(META) 2023 Q4 法說會逐字稿

Good afternoon. My name is Krista and I'll be your conference operator today. At this time, I would like to welcome everyone to the Meta Fourth Quarter and Full Year 2023 Earnings Conference Call. (Operator Instructions) This call will be recorded. Thank you very much.

午安.我叫克里斯塔，今天我將擔任你們的會議接線生。此時，我歡迎大家參加 Meta 第四季和 2023 年全年財報電話會議。 （操作員說明）此通話將會被錄音。非常感謝。

Ken Dorell, Meta's Director of Investor Relations, you may begin.

Meta 的投資者關係總監 Ken Dorell，您可以開始了。

Thank you. Good afternoon and welcome to Meta Platform's Fourth Quarter and Full Year 2023 Earnings Conference Call. Joining me today to discuss our results are Mark Zuckerberg, CEO; and Susan Li, CFO.

謝謝。下午好，歡迎參加 Meta Platform 2023 年第四季和全年財報電話會議。今天與我一起討論我們的結果的是執行長馬克·祖克柏 (Mark Zuckerberg)；和財務長李蘇珊。

Before we get started, I would like to take this opportunity to remind you that our remarks today will include forward-looking statements. Actual results may differ materially from those contemplated by these forward-looking statements. Factors that could cause these results to differ materially are set forth in today's earnings press release and in our quarterly report on Form 10-Q filed with the SEC. Any forward-looking statements that we make on this call are based on assumptions as of today, and we undertake no obligation to update these statements as a result of new information or future events.

在我們開始之前，我想藉此機會提醒您，我們今天的言論將包含前瞻性陳述。實際結果可能與這些前瞻性陳述預期的結果有重大差異。今天的收益新聞稿和我們向 SEC 提交的 10-Q 表格季度報告中列出了可能導致這些結果出現重大差異的因素。我們在本次電話會議中所做的任何前瞻性陳述均基於截至目前的假設，我們不承擔因新資訊或未來事件而更新這些陳述的義務。

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.fb.com.

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 指標的調整表。收益新聞稿和隨附的投資者簡報可在我們的網站 Investor.fb.com 上取得。

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. Hi, everyone. Thanks for joining us. This was a good quarter, and it wrapped up an important year for our community and our company. We estimate that there are more than 3.1 billion people who use at least one of our apps each day. 2023 was our year of efficiency, which focused on making Meta a stronger technology company and improving our business to give us the stability to deliver our ambitious long-term vision for AI and the metaverse.

好的。大家好。感謝您加入我們。這是一個很好的季度，它結束了對我們的社區和公司來說重要的一年。我們估計每天有超過 31 億人至少使用我們的一款應用程式。 2023 年是我們的效率年，我們的重點是使 Meta 成為一家更強大的科技公司，並改善我們的業務，使我們能夠穩定地實現人工智慧和元宇宙雄心勃勃的長期願景。

And last year, not only did we achieve our efficiency goals, but we returned to strong revenue growth, saw strong engagement across our apps; shipped a number of exciting new products like Threads, Ray-Ban Meta smart glasses and mixed reality in Quest 3; and of course, established a world-class AI effort that's going to be the foundation for many of our future products. I think that being a leaner company is helping us execute better and faster, and we will continue to carry these values forward as a permanent part of how we operate.

去年，我們不僅實現了效率目標，而且恢復了強勁的收入成長，我們的應用程式的參與度也很高；發布了許多令人興奮的新產品，例如 Threads、Ray-Ban Meta 智慧眼鏡和 Quest 3 中的混合實境；當然，我們建立了世界一流的人工智慧工作，這將成為我們未來許多產品的基礎。我認為，成為一家更精簡的公司有助於我們更好、更快地執行，我們將繼續發揚這些價值觀，將其作為我們營運方式的永久組成部分。

Now moving forward, a major goal, we'll be building the most popular and most advanced AI products and services. And if we succeed, everyone who uses our services will have a world-class AI assistant to help get things done, every creator will have an AI that their community can engage with, every business will have an AI that their customers can interact with to buy goods and get support, and every developer will have a state-of-the-art open-source model to build with.

