定匯率計算成長 16% 或 20%。

Within ad revenue, the online commerce vertical was the largest contributor to year-over-year growth. On a user geography basis, ad revenue growth was strongest in Rest of World and North America at 19% and 18%, respectively. Europe and Asia Pacific grew 14% and 12%. In Q1, the total number of ad impressions served across our services increased 5%, and the average price per ad increased 10%. Impression growth was mainly driven by Asia Pacific.

在廣告收入中，線上商務垂直領域是年成長的最大貢獻者。從用戶地理來看，全球其他地區和北美的廣告收入成長最為強勁，分別為 19% 和 18%。歐洲和亞太地區分別成長了14%和12%。第一季度，我們各項服務的廣告展示總數增加了 5%，每個廣告的平均價格上漲了 10%。印象成長主要受到亞太地區的推動。

Pricing growth benefited from increased advertiser demand, in part driven by improved ad performance. This was partially offset by impression growth, particularly from lower monetizing regions and surfaces. Family of Apps other revenue was $510 million, up 34%, driven mostly by business messaging revenue growth from our WhatsApp business platform as well as Meta verified subscriptions. We continue to direct the majority of our investments toward the development and operation of our Family of Apps. In Q1, Family of Apps expenses were $20.1 billion, representing 81% of our overall expenses.

定價成長受惠於廣告商需求的增加，部分原因是廣告效果的提升。這被印象增長部分抵消，特別是來自貨幣化較低的地區和表面的印象增長。應用程式系列的其他收入為 5.1 億美元，成長 34%，主要得益於我們的 WhatsApp 商業平台的商業訊息收入成長以及 Meta 驗證訂閱。我們繼續將大部分投資用於應用程式系列的開發和營運。第一季度，應用程式系列的支出為 201 億美元，占我們整體支出的 81%。

Family of Apps expenses were up 10% mainly due to growth in employee compensation and infrastructure costs, which were partially offset by lower legal-related expenses. Family of Apps operating income was $21.8 billion, representing a 52% operating margin.

應用程式系列費用上漲 10%，主要原因是員工薪酬和基礎設施成本的成長，但法律相關費用的降低部分抵消了這一增長。應用程式系列營運收入為 218 億美元，營運利潤率為 52%。

Within our Reality Labs segment, Q1 revenue was $412 million, down 6% year-over-year due to lower Meta Quest sales, which were partially offset by increased sales of Ray-Ban Meta AI glasses. Reality Labs expenses were $4.6 billion, up 8% year-over-year, driven primarily by higher employee compensation. Reality Labs operating loss was $4.2 billion.

在我們的現實實驗室部門，第一季營收為 4.12 億美元，年減 6%，原因是 Meta Quest 銷量下降，但雷朋 Meta AI 眼鏡銷量成長部分抵消了這一影響。Reality Labs 的支出為 46 億美元，較去年同期成長 8%，主要原因是員工薪資增加。Reality Labs 的營運虧損為 42 億美元。

Turning now to the business outlook. There are two primary factors that drive our revenue performance, our ability to deliver engaging experiences for our community and our effectiveness at monetizing that engagement over time. On the first, we're focused both on enhancing our core Family of Apps today and building the next generation of devices and experiences through Reality Labs. I'll start with our Family of Apps. In the first quarter, we saw strong growth in video consumption across both Facebook and Instagram, particularly in the US, where video time spent grew double digits year-over-year.

現在談談商業前景。推動我們收入表現的主要因素有兩個：我們為社區提供引人入勝的體驗的能力，以及我們隨著時間的推移將這種參與貨幣化的有效性。首先，我們專注於增強當前的核心應用系列，並透過現實實驗室建立下一代設備和體驗。我將從我們的應用程式系列開始。第一季度，我們看到 Facebook 和 Instagram 上的影片消費強勁成長，尤其是在美國，影片消費時間比去年同期成長了兩位數。

This growth continues to be driven primarily by ongoing enhancements to our recommendation systems, and we see opportunities to deliver further gains this year. We're also progressing on longer-term efforts to develop innovative new approaches to recommendations. A big focus of this work will be on developing increasingly efficient recommendation systems so that we can continue scaling up the complexity and compute used to train our models while avoiding diminishing returns. There are promising techniques we're working on that will incorporate the innovations from LLM model architectures to achieve this.

這一成長主要得益於我們推薦系統的持續增強，我們看到了今年進一步成長的機會。我們也在長期努力中取得進展，以開發創新的新建議方法。這項工作的重點是開發越來越有效率的推薦系統，以便我們能夠繼續擴大用於訓練模型的複雜性和運算能力，同時避免效益遞減。我們正在研究一些有前景的技術，這些技術將結合 LLM 模型架構的創新來實現這一目標。

Another area that is showing early promise is integrating LLM technology into our content recommendation systems. For example, we're finding that LLM's ability to understand a piece of content more deeply than traditional recommendation systems, can help better identify what is interesting to someone about a piece of content, leading to better recommendations. We began testing using Llama in Threads recommendation systems at the end of last year given the app's text-based content and have already seen a 4% lift in time spent from the first launch.

另一個顯示早期前景的領域是將 LLM 技術整合到我們的內容推薦系統中。例如，我們發現 LLM 能夠比傳統推薦系統更深入地理解內容，可以幫助更好地識別人們對內容的興趣，從而提供更好的推薦。考慮到該應用的基於文字的內容，我們從去年年底開始在 Threads 推薦系統中使用 Llama 進行測試，並且已經看到自首次啟動以來的使用時間增加了 4%。

It remains early here, but a big focus this year will be on exploring how we can deploy this for other content types, including photos and videos. We also expect this to be complementary to Meta AI as it can provide more relevant responses to people's queries by better understanding their interests and preferences through their interactions across Facebook, Instagram and Threads.

現在還為時過早，但今年的重點將放在探索如何將其部署到其他內容類型，包括照片和影片。我們也希望這與 Meta AI 相輔相成，因為它可以透過人們在 Facebook、Instagram 和 Threads 上的互動來更好地了解他們的興趣和偏好，從而為人們的查詢提供更相關的回應。

Earlier this year, we began testing the ability for Meta AI to better personalize its responses by remembering certain details from people's prior queries and considering what that person engages with on our apps. We are already seeing this led to deeper engagement with people we've rolled it out to. And it is now built into Meta AI across Facebook, Instagram, Messenger and our new stand-alone Meta AI app in the US and Canada. We're also continuing to focus on helping people connect over content. In Q1, we launched a new experience on Instagram in the US that consists of a feed of content your friends have left a note on or light, and we're seeing good results.

今年早些時候，我們開始測試 Meta AI 的能力，透過記住人們先前查詢的某些細節並考慮該人在我們的應用程式上進行的操作，使其能夠更好地個性化其回應。我們已經看到，這促進了我們與推廣對象的更深層次的互動。現在，它已內建在 Facebook、Instagram、Messenger 以及我們在美國和加拿大推出的全新獨立 Meta AI 應用程式中。我們也將繼續致力於幫助人們透過內容建立聯繫。在第一季度，我們在美國 Instagram 上推出了一項新體驗，其中包括好友留言或點亮的內容，我們看到了良好的結果。

We also just launched Blend, which is an opt-in experience in direct messages that enables you to blend your Reels' algorithm with your friends to spark conversations over each other's interest. These features all lean into Instagram's position at the intersection of entertainment and social connection. WhatsApp remains at its core, a private messaging app, but it has evolved to also become a place people come to get updates from accounts they are connected to or follow.