現在，我們的一個主要目標是打造最受歡迎、最先進的人工智慧產品和服務。如果我們成功，每個使用我們服務的人都將擁有世界一流的人工智慧助理來幫助完成工作，每個創作者都將擁有一個可以與他們的社區互動的人工智慧，每個企業都將擁有一個可以與他們的客戶互動的人工智慧購買商品並獲得支持，每個開發人員都將擁有一個最先進的開源模型來建造。

I also think that everyone will want a new category of computing devices that let you frictionlessly interact with AIs that can see what you see and hear what you hear, like smart glasses. And one thing that became clear to me in the last year is that this next generation of services requires building full general intelligence. Previously, I thought that because many of the tools were social-, commerce- or maybe media-oriented that it might be possible to deliver these products by solving only a subset of AI's challenges.

我還認為，每個人都會想要一種新的計算設備，讓你能夠與人工智慧無縫交互，讓人工智慧能夠看到你所看到的、聽到你所聽到的，例如智慧眼鏡。去年我清楚的一件事是，下一代服務需要建立完整的通用智慧。先前，我認為，由於許多工具都是面向社交、商業或媒體的，因此可以透過僅解決人工智慧挑戰的一部分來交付這些產品。

But now it's clear that we're going to need our models to be able to reason, plan, code, remember and many other cognitive abilities in order to provide the best versions of the services that we envision. We've been working on general intelligence research and FAIR for more than a decade. But now general intelligence will be the theme of our product work as well. Meta has a long history of building new technologies into our services, and we have a clear long-term playbook for becoming leaders.

但現在很明顯，我們需要我們的模型能夠推理、計劃、編碼、記憶和許多其他認知能力，以便提供我們設想的最佳版本的服務。十多年來，我們一直致力於通用情報研究和 FAIR。但現在通用智能也將成為我們產品工作的主題。 Meta 在將新技術融入我們的服務方面有著悠久的歷史，我們為成為領導者制定了明確的長期策略。

And there are a few key aspects of this that I want to take some time to go through today. The first is world-class compute infrastructure. I recently shared that, by the end of this year, we'll have about 350,000 H100s, and including other GPUs, that will be around 600,000 H100 equivalents of compute. We're well positioned now because of the lessons that we learned from Reels. We initially underbuilt our GPU clusters for Reels. And when we were going through that, I decided that we should build enough capacity to support both Reels and another Reels-sized AI service that we expected to emerge so we wouldn't be in that situation again.

今天我想花一些時間來討論其中的幾個關鍵方面。首先是世界一流的運算基礎設施。我最近表示，到今年年底，我們將擁有約 35 萬個 H100，包括其他 GPU，這將相當於約 60 萬個 H100 的計算量。由於我們從 Reels 中吸取了教訓，我們現在已經處於有利位置。我們最初為 Reels 建立了 GPU 叢集。當我們經歷這個過程時，我決定我們應該建立足夠的能力來支援 Reels 和我們預計會出現的另一個 Reels 大小的人工智慧服務，這樣我們就不會再遇到這種情況了。

And at the time, the decision was somewhat controversial, and we faced a lot of questions about CapEx spending, but I'm really glad that we did this. Now going forward, we think that training and operating future models will be even more compute-intensive. We don't have a clear expectation for exactly how much this will be yet, but the trend has been that state-of-the-art large language models have been trained on roughly 10x the amount of compute each year.

當時，這個決定有些爭議，我們面臨著許多關於資本支出的問題，但我真的很高興我們這麼做了。現在，我們認為未來模型的訓練和操作將更加計算密集。我們還沒有明確預期具體會達到多少，但趨勢是，最先進的大型語言模型每年訓練的計算量大約是原來的 10 倍。

And our training clusters are only part of our overall infrastructure, and the rest, obviously, isn't growing as quickly. But overall, we're playing to win here, and I expect us to continue investing aggressively in this area. In order to build the most advanced clusters, we're also designing novel data centers and designing our own custom silicons specialized for our workloads.