我們也剛推出了 Blend，這是一種可選擇的直接訊息體驗，可讓您將自己的 Reels 演算法與朋友的演算法融合，從而激發彼此感興趣的對話。這些功能都有利於 Instagram 在娛樂和社交聯繫的交匯處佔據一席之地。WhatsApp 的核心仍然是一個私人訊息應用程序，但它已經發展成為人們從他們所連接或關注的帳戶獲取更新的地方。

Today, there are tens of billions of views of status posts on WhatsApp each day, and we continue to invest in the updates tab as a place people can go to do more. Creators remain another big focus for us and we're investing in tools to help them produce the best original content on our platforms. Last week, we launched our stand-alone Edits app, which supports the full creative process for video creators from inspiration and creation to performance insights. Edits has an ultra-high resolution, short-form video camera and includes generative AI tools that enable people to remove the background of any video or animate still images with more features coming soon.

如今，WhatsApp 上的狀態貼文每天有數百億次瀏覽量，我們將繼續投資更新標籤，讓人們可以在這裡做更多的事情。創作者仍然是我們關注的另一個重點，我們正在投資工具來幫助他們在我們的平台上製作最好的原創內容。上週，我們推出了獨立的 Edits 應用程序，它支援影片創作者從靈感、創作到性能洞察的完整創作過程。Edits 擁有超高解析度的短影片相機，並包含生成式 AI 工具，使人們能夠刪除任何影片的背景或為靜態圖像製作動畫，更多功能即將推出。

Moving to Reality Labs. We're seeing very strong traction with Ray-Ban Meta AI glasses, with over 4x as many monthly actives as a year ago, and the number of people using voice commands is growing even faster as people use it to answer questions and control their glasses. This month, we fully rolled out live translations on Ray-Ban Meta AI glasses to all markets for English, French, Italian and Spanish.

轉向現實實驗室。我們看到雷朋 Meta AI 眼鏡的吸引力非常大，其每月活躍用戶數量是去年的 4 倍多，而且使用語音命令的人數增長速度更快，因為人們用它來回答問題和控制眼鏡。本月，我們向所有市場全面推出了雷朋 Meta AI 眼鏡的英語、法語、義大利語和西班牙語即時翻譯功能。

Now when you are speaking to someone in one of these languages, you'll hear what they say in your preferred language through the glasses in real time. Now to the second driver of our revenue performance, increasing monetization efficiency. The first part of this work is optimizing the level of ads within organic engagement. We continue to optimize ad supply across each service to better deliver ads at the time and place they are most relevant to people. We are also starting to introduce ads on unmonetized surfaces like Threads, which we opened up to all eligible advertisers this month to reach people in over 30 different markets to start including the US.

現在，當您用其中一種語言與某人交談時，您將透過眼鏡即時聽到他們用您喜歡的語言說的話。現在談談我們收入表現的第二個驅動力，即提高貨幣化效率。這項工作的第一部分是優化有機參與中的廣告水平。我們持續優化每項服務的廣告供應，以便在與人們最相關的時間和地點更好地投放廣告。我們也開始在 Threads 等非貨幣化平台上推出廣告，本月我們向所有符合條件的廣告商開放了這些平台，以覆蓋包括美國在內的 30 多個不同市場的人們。

As we do for any newly monetized surface, we expect to gradually ramp ad supply as we optimize the ad formats and ensure they feel native to the app. We don't expect Threads to be a meaningful driver of overall impression or revenue growth in 2025.

正如我們對任何新貨幣化平台所做的那樣，我們希望在優化廣告格式並確保其適合應用程式的同時，逐步增加廣告供應。我們預期 Threads 不會成為 2025 年整體印像或營收成長的重要推手。

The second part of increasing monetization efficiency is improving marketing performance. We're continuing to improve our ad systems by developing new modeling technologies to more efficiently predict the right ad to show. In Q1, we introduced our new generative ads recommendation model, or GEM, for ads ranking. This model uses a new architecture we developed that is twice as efficient at improving ad performance for a given amount of data and compute. This efficiency gain enabled us to significantly scale up the amount of compute we use for model training, with GEM trained on thousands of GPUs, our largest cluster for ads training to date.

提升變現效率的第二部分是提升行銷績效。我們正在透過開發新的建模技術來繼續改進我們的廣告系統，以便更有效地預測要顯示的正確廣告。在第一季度，我們推出了新的生成廣告推薦模式（GEM），用於廣告排名。該模型採用了我們開發的新架構，在給定資料量和計算量的情況下，其廣告效果的提升效率提高了一倍。這種效率的提升使我們能夠大幅擴大用於模型訓練的運算量，GEM 在數千個 GPU 上進行訓練，這是我們迄今為止最大的廣告訓練群集。

We began testing the new model for ads recommendations on Facebook Reels earlier this year and have seen up to a 5% increase in ad conversions. We're now rolling it out to additional services across our apps.

我們今年稍早開始在 Facebook Reels 上測試新的廣告推薦模型，廣告轉換率提高了 5%。我們現在正在將其推廣到我們應用程式的其他服務中。

On the ads product side, we're seeing continued momentum with our Advantage suite of AI-powered solutions. We've been encouraged by the initial test of our streamlined campaign creation flow for sales, app and lead campaigns, which starts with Advantage+ turned on from the beginning for advertisers. In April, we rolled this out to more advertisers and expect to complete the global rollout later this year. We're also seeing strong adoption of Advantage+ Creative.

在廣告產品方面，我們看到 Advantage 人工智慧解決方案套件持續保持強勁發展勢頭。我們對銷售、應用程式和潛在客戶行銷活動的簡化行銷活動創建流程的初步測試感到鼓舞，該流程從一開始就為廣告商啟用 Advantage+。今年 4 月，我們向更多廣告商推出了這項服務，預計今年稍後完成全球推廣。我們也看到 Advantage+ Creative 的廣泛採用。

This week, we are broadening access of video expansion to Facebook Reels for all eligible advertisers, enabling them to automatically adjust the aspect ratio of their existing videos by generating new pixels in each frame to optimize their ads for full screen surfaces. We also rolled out image generation to all eligible advertisers.

本週，我們將向所有符合條件的廣告商開放 Facebook Reels 視訊擴充存取權限，使他們能夠透過在每個畫面中產生新像素來自動調整現有影片的寬高比，從而優化其廣告以適應全螢幕表面。我們也向所有符合條件的廣告主推出了圖像生成功能。

And this quarter, we plan to continue testing a new virtual try-on feature that uses Gen AI to place clothing on virtual models, helping customers visualize how an item may look and fit. Last, we continue to evolve our ads platform to drive results that are optimized for each business' objectives and the way they measure value. One example of this is our incremental attribution feature, which enables advertisers to optimize for driving incremental conversions or conversions we believe would not have occurred without an ad being shown.

本季度，我們計劃繼續測試一項新的虛擬試穿功能，該功能使用 Gen AI 將衣服穿在虛擬模型上，幫助顧客直觀地看到衣服的外觀和穿著效果。最後，我們將繼續改進我們的廣告平台，以推動針對每個企業的目標和衡量價值的方式進行最佳化的結果。其中一個例子就是我們的增量歸因功能，它使廣告主能夠進行最佳化以推動增量轉換或我們認為在沒有展示廣告的情況下就不會發生的轉換。

We're seeing strong results in testing so far, with advertisers using incremental attribution in tests, seeing an average 46% lift in incremental conversions compared to their business as usual approach. We expect to make this available to all advertisers in the coming weeks.

到目前為止，我們在測試中看到了強勁的結果，廣告主在測試中使用增量歸因，與常規業務方法相比，增量轉換率平均提升了 46%。我們預計將在未來幾週內向所有廣告商提供此功能。

Next, I would like to discuss our approach to capital allocation. Our primary focus remains investing capital back into the business with infrastructure and talent being our top priorities. Starting with headcount, our hiring continues to be targeted at technical roles within our company priorities. In the first quarter, the significant majority of the roughly 2,800 employees we added were to support our priorities of monetization, infrastructure, generative AI, regulation and compliance and Reality Labs. On infrastructure, we have two primary focuses to meet the growing compute needs of our services and AI initiatives.

接下來我想討論一下我們的資本配置方法。我們的首要重點仍然是將資本重新投入到業務中，而基礎設施和人才是我們的首要任務。從員工人數開始，我們的招募繼續以公司優先考慮的技術職位為目標。在第一季度，我們新增了大約 2,800 名員工，其中絕大多數是為了支援我們的貨幣化、基礎設施、生成式人工智慧、監管與合規以及現實實驗室等優先事項。在基礎設施方面，我們主要關注兩個方面，以滿足我們的服務和人工智慧計劃日益增長的運算需求。

The first way is by significantly scaling up our infrastructure footprint. Our CapEx growth this year is going towards both generative AI and core business needs with the majority of overall CapEx supporting the core. We expect the significant infrastructure footprint we are building will not only help us meet the demands of our business in the near term but also provide us an advantage in the quality and scale of AI services we can deliver. We continue to build this capacity in a way that grants us maximum flexibility in how and when we deploy it to ensure we have the agility to react to how the technology and industry develop in the coming years.

第一種方式是大幅擴大我們的基礎設施覆蓋範圍。我們今年的資本支出成長將用於產生人工智慧和核心業務需求，其中大部分資本支出用於支援核心業務。我們預計，我們正在建造的重要基礎設施不僅將幫助我們在短期內滿足業務需求，而且還將在我們提供的人工智慧服務的品質和規模方面為我們提供優勢。我們將繼續以一種賦予我們在部署方式和時間方面最大靈活性的方式來建立這種能力，以確保我們能夠靈活地應對未來幾年技術和行業的發展。

The second way we're meeting our compute needs is by increasing the efficiency of our workloads. In fact, many of the innovations coming out of our ranking work are focused on increasing the efficiency of our systems. This emphasis on efficiency is helping us deliver consistently strong returns from our core AI initiatives.

我們滿足計算需求的第二種方法是提高工作負載的效率。事實上，我們的排名工作產生的許多創新都集中在提高系統效率上。這種對效率的重視有助於我們從核心人工智慧計畫中獲得持續強勁的回報。

For example, we shared on the Q3 2024 call that improvements to our AI-driven feed and video recommendations drove a roughly 8% lift in time spent on Facebook and and a 6% lift on Instagram over the first 9 months of last year. Since then, we've been able to deliver similar gains in just six months' time, with improvements to our AI recommendations, delivering 7% and 6% time spent gains on Facebook and Instagram, respectively.

例如，我們在 2024 年第三季電話會議上分享了，我們人工智慧驅動的資訊流和影片推薦的改進使得去年前 9 個月在 Facebook 上花費的時間增加了約 8%，在 Instagram 上花費的時間增加了 6%。從那時起，我們僅用六個月的時間就實現了類似的收益，並改進了我們的人工智慧推薦，分別使 Facebook 和 Instagram 上的使用時間增加了 7% 和 6%。

Before moving to our financial guidance, I want to acknowledge the dynamic macro environment and note that our range reflects the potential for a wider set of outcomes. We continue to feel good about the fundamental drivers of revenue growth and believe the past work we've done to streamline our operations and cost profile puts us in a strong position to navigate a variety of outcomes.

在談到我們的財務指導之前，我想先了解動態的宏觀環境，並指出我們的範圍反映了更廣泛結果的潛力。我們繼續對收入成長的基本驅動力感到樂觀，並相信我們過去為簡化營運和成本狀況所做的工作使我們處於有利地位，可以應對各種結果。

Moving to our financial outlook. We expect second quarter 2025 total revenue to be in the range of $42.5 billion to $45.5 billion. Our guidance assumes foreign currency is an approximately 1% tailwind to year-over-year total revenue growth based on current exchange rates.

轉向我們的財務展望。我們預計 2025 年第二季總營收將在 425 億至 455 億美元之間。我們的指導假設是，根據當前匯率，外幣將為年比總收入成長帶來約 1% 的順風。

Turning now to the expense outlook. We expect full year 2025 total expenses to be in the range of $113 billion to $118 billion, lowered from our prior outlook of $114 billion to $119 billion.

現在來談談費用前景。我們預計 2025 年全年總支出將在 1,130 億美元至 1,180 億美元之間，低於先前預測的 1,140 億美元至 1,190 億美元。

Turning now to the CapEx outlook. We anticipate our full year 2025 capital expenditures, including principal payments on finance leases will be in the range of $64 billion to $72 billion, increased from our prior outlook of $60 billion to $65 billion. This updated outlook reflects additional data center investments to support our AI efforts as well as an increase in the expected cost of infrastructure hardware. The majority of our CapEx in 2025 will continue to be directed to our core business.

現在來談談資本支出前景。我們預計 2025 年全年資本支出（包括融資租賃的本金支付）將在 640 億美元至 720 億美元之間，高於我們先前預測的 600 億美元至 650 億美元。這項更新的展望反映了為支援我們的人工智慧工作而增加的資料中心投資以及基礎設施硬體預期成本的增加。我們 2025 年的大部分資本支出將繼續用於我們的核心業務。

On to tax. Absent any changes to our tax landscape, we expect our full year 2025 tax rate to be in the range of 12% to 15%. In addition, we continue to monitor an active regular landscape, including legal and regulatory headwinds in the EU and the US that could significantly impact our business and our financial results. The European Commission recently announced its decision that our subscription for no ads model is not compliant with the DMA.

談稅。如果我們的稅收狀況沒有任何變化，我們預計 2025 年全年稅率將在 12% 至 15% 之間。此外，我們將繼續關注活躍的常規環境，包括可能對我們的業務和財務表現產生重大影響的歐盟和美國的法律和監管阻力。歐盟委員會最近宣布其決定，認為我們的無廣告訂閱模式不符合 DMA。

Based on feedback from the European Commission in connection with the DMA, we expect we will need to make some modifications to our model, which could result in a materially worse user experience for European users, and a significant impact to our European business and revenue as early as the third quarter of 2025. We will appeal the commission's DMA decision, but any modifications to our model may be imposed before or during the appeal process. In closing, this was another solid quarter for our business. We believe the investments we're making across our company priorities will position us well in the coming years to continue delivering engaging services for our community, compelling results for advertisers and strong business performance.

根據歐盟委員會對 DMA 的回饋，我們預計需要對我們的模型進行一些修改，這可能會導致歐洲用戶的用戶體驗大幅下降，並對我們最早在 2025 年第三季的歐洲業務和收入產生重大影響。我們將對委員會的 DMA 決定提出上訴，但對我們的模型的任何修改都可能在上訴過程之前或期間實施。總而言之，這對我們的業務來說又是一個穩健的季度。我們相信，我們在公司優先事項上所做的投資將使我們在未來幾年處於有利地位，繼續為我們的社區提供引人入勝的服務，為廣告商提供令人信服的成果並實現強勁的業務表現。

With that, Krista, let's open up the call for questions.

克里斯塔，讓我們開始提問吧。

(Operator Instructions) Brian Nowak, Morgan Stanley.

（操作員指示）摩根士丹利的 Brian Nowak。

I have two. The first one is on Llama. Mark, can you -- the LLM landscape continues to sort of evolve and be somewhat competitive. Can you sort of talk us through some of the key areas of advancement you are most focused on and excited about as we sort of think about Behemoth and next versions of Llama to come? And then the second one on Meta AI, almost 1 billion users globally. Any help on sort of how you're seeing US traction there and the types of recurring user behaviors that you're seeing in the early Meta AI use cases?