我們的培訓集群只是我們整體基礎設施的一部分，顯然，其餘部分的成長速度沒有那麼快。但總的來說，我們正在努力贏得勝利，我預計我們將繼續在這一領域積極投資。為了建造最先進的集群，我們還設計了新穎的資料中心，並設計了專門針對我們的工作負載的客製化晶片。

The second part of our playbook is open-source software infrastructure. Our long-standing strategy has been to build an open-source general infrastructure while keeping our specific product implementations proprietary. In the case of AI, the general infrastructure includes our Llama models, including Llama 3, which is training now, and it's looking great so far, as well as industry standard tools like PyTorch that we've developed.

我們的策略的第二部分是開源軟體基礎設施。我們的長期策略是建立開源通用基礎設施，同時保持我們特定產品實現的專有性。就人工智慧而言，通用基礎設施包括我們的 Llama 模型，包括現在正在訓練的 Llama 3，到目前為止它看起來很棒，以及我們開發的 PyTorch 等行業標準工具。

And this approach to open source has unlocked a lot of innovation across the industry, and it's something that we believe in deeply. And I know that some people have questions about how we benefit from open sourcing, the results of our research and large amounts of compute. So I thought it might be useful to lay out the strategic benefits here.

這種開源方法已經釋放了整個行業的大量創新，這是我們深信不疑的。我知道有些人對我們如何從開源、我們的研究結果和大量計算中受益有疑問。所以我認為在這裡列出戰略利益可能會很有用。

The short version is that open sourcing improves our models. And because there's still significant work to turn our models into products because there will be other open-source models available anyway, we find that there are mostly advantages to being the open-source leader, and it doesn't remove differentiation for our products much anyway. And more specifically, there are several strategic benefits.

簡而言之，開源改進了我們的模型。而且，由於將我們的模型轉化為產品仍有大量工作要做，因為無論如何都會有其他開源模型可用，所以我們發現成為開源領導者有很多優勢，並且它並沒有消除我們產品的差異化反正。更具體地說，有幾個戰略好處。

First, open-source software is typically safer and more secure as well as more compute-efficient to operate due to all the ongoing feedback, scrutiny and development from the community. Now this is a big deal because safety is one of the most important issues in AI. Efficiency improvements and lowering the compute costs also benefit everyone, including us.

首先，由於社群的所有持續回饋、審查和開發，開源軟體通常更安全、更可靠，並且運行時計算效率更高。現在這是一件大事，因為安全是人工智慧中最重要的問題之一。效率的提高和計算成本的降低也使包括我們在內的每個人受益。

Second, open-source software often becomes an industry standard. And when companies standardize on building with our stack, that then becomes easier to integrate new innovations into our products. That's subtle, but the ability to learn and improve quickly is a huge advantage. And being an industry standard enables that.

其次，開源軟體往往成為業界標準。當公司使用我們的堆疊進行標準化建置時，將新的創新整合到我們的產品中就會變得更容易。這很微妙，但快速學習和提高的能力是一個巨大的優勢。作為行業標準可以實現這一點。

Third, open source is hugely popular with developers and researchers. And we know that people want to work on open systems that will be widely adopted. So this helps us recruit the best people at Meta, which is a very big deal for leading in any new technology area. And again, we typically have unique data and build unique product integrations anyway, so providing infrastructure like Llama as open source doesn't reduce our main advantage. This is why our long-standing strategy has been to open source general infrastructure and why I expect it to continue to be the right approach for us going forward.

第三，開源深受開發人員和研究人員的歡迎。我們知道人們希望在將被廣泛採用的開放系統上工作。因此，這有助於我們在 Meta 招募最優秀的人才，這對於任何新技術領域的領先來說都是非常重要的。再說一次，我們通常擁有獨特的數據並建立獨特的產品集成，因此提供像 Llama 這樣的開源基礎設施並不會減少我們的主要優勢。這就是為什麼我們的長期策略是開源通用基礎設施，也是為什麼我希望它繼續成為我們前進的正確方法。

The next part of our playbook is just taking a long-term approach towards the development. While we're working on today's products and models, we're also working on the research that we need to advance for Llama 5, 6 and 7 in the coming years and beyond to develop full general intelligence. It's important to have a portfolio of multiyear investments in research projects, but it's also important to have clear launch vehicles like future Llama models that help focus our work.