我有兩個。第一個是關於 Llama 的。馬克，你能嗎——法學碩士學位領域正在繼續發展，並且競爭越來越激烈。當我們考慮 Behemoth 和即將推出的 Llama 的下一個版本時，您能否向我們介紹您最關注和最興奮的一些關鍵進步領域？第二個是關於 Meta AI 的，全球有近 10 億用戶。您能談談您如何看待美國在該領域的發展勢頭，以及在早期 Meta AI 用例中看到的重複使用者行為類型嗎？

Sure. I can talk about the LLMs. On the Meta AI usage, I'm not sure if we have more stats to share on that now. Yes, it's -- I'll defer to Susan on if there's anything that we're ready on that.

當然。我可以談論法學碩士學位。關於 Meta AI 的使用，我不確定我們現在是否有更多統計數據可以分享。是的——如果我們準備好了什麼的話，我會向蘇珊報告。

On the LLM, yes, there's a lot of progress being made in a lot of different dimensions. And the reason why we want to build this out is one, is that we think it's important that for kind of how critical this is for our business that we sort of have control of our own destiny and are not depending on another company for something so critical. But two, we want to make sure that we can shape the development to be optimized for our infrastructure and the use cases that we want.

是的，就法學碩士學位而言，我們在許多不同方面都取得了巨大進展。我們想要建立這個體系的原因之一是，我們認為這對我們的業務至關重要，我們必須掌控自己的命運，而不是依賴其他公司來處理如此重要的事情。其次，我們要確保我們能夠優化開發，以適應我們的基礎設施和我們想要的用例。

So to that end, Llama 4, the shape of the model with 17 billion parameters per expert was designed specifically for the infrastructure that we have in order to provide the low latency experience to be voice optimized. One of the key things, if you're having a voice conversation with AI is it needs to be low latencies. That way when you're having a conversation with it, there's isn't a large gap between when you stop speaking and it starts.

因此，為此，Llama 4，每位專家具有 170 億個參數的模型形狀是專門為我們現有的基礎設施設計的，以提供低延遲的語音優化體驗。如果你要與人工智慧進行語音對話，那麼關鍵的一點就是延遲要低。這樣，當你與它交談時，你停止說話和它開始說話之間就不會存在很大的時間間隔。

So everything from the shape of the model to the research that we're doing to techniques that go into it are kind of fit into that. Similarly, another thing that we focused on was context window length. And in some of our models, we have really -- we're industry-leading on context window length.

因此，從模型的形狀到我們正在進行的研究，再到其中所採用的技術，一切都與此相適應。同樣，我們關注的另一件事是上下文視窗長度。在我們的一些模型中，我們在上下文視窗長度方面確實處於業界領先地位。

And part of the reason why we think that, that's important is because we're very focused on providing a personalized experience. And there are different ways that you can put personalization context into an LLM, but one of the ways to do it is to include some of that context in the context window. And having a long context window that can incorporate a lot of the background that the person has shared across our apps is one way to do that. So that's like -- it kind of is giving you a flavor of the products that we're trying to build and then some specific technical architecture decisions and research prioritization that we basically have made in order to deliver the specific experience that we're going for.

我們之所以認為這一點很重要，部分原因在於我們非常注重提供個人化的體驗。您可以透過不同的方式將個人化上下文放入 LLM，但其中一種方法是將部分上下文包含在上下文視窗中。擁有一個較長的上下文窗口，可以整合用戶在我們的應用程式上分享的大量背景信息，這是實現這一目標的方法之一。所以這就像是——它讓你了解我們正在嘗試建立的產品，然後是一些特定的技術架構決策和研究優先級，這些基本上是我們為了提供我們想要的特定體驗而做出的。

I could go on and add a lot more. The reason I think it's also very important to deliver big models like Behemoth, not because we're going to end up serving them in production, but because of the technique of distilling from larger models, right. The Llama 4 models that we've published so far and the ones that we're using internally and some of the ones that we'll build in the future, are basically distilled from the behemoth model in order to get the 90%, 95% of the intelligence of the large model in a form factor that is much lower latency and much more efficient. So these things are all very important. Obviously, we wouldn't be able to do that kind of distillation from other closed models.

我還可以繼續添加更多內容。我認為提供像 Behemoth 這樣的大型模型也非常重要，這並不是因為我們最終會在生產中使用它們，而是因為從更大的模型中進行提煉的技術，對吧。我們迄今為止發布的 Llama 4 模型、我們內部正在使用的模型以及我們將來會構建的一些模型，基本上都是從龐然大物模型中提煉出來的，目的是以延遲更低、效率更高的形式獲得大型模型 90%、95% 的智能。所以這些事情都非常重要。顯然，我們無法從其他封閉模型中進行這種提煉。

So that kind of gives you a flavor for how we're thinking about the development of this and then, of course, the models and the infrastructure that we're building out power all of the opportunities that I mentioned before.

因此，這可以讓您了解我們對此發展的看法，當然，我們正在建造的模型和基礎設施為之前提到的所有機會提供動力。

Brian, I'm happy to answer your second question about Meta AI. The top use case right now for Meta AI from a query perspective is really around information gathering as people are using it to search for and understand and analyze information followed by social interactions from, ranging from casual chatting to more in-depth discussion or debate. We also see people use it for writing assistance, interacting with visual content, seeking help. And we see Meta -- people engage with Meta AI from several different entry points. WhatsApp continues to see the strongest Meta AI usage across our Family of Apps.

布萊恩，我很高興回答你關於 Meta AI 的第二個問題。從查詢角度來看，Meta AI 目前的主要用例實際上是圍繞著資訊收集，因為人們使用它來搜尋、理解和分析訊息，然後進行社交互動，從隨意聊天到更深入的討論或辯論。我們也看到人們使用它來輔助寫作、與視覺內容互動、尋求幫助。我們看到 Meta——人們從幾個不同的切入點與 Meta AI 互動。WhatsApp 在我們的應用程式系列中繼續見證著最強大的 Meta AI 使用率。

Most of that, WhatsApp engagement is in one-on-one threads, followed by Facebook, which is the second largest driver of Meta AI engagement, where we're seeing strong engagement from our feed deep dives integration that lets people ask Meta AI questions about the content that's recommended to them. And we're obviously excited about the launch of the Meta AI stand-alone app.

其中，WhatsApp 的參與大多是一對一的，其次是 Facebook，它是 Meta AI 參與的第二大驅動力，我們從 feed 深度挖掘整合中看到了強大的參與度，它允許人們向 Meta AI 詢問有關推薦給他們的內容的問題。我們對 Meta AI 獨立應用程式的推出感到非常興奮。

Eric Sheridan, Goldman Sachs.

高盛的艾瑞克·謝裡丹。

Maybe following up on Brian's question and coming at it from a different angle. I appreciate the color on the use cases you're seeing today for Meta AI. How would you suspect those use cases evolve with a stand-alone app. Can you bring us into a little bit the decision process to do a stand-alone app, what that might change in terms of utility, frequency or scale relative to what you see inside Family of Apps today? And how do you think about positioning Meta AI as a stand-alone app, I guess, the competitive landscape today of other stand-alone sort of consumer AI apps.