我們的策略的下一部分只是採取長期的發展方法。在我們致力於開發當今的產品和模型的同時，我們也致力於在未來幾年及更長時間內為 Llama 5、6 和 7 推進研究，以開發全面的通用智慧。擁有對研究項目的多年投資組合很重要，但擁有明確的運載火箭（例如未來的美洲駝模型）也很重要，這有助於我們集中工作。

We've worked on general intelligence in our lab, FAIR, for more than a decade, as I mentioned, and we produced a lot of valuable work. But having clear product targets for delivering general intelligence really focuses this work and helps us build the leading research program.

正如我所提到的，我們在 FAIR 實驗室研究通用智慧已經十多年了，我們做了許多有價值的工作。但是，擁有明確的產品目標來提供通用智慧確實可以集中這項工作，並幫助我們建立領先的研究計劃。

Now the next key part of our playbook is learning from unique data and feedback loops in our products. When people think about data, they typically think about the corpus that you might use to train a model upfront. And on Facebook and Instagram, there are hundreds of billions of publicly shared images and tens of billions of public videos, which we estimate is greater than the common crawl data set. And people share large numbers of public text posts and comments across our services as well.

現在，我們的策略的下一個關鍵部分是從我們產品中的獨特數據和回饋循環中學習。當人們考慮資料時，他們通常會考慮可用於預先訓練模型的語料庫。而在 Facebook 和 Instagram 上，有數千億個公開分享的圖像和數百億個公共視頻，我們估計比常見的爬行數據集還要多。人們也在我們的服務中分享大量的公共文字貼文和評論。

But even more important in the upfront training corpus is the ability to establish the right feedback loops with hundreds of millions of people interacting with AI services across our products. And this feedback is a big part of how we've improved our AI systems so quickly with Reels and Ads, especially over the last couple of years when we had to re-architect it around new rules.

但在前期訓練語料庫中，更重要的是能夠建立正確的回饋循環，讓數億人與我們產品中的人工智慧服務互動。這種反饋是我們如何透過捲軸和廣告如此快速地改進人工智慧系統的重要組成部分，特別是在過去幾年中，當我們必須圍繞新規則重新建構它時。

Now that brings me to the last part of our playbook for building leading services, which is our culture of rapid learning and experimentation across our apps. When we decide that a new technology, like AI-recommended Reels, is going to be an important part of the future, we're not shy about having multiple teams experimenting with different versions across our apps until we get it right. And then we learn what works and we roll it out to everyone.

現在，我要講的是我們建立領先服務的手冊的最後一部分，這就是我們在應用程式中快速學習和實驗的文化。當我們確定一項新技術（例如 AI 推薦的 Reels）將成為未來的重要組成部分時，我們會毫不猶豫地讓多個團隊在我們的應用程式中嘗試不同版本，直到我們做到正確為止。然後我們了解什麼是有效的，並將其推廣給每個人。

And there used to be this meme that we'd probably launch Stories on our Settings page at some point. And look, I think it's kind of funny because it gets to a core part of our approach. We start by learning and tuning our products until they perform the way we want, and then we roll them out very broadly.

曾經有這樣一個迷因，我們可能會在某個時候在我們的「設定」頁面上啟動故事。看，我認為這有點有趣，因為它涉及我們方法的核心部分。我們首先學習和調整我們的產品，直到它們按照我們想要的方式運行，然後我們將它們廣泛推廣。

And sometimes, occasionally, products will blow up before we're ready for them to, like Threads, although I'll note that Threads now has more people actively using it today than it did during its initial launch peak. So that one is, I think, on track to be a major success.