也許可以繼續回答布萊恩的問題，並從不同的角度來回答它。我很欣賞您今天看到的 Meta AI 用例的豐富色彩。您認為這些用例會如何隨著獨立應用程式的發展而演變。您能否向我們稍微介紹一下開發獨立應用程式的決策過程？與您目前在應用程式系列中看到的情況相比，它在實用性、頻率或規模方面可能發生哪些變化？您如何看待將 Meta AI 定位為獨立應用程序，我想，當今其他獨立消費者 AI 應用程式的競爭格局是怎樣的。

Yes, I can talk about that. We're going to focus on both integrating it into our Family of Apps in more ways and building a stand-alone experience. I think some people want faster access to it or a more built out feature set than you can build into an app like WhatsApp. So the stand-alone app will be valuable for that. I also think the stand-alone app is going to be particularly important in the United States because WhatsApp, as Susan said, is the largest surface that people use Meta AI in which makes sense if you want to text an AI having that be closely integrated and a good experience in the messaging app that you use makes a lot of sense.

是的，我可以談論這個。我們將專注於以更多方式將其整合到我們的應用程式系列中並建立獨立的體驗。我認為有些人希望更快地訪問它或獲得比 WhatsApp 等應用程式內建的更豐富的功能集。因此，獨立的應用程式將會很有價值。我還認為獨立應用程式在美國將特別重要，因為正如蘇珊所說，WhatsApp 是人們使用元人工智慧的最大平台，如果你想要給人工智慧發送短信，那麼將其緊密整合並在你使用的通訊應用程式中獲得良好的體驗是非常有意義的。

But we're -- while we have more than 100 million people use WhatsApp in the United States, we're clearly not the primary messaging app in the United States at this point, iMessages. We hope to become the leader over time. But we're in a different position there than we are in most of the rest of the world on WhatsApp. So I think that the Meta AI app as a stand-alone is going to be particularly important in the United States to establishing leadership in -- as the main personal AI that people use. But we're going to keep on advancing the experiences across the board in all of these different areas.

但是，雖然美國有超過 1 億人使用 WhatsApp，但目前我們顯然不是美國的主要訊息應用程式 iMessages。我們希望隨著時間的推移成為領導者。但我們在那裡的地位與 WhatsApp 在世界其他大部分地區的地位不同。因此，我認為 Meta AI 應用程式作為獨立應用程式將在美國發揮特別重要的作用，確立其作為人們使用的主要個人人工智慧的領導地位。但我們將繼續在所有這些不同領域全面提升經驗。

Justin Post, Bank of America.

美國銀行的賈斯汀·波斯特 (Justin Post)。

A couple of questions. Just on the guide in the second quarter, reports of potential supply issues in e-commerce, how you thought about that in the guide and maybe how you're thinking about it for the back half. And then on a bigger picture question. Your CapEx spend is now close to some hyperscalers with very big client bases. Just help us conceptualize the kind of ecosystem you're building with your CapEx. I know you gave a lot of help on the intro, but maybe the ROI works without direct enterprise spend to drive revenues. How you're thinking about that?

有幾個問題。僅在第二季的指南中，就有關於電子商務中潛在供應問題的報告，您在指南中是如何看待這個問題的，以及您對下半年是如何考慮這個問題的。然後討論一個更大的問題。您的資本支出現已接近一些擁有龐大客戶群的超大規模企業。只需幫助我們概念化您正在使用資本支出建立的生態系統類型。我知道您在介紹中給予了很多幫助，但也許投資回報率不需要直接的企業支出來推動收入。您對此有何看法？

Thanks, Justin. On the Q2 guide, there's uncertainty, obviously, in how the macro environment will evolve over time and how that could impact different segments of our business. Our Q2 revenue outlook aims to factor that in. And partly -- that's partly why the $3 billion range reflects the potential for a wider range of outcomes.

謝謝，賈斯汀。對於第二季的指引，顯然存在不確定性，包括宏觀環境將如何隨時間演變，以及這將如何影響我們業務的不同部分。我們的第二季營收展望旨在將此因素納入考量。這也是為什麼 30 億美元的範圍反映了更廣泛的結果的可能性。

Specifically, we have seen some reduced spend in the US from Asia-based e-commerce exporters, which we believe is in anticipation of the de minimis exemption going away on May 2. A portion of that spend has been redirected to other markets but overall spend for those advertisers is below the levels prior to April. But our Q2 outlook reflects the trends we're seeing so far in April, which have generally been healthy. So it's very early. Hard to know how things will play out over the quarter and certainly harder to know that for the rest of the year.

具體來說，我們看到亞洲電子商務出口商在美國的支出有所減少，我們認為這是由於預期最低限度豁免將於 5 月 2 日取消。其中一部分支出已轉移到其他市場，但這些廣告商的總體支出低於 4 月之前的水平。但我們對第二季度的展望反映了四月份迄今為止所看到的趨勢，總體而言是健康的。所以現在還很早。很難知道本季的情況將如何發展，當然更難知道今年剩餘時間的情況將如何發展。

Your second question is about why we're investing more in CapEx. And we really believe that our ability to build world-class infrastructure gives us a meaningful advantage in both developing the leading AI technology and services over the coming years. And there are a lot of opportunities also for us to improve our core business by putting more compute against our ads and recommendation work.

您的第二個問題是為什麼我們要加大資本支出。我們堅信，我們建立世界一流基礎設施的能力將為我們在未來幾年開發領先的人工智慧技術和服務方面帶來顯著優勢。我們還有很多機會透過在廣告和推薦工作中投入更多計算來改善我們的核心業務。

So even with the capacity that we're bringing online in 2025, we are having a hard time meeting the demand that teams have for compute resources across the company. So we are going to continually invest meaningfully here across our infrastructure footprint, but we are also really looking to build this capacity in a way that gives us the maximum flexibility in how and when we deploy it over the coming years, so we can respond to how the market and technology develop.

因此，即使我們在 2025 年實現了上線容量，我們也很難滿足整個公司團隊對運算資源的需求。因此，我們將繼續在我們的基礎設施領域進行有意義的投資，但我們也希望以一種方式來建立這種能力，以便我們在未來幾年內如何以及何時部署它方面擁有最大的靈活性，以便我們能夠應對市場和技術的發展。

Douglas Anmuth, JPMorgan.

摩根大通的道格拉斯·安穆斯。

I just wanted to follow up on CapEx and infrastructure spending. Just on the higher range for CapEx. Can you just help us understand how much of that is tied to the additional data center investments versus the increased hardware costs and really what's driving those higher hardware costs? And then separately, there have been some articles suggesting that you've been looking to partner to share some of the cost of the AI infrastructure build-out. Can you just help us understand your thought process there and some of the pros and cons of going alone versus partnering?

我只是想跟進資本支出和基礎設施支出。資本支出處於較高範圍。您能否幫助我們了解其中有多少與額外的資料中心投資和增加的硬體成本有關，以及真正導致硬體成本上升的原因是什麼？另外，有一些文章表明，你們一直在尋求合作夥伴來分擔人工智慧基礎設施建設的部分成本。您能否幫助我們了解您的思考過程以及單獨行動與合作行動的優缺點？

Thanks, Doug. So our increased CapEx outlook reflects both of those updates, the increased data center spend this year as we have made some adjustments to flex our build strategy that will enable us to really stand up capacity more quickly, both in '25 and '26. We haven't broken down sort of the exact drivers. The higher costs we expect to incur for infrastructure hardware this year really comes from suppliers who source from countries around the world. And there's just a lot of uncertainty around this, given the ongoing trade discussions.

謝謝，道格。因此，我們增加的資本支出前景反映了這兩項更新，即今年數據中心支出的增加，因為我們做出了一些調整來靈活調整我們的建設策略，這將使我們能夠在 25 年和 26 年更快地真正提高容量。我們還沒有分析出具體的驅動因素。我們預計今年基礎設施硬體成本的增加實際上來自於來自世界各國的供應商。鑑於正在進行的貿易談判，這方面存在許多不確定性。

And so that is both reflected in the wider range that we are giving. And we're also working on our end on mitigations by optimizing our supply chain, and our outlook is really trying to reflect our best understanding of the potential impact this year across all of that uncertainty.

因此，這反映在我們所提供的更廣泛的範圍內。我們也正在透過優化供應鏈來努力緩解影響，我們的展望實際上是在試圖反映我們對今年所有不確定性可能產生的影響的最佳理解。

On the second part of your question, we are -- we're pleased to have partners investing alongside us and bringing Llama to market like AWS and Azure who are helping us host Llama. We're always looking for opportunities to continue deepening or expanding those partnerships. But we are funding the infrastructure that is being used to train Llama, and we don't have any expectation that, that will change at this point.

關於您問題的第二部分，我們很高興有合作夥伴與我們一起投資並將 Llama 推向市場，例如 AWS 和 Azure 協助我們託管 Llama。我們一直在尋找機會繼續深化或擴大這些合作關係。但我們正在為訓練 Llama 的基礎設施提供資金，而且我們並不期望這種情況會就此改變。

Mark Shmulik, Bernstein.

馬克‧施穆里克，伯恩斯坦。

Mark, in your conversation last night with Satya, I think you both discussed a bit around kind of the portion of code being written internally by AI. Kind of back to some of your previous comments around this being the year where we might see AI kind of the place of a mid-level engineer. With the world evolving so quickly, can you share some places where you've seen strong traction there? And are we progressing kind of faster, slower or as you expected towards this milestone?

馬克，昨晚你和薩蒂亞的談話中，我想你們都討論了一些有關人工智慧內部編寫的程式碼部分。回到您之前的一些評論，今年我們可能會看到人工智慧取代中級工程師。世界發展如此之快，您能否分享一些您看到強勁發展的地區？我們在實現這一里程碑的過程中進展得更快、更慢還是如您預期的那樣？

And then, Susan, with the expense guidance coming down just to touch, how should we think about just the overall cadence of expected spending really as it relates to kind of core business performance and the realities of the day-to-day world we're living in?

然後，蘇珊，隨著支出指導的下降，我們應該如何看待預期支出的整體節奏，因為它與核心業務績效以及我們日常生活的現實有關？

I can talk about the coding agent work. I don't think that there's been any real change in our prediction for the timing of this. So I'd say it's basically still on track for something around a mid-level engineer kind of starting to become possible sometime this year, scaling into next year. So I'd expect that by the middle to end of next year, AI coding agents are going to be doing a substantial part of AI research and development. So we're focused on that.

我可以講一下編碼代理的工作。我認為我們對此時間的預測沒有任何實際改變。所以我想說，中級工程師的職位基本上仍在按計劃進行，今年某個時候開始成為可能，並擴展到明年。因此我預計到明年年中到年底，人工智慧編碼代理將在人工智慧研究和開發中發揮重要作用。所以我們專注於此。

Internally, we're also very focused on building AI agents or systems that can help run different experiments to increase recommendations across our other AI products like the ones that do recommendations across our feeds and things like that. So I think that if it works, should just accelerate our progress in those areas, that's the basic bet that we're making.

在內部，我們也非常注重建構人工智慧代理或系統，以幫助運行不同的實驗，以增加我們其他人工智慧產品的推薦，例如那些在我們的資訊流中進行推薦的產品等等。所以我認為，如果它有效的話，就應該加速我們在這些領域的進步，這是我們所做的基本賭注。

Your second question about our lowered expense outlook. Really, we are four months into the year. The lowered outlook reflects more refined forecasts, including updated expectations for both employee compensation as well as some other non-headcount-related operating expenses this year. And that's partially offset by higher expected infrastructure costs related to our increased CapEx outlook as well as higher expected Reality Labs cost of goods sold.

您的第二個問題是關於我們降低費用前景。確實，今年已經過去四個月了。下調的前景反映了更精細的預測，包括今年員工薪酬以及一些其他與員工人數無關的營運費用的最新預期。這部分被與我們增加的資本支出前景相關的更高的預期基礎設施成本以及更高的預期 Reality Labs 銷售成本所抵消。

And we've maintained our $5 billion range just given the more dynamic operating environment that we're in. And what I would say is our investment posture today reflects the significant opportunities that we see across either the company and priorities that we're investing in this year.

考慮到我們所處的更動態的營運環境，我們維持了 50 億美元的範圍。我想說的是，我們今天的投資態勢反映了我們在公司內部看到的重大機會以及我們今年投資的重點。

We will obviously continue evaluating depending on how macro conditions more broadly evolve. But we really feel like these are big strategic priorities for us and are critical for us to continue investing in. And in fact, I think one of the aims of our efficiency work over the last two years was to put us in a stronger financial position so that we can continue investing in key priorities through tougher financial cycles.

顯然，我們將根據宏觀條件的更廣泛發展繼續進行評估。但我們確實覺得這些對我們來說是重大的策略重點，並且對我們繼續投資至關重要。事實上，我認為我們過去兩年效率工作的目標之一是讓我們的財務狀況更加穩健，以便我們能夠在更艱難的金融週期中繼續對關鍵優先事項進行投資。

Ross Sandler, Barclays.

巴克萊銀行的羅斯桑德勒。

Great. Mark, yesterday in one of your many kind of podcast or keynote presentations, you mentioned that like a bunch of projects that your teams want to or aspire to do are kind of bottlenecked by the AI capacity, which Susan just talked about earlier and that even some of the testing that the ad ranking team wants to run is just getting kind of delayed. So I guess looking out either this year or next year or whatever, when do you kind of see some of this constraint being eased back?

偉大的。馬克，昨天，在您眾多的播客或主題演講中，您提到，您的團隊想要或渴望做的很多項目都受到了人工智能能力的瓶頸限制，蘇珊剛才也談到了這一點，甚至廣告排名團隊想要運行的一些測試也受到了某種程度的延遲。所以我想展望今年或明年或其他什麼年份，您什麼時候會認為這些限制會有所放寬？

And more broadly, we're kind of three years past the IDFA impact your business. So where do you think we are in terms of just the overall improvements to the ad ranking system, the ROI that you guys are able to deliver? And like what inning are we in on that in your opinion?

更廣泛地說，IDFA 對您的業務的影響已經過去三年了。那麼，您認為就廣告排名系統的整體改進以及您能夠提供的投資回報率而言，我們處於什麼位置？您認為我們目前處於什麼階段？

I can take a shot at both of those and Mark, you can obviously chime in. On the first question, the capacity landscape we are in is pretty dynamic, both in terms of the many moving parts in terms of us bringing capacity online, but also in terms of the demand from different product groups in our company, whether they are in the gen AI teams or whether they're doing more of the core AI work around ranking and recommendations. So both the supply and demand are quite fluid.

我可以嘗試這兩個，馬克，你顯然也可以加入。關於第一個問題，我們所處的產能格局非常動態，不僅體現在我們上線產能的許多活動部件方面，還體現在我們公司不同產品組的需求方面，無論他們是在人工智慧團隊中，還是在圍繞排名和推薦做更多核心人工智慧工作。因此，供給和需求都相當不穩定。

And so we don't have a sort of fixed answer in terms of when we expect that we will sort of have enough supply to meet all demand, but that's something that we are working very hard to alleviate and it's part of why we accelerated bringing more data center space online this year. And also, we're very focused on increasing the efficiency of our workloads over the course of the year.

因此，對於我們何時能夠擁有足夠的供應來滿足所有需求，我們沒有一個固定的答案，但我們正在努力緩解這個問題，這也是我們今年加速上線更多資料中心空間的原因之一。此外，我們非常注重在一年內提高工作量的效率。

On your second question about ads, performance ads ranking. We have invested for many years and continue to invest in driving ad performance improvements. Year-over-year conversion growth remains strong. And in fact, we continue to see conversions grow at a faster rate than ad impressions in Q1, so reflecting increased conversion rates and ads ranking and modeling improvements are a big driver of overall performance gains. We have a lot of innovations and model architecture in both the ads retrieval and ranking stages of the ads delivery process to serve more relevant ads to people. We talked about the introduction of the new GEM ads recommendation model in Q1.

關於廣告的第二個問題，效果廣告排名。我們已經投資多年，並將繼續投資於推動廣告效果的改進。年比轉換率成長依然強勁。事實上，我們繼續看到第一季轉換率的成長速度快於廣告展示率，因此反映轉換率的提高和廣告排名和建模的改進是整體效能提升的重要驅動力。我們在廣告投放過程的廣告檢索和排名階段都有很多創新和模型架構，以便為人們提供更相關的廣告。我們在第一季談到了新的GEM廣告推薦模型的推出。

And we have talked about some of the prior model architecture improvements like Lattice and Andromeda in past quarters. For us, we really believe, first and foremost, that advertising is a relative performance game. That's especially important for us because the vast majority of our business is direct response advertising. So we feel good about how the prior investments are paying off and we continue to invest in a lot of different work to constantly improve our ads ranking and recommendations performance.

在過去的幾個季度中，我們討論了一些先前的模型架構改進，例如 Lattice 和 Andromeda。對我們來說，我們首先確實相信，廣告是一場相對績效的遊戲。這對我們來說尤其重要，因為我們的絕大部分業務是直接回應廣告。因此，我們對先前的投資回報感到滿意，我們將繼續投入大量不同的工作來不斷提高我們的廣告排名和推薦效果。

Kenneth Gawrelski, Wells Fargo.

富國銀行的肯尼斯‧加夫雷爾斯基 (Kenneth Gawrelski)。

Two for me, please. First, maybe, Mark, how should we think about the timing of AI capabilities necessary to drive WhatsApp for business adoption in higher labor -- cost labor markets. What is Meta doing to accelerate that adoption? And do you see this as mostly incremental to SME ad spend that you're already capturing? And then for Susan, one, please.

請給我兩份。首先，馬克，我們應該如何考慮推動 WhatsApp 在勞動成本較高的勞動市場中被企業採用所需的人工智慧能力的時機。Meta 正在採取什麼措施來加速這項採用？您是否認為這對您已經獲得的中小企業廣告支出來說主要是增量？然後請蘇珊喝一杯。

What is the revised CapEx outlook for this year for '25 mean about future years? Does it mean anything? Or you talked about this being an acceleration in your revised outlook statement. Should we think about this as a new starting point for -- to think about '26 and beyond? Or should we just start fresh in '26 and think about the needs and capacity at that point?

今年 25 年資本支出前景的修訂對未來幾年意味著什麼？這有什麼意義嗎？或者您在修改後的展望聲明中談到了這是一種加速。我們是否應該將此視為一個新的起點——思考 26 年及以後的事？或者我們應該從 26 年重新開始，考慮當時的需求和能力？

I'm happy to take -- I'll go ahead and take both of those. And Mark, you should feel free to chime in wherever you would like. So Mark talked a little bit about our general vision that every business will soon have an AI that is an expert in their business for their customers, to talk to you in the same way that today, they've got e-mail and websites, social media presences, et cetera. We are currently testing business AIs with a limited set of businesses in the US and a few additional countries on WhatsApp Messenger and on ads on Facebook and Instagram.

我很樂意接受——我會繼續接受這兩個。馬克，你可以隨意發表你的看法。因此，馬克談到了我們的整體願景，即每個企業很快都會擁有一個人工智慧，這是他們業務方面的專家，可以為客戶提供服務，就像今天透過電子郵件、網站、社交媒體等方式與您交談一樣。我們目前正在透過 WhatsApp Messenger 以及 Facebook 和 Instagram 上的廣告與美國和其他幾個國家的有限企業測試商業 AI。

We've been starting with small business and focusing first on helping them sell their goods and services with business AIs. But ultimately, we are working on tools to support businesses at every stage of the customer funnel from lead generation to order management and customer service. And a core area that we're addressing right now is really the ability for businesses to customize and control the agent to achieve the outcome that they want.

我們從小型企業開始，首先專注於利用商業人工智慧來幫助他們銷售商品和服務。但最終，我們正在開發工具來支援客戶漏斗的每個階段的業務，從潛在客戶生成到訂單管理和客戶服務。我們目前正在解決的一個核心領域實際上是企業客製化和控制代理商以實現他們想要的結果的能力。

So we've launched a new agent management experience and dashboard that makes it easier for businesses to train their AI based on existing information on their website or WhatsApp profile or their Instagram and Facebook pages and we're starting with the ability for businesses to activate AI in their chats with customers. We are also testing business AIs on Facebook and Instagram ads that you can ask about product and return policies or assist you in making a purchase within our in-app browser.

因此，我們推出了一種新的代理管理體驗和儀表板，使企業能夠更輕鬆地根據其網站或 WhatsApp 個人資料或 Instagram 和 Facebook 頁面上的現有資訊來訓練他們的人工智慧，並且我們開始讓企業能夠在與客戶的聊天中啟動人工智慧。我們也會在 Facebook 和 Instagram 廣告上測試商業 AI，您可以在我們的應用程式內瀏覽器中詢問產品和退貨政策，或協助您購買。

So again, the ultimate vision is to build an experience that serves customers across all of these different services and apps. No matter where you engage with the business AI, it should be one agent that recalls your history and your preferences. And we're hearing encouraging feedback, particularly that adopting these AIs or saving the business that we're testing with a lot of time in helping to determine which conversations make sense for them to spend more time on.

所以，最終的願景是打造一種服務所有不同服務和應用程式的客戶體驗。無論您在何處與商業 AI 互動，它都應該是一個能夠回憶您的歷史和偏好的代理商。我們聽到了令人鼓舞的回饋，特別是採用這些人工智慧或為我們正在測試的企業節省大量時間，幫助確定哪些對話值得他們花更多時間。

And then your second question, right, was about 2026 CapEx. Infrastructure, as I alluded to earlier, just is a very dynamic planning area given the continued advances in AI and also for us, the fact that we continue to find a lot of good use cases to put capacity toward a core AI ranking and recommendations work. So I would say it's too early to discuss plans beyond 2025.

然後你的第二個問題是關於 2026 年資本支出。正如我之前提到的，鑑於人工智慧的持續進步，基礎設施是一個非常動態的規劃領域，對我們來說，我們繼續發現許多良好的用例，將容量用於核心人工智慧排名和推薦工作。所以我認為現在討論 2025 年以後的計劃還為時過早。

Youssef Squali, Truist Securities.

Youssef Squali，Truist Securities。

So Mark, in a world where we now have maybe 5 to 10 chatbots, including Meta AI on our smartphone is doing in virtually the same thing. Do you think this is a market much like search where the winner takes most or is it likely to be much more fragmented. But in either case, what would you say are the top two or three competitive advantages of Meta AI. And then Susan, on the EU decision in connection with the DMA, what kind of modification will you need to make to the apps? And can you maybe just help us gauge the potential financial fallout, understanding that it may obviously be too early.

所以馬克，我們現在的世界可能有 5 到 10 個聊天機器人，包括我們智慧型手機上的 Meta AI 也在做幾乎同樣的事情。您是否認為這是一個類似搜尋的市場，贏家通吃，還是說這個市場可能會更加分散？但無論如何，您認為 Meta AI 的兩到三個最大競爭優勢是什麼？然後蘇珊，關於歐盟與 DMA 相關的決定，您需要對應用程式進行哪些修改？您能否幫助我們評估潛在的財務影響，請理解現在顯然還為時過早。

Yes. On Meta AI, I mean, I think that there are going to be a number of different agents that people use, just like people use different apps for different things. I'm not sure that people are going to use multiple agents for the same exact things, but I'd imagine that something that is more focused on kind of enterprise productivity might be different from something that is somewhat more optimized for personal productivity and that might be somewhat different from something that is optimized for entertainment and social connectivity.

是的。關於 Meta AI，我的意思是，我認為人們會使用許多不同的代理，就像人們使用不同的應用程式來做不同的事情一樣。我不確定人們是否會使用多個代理商來做同樣的事情，但我認為，更注重企業生產力的東西可能與針對個人生產力進行優化的東西有所不同，也可能與針對娛樂和社交連接進行優化的東西有所不同。

So then there were different experiences. One of the trends that I think we're starting to see now is personalization across these. Right now, if the experience is unpersonalized, then you can kind of just go to different apps and get reasonably similar answers to different questions.

所以那時就有了不同的經驗。我認為我們現在開始看到的趨勢之一就是這些方面的個人化。現在，如果體驗不是個人化的，那麼你只能使用不同的應用程式並獲得不同問題的相當相似的答案。

But once an AI starts getting to know you and what you care about in context and can build up memory from the conversations that you've had with it over time, I think that will start to become somewhat more of a differentiator. So that's one thing that we think will matter. And then, of course, there's all the different modalities being able to not just answer questions about in text but being able to do voice and multimodal and be able to produce images and videos and understand all those things and have good conversations about that, I think, is going to be important overall. So yes, I mean I think Meta AI is well positioned, but we have a lot of work to do in order to make it the leading personal AI.

但是，一旦人工智慧開始了解你以及你在特定情境下所關心的事物，並能從你與它長期進行的對話中建立記憶，我認為這將開始成為更具差異化的因素。所以我們認為這是一件很重要的事。當然，還有各種不同的模式，不僅能夠回答有關文本的問題，還能夠進行語音和多模式處理，能夠製作圖像和視頻，理解所有這些內容，並就此進行良好的對話，我認為，總的來說，這將非常重要。所以是的，我的意思是我認為 Meta AI 定位很好，但為了使其成為領先的個人 AI，我們還有很多工作要做。

And Youssef, on your second question, it is really too early to speak about what those changes could be because we are in the process of engaging with the European Commission. I think maybe the most useful sort of metric I could give is just that our advertising revenue in the European economic area in Switzerland, which would be the geographies impacted here was 16% of our worldwide total revenue in 2024. Again, we are continuing to engage actively with the European Commission further on this, so we hope to have more clarity by next quarter's call.

尤素夫，關於你的第二個問題，現在談論這些變化可能是什麼還為時過早，因為我們正在與歐盟委員會接觸。我認為我能給的最有用的指標是，我們在瑞士歐洲經濟區（受影響的地區）的廣告收入占我們 2024 年全球總收入的 16%。再次，我們將繼續就此事與歐盟委員會進行積極接觸，因此我們希望在下個季度的電話會議中獲得更清晰的答案。

Krista, we have time for one last question.

克里斯塔，我們還有時間問最後一個問題。

Mark Mahaney, Evercore ISI.

馬克‧馬哈尼 (Mark Mahaney)，Evercore ISI。

I'll just throw in two. I think you called out that the China-based retailers as one sort of potentially soft advertising vertical. Anything else you'd call out? And I would just suggest autos. Is that an area of any softness?

我只會投入兩個。我認為您稱中國零售商為一種潛在的軟廣告垂直行業。還有什麼想說的嗎？我只想建議汽車。那是柔軟的區域嗎？

And then on the Reality Labs and on the losses associated with Reality Labs, they've been very consistent $4 billion a quarter for quite some time. Is there light at the end of the tunnel? Is there a reason to think, is there a factor that would occur that would cause those losses to come down when would that be? But maybe more importantly, what is going to cause those losses to come down?

然後就現實實驗室以及與現實實驗室相關的損失而言，相當長一段時間內，它們一直保持著每季 40 億美元的穩定水平。隧道盡頭有光嗎？是否有理由去思考，是否存在一個因素會導致這些損失下降，那什麼時候會下降？但也許更重要的是，什麼會導致這些損失下降？

Mark, let me take your first question about other verticals. We generally saw healthy growth in most verticals in Q1. We did see some weakness in gaming and politics. So year-over-year growth in gaming was negative in Q1 as we lapped a period of strong spend from China-based advertisers that were promoting a larger volume of game titles in Q1 of 2024. And then year-over-year growth in the government and politics vertical dropped sharply as expected with the conclusion of US elections, but that continues to just be a very small vertical overall. And then your second question on Reality Labs.

馬克，讓我來回答你關於其他垂直行業的第一個問題。我們總體上看到第一季大多數垂直行業都實現了健康成長。我們確實看到了遊戲和政治方面的一些弱點。因此，第一季遊戲業務的同比增長為負，因為我們經歷了中國廣告商的強勁支出時期，這些廣告商在 2024 年第一季推廣了大量遊戲。隨著美國大選的結束，政府和政治垂直領域的同比增長如預期般急劇下降，但總體而言，這仍然是一個非常小的垂直領域。然後是關於現實實驗室的第二個問題。

Yes. Reality Labs one. I mean we're basically focused on doing the work more efficiently. But as the AI glasses have really taken off, I've talked about this on a number of calls. There are more investments that I think makes sense to make around making sure that we can distribute this and grow it very quickly.

是的。現實實驗室一號。我的意思是我們基本上專注於更有效率地完成工作。但隨著人工智慧眼鏡的真正普及，我在多次電話會議中都談到了這個問題。我認為進行更多投資是有意義的，以確保我們能夠快速分銷並實現成長。

I mean some of the -- if you look at some of the leading consumer electronics products of other categories, by the time they get to their third generation, they're often selling 10 million units and scaling from there. And I'm not sure if we're going to do exactly that, but I think that that's like the ballpark of the opportunity that we have. And that's something that I think we're kind of focused on scaling to that and then scaling beyond that for the generations after that.

我的意思是——如果你看看其他類別的一些領先的消費電子產品，當它們發展到第三代時，它們的銷量通常已經達到 1000 萬台，並且還在不斷擴大。我不確定我們是否會這樣做，但我認為這就是我們所擁有的機會的大致範圍。我認為，我們關注的重點是將其擴展到這個水平，然後繼續為後代擴展這一水平。

So I think some of the effort that we're doing is going to -- we're going to get more efficient in some parts of the work that we do. But then as a bunch of the products start to hit and start to grow even bigger than the number that I just said is just sort of like the sort of a near-term milestone, then I think we'll continue scaling in terms of distribution.

所以我認為我們正在做的一些努力將會——我們將在工作的某些部分變得更有效率。但是，隨著大量產品開始暢銷，並且開始成長，甚至超過我剛才所說的近期里程碑的數字，那麼我認為我們將繼續擴大分銷規模。

And then at some point, just like the other products that we build out, we will feel like we're at a sufficient scale that we're going to primarily focus on making sure that we're monetizing and building an efficient business around it. So that's kind of where we're at on it. We're definitely focused on doing the work more efficiently, but also very optimistic about what we're seeing in the results, especially on the AI glass side.

然後到了某個時候，就像我們開發的其他產品一樣，我們會覺得我們已經達到了足夠的規模，我們將主要集中精力確保我們能夠實現盈利並圍繞它建立高效的業務。這就是我們現在的處境。我們確實專注於更有效率地完成工作，但對所看到的結果也非常樂觀，特別是在人工智慧玻璃方面。

Thank you, everyone, for joining us and we look forward to speaking to you again soon.

感謝大家的參與，我們期待很快能再次與你們交談。