有時，有時候，產品在我們準備好之前就會爆炸，就像 Threads 一樣，儘管我會注意到 Threads 現在有更多的人積極使用它，而不是在最初的發布高峰期。因此，我認為，這項舉措有望取得重大成功。

But normally, we learn and we iterate methodically. And we started doing that with our AI services in the fall launching Meta AI, our assistant; AI Studio, which the precursor to Creator AIs; our Alpha with Business AIs; and the Ray-Ban Meta smart glasses. And we've been tuning each of these, and we're getting closer to rolling them out more widely. So you should expect that in the coming months.

但通常情況下，我們會有條不紊地學習和迭代。我們從秋季開始透過我們的人工智慧服務來做到這一點，推出了我們的助理 Meta AI； AI Studio，Creator AI 的前身；我們的 Alpha 與商業人工智慧；以及雷朋 Meta 智慧眼鏡。我們一直在調整其中的每一個，並且越來越接近更廣泛地推廣它們。所以你應該在接下來的幾個月預料到這一點。

From there, we'll focus on rolling out services until they reach hundreds of millions or billions of people. And usually, only when we reach that kind of scale do we start focusing on what monetization will look like. And although in this case, the way the business AIs will help business messaging grow and WhatsApp, Messenger and Instagram is pretty clear. But that's our basic approach, and I'm really excited about pointing our company at developing so many of these awesome things.

從那時起，我們將專注於推出服務，直到它們惠及數億或數十億人。通常，只有當我們達到這種規模時，我們才會開始關注貨幣化的情況。儘管在這種情況下，商業人工智慧將幫助商業訊息傳遞以及 WhatsApp、Messenger 和 Instagram 發展的方式非常明確。但這是我們的基本方法，我真的很高興能指導我們公司開發這麼多很棒的東西。

Now we have 2 major parts of our long-term vision, and in addition to AI, the other part is the metaverse. We've invested heavily in both AI and the metaverse for a long time, and we will continue to do so. These days, there are a lot of questions more about AI that I get, and that field is moving very quickly. But I still expect that this next generation of AR, VR and MR computing platforms to deliver a realistic sense of presence that will be the foundation for the future of social experiences and almost every other category of experiences as well.

現在我們的長期願景有兩個主要部分，除了AI之外，另一個部分是元宇宙。長期以來，我們在人工智慧和虛擬宇宙方面投入了大量資金，並且我們將繼續這樣做。這些天，我收到了很多關於人工智慧的問題，而且這個領域發展得非常快。但我仍然期望下一代 AR、VR 和 MR 運算平台能夠提供真實的臨場感，這將成為未來社交體驗以及幾乎所有其他體驗類別的基礎。

Reality Labs crossed $1 billion in revenue in Q4 for the first time with Quest having a strong holiday season. Quest 3 is off to a strong start, and I expect it to continue to be the most popular mixed reality device. With Quest 3 and Quest 2 both performing well, we saw that the Quest app was actually the most downloaded app in the App Store on Christmas Day. I want to give a shout-out to Asgard's Wrath2, which was developed by one of our in-house studios and received IGN's 10 out of 10 masterpiece rating, making it one of the best-rated games out there, not just in VR, but of any game on any platform ever.

隨著 Quest 假期季節的強勁表現，Reality Labs 第四季營收首次突破 10 億美元。 Quest 3 已經有了一個良好的開端，我預計它將繼續成為最受歡迎的混合實境裝置。由於 Quest 3 和 Quest 2 都表現出色，我們發現 Quest 應用程式實際上是聖誕節當天 App Store 中下載次數最多的應用程式。我想對《Asgard's Wrath2》表示讚揚，它是由我們的一個內部工作室開發的，並獲得了IGN 的10 分（滿分10 分）傑作評級，使其成為評分最高的遊戲之一，不僅僅是在VR 領域，但任何平台上的任何遊戲。

So it's a really good sign that we're able to deliver that quality of work at Meta. Horizon is growing quickly, too. It is now on Top 10 Most Used Apps on Quest, and we have an exciting road map ahead. This is another example of applying the long-term playbook that I discussed earlier with AI but in another area. We take the time to build up the core technology and tune the experience. And then when it's ready, we're good at growing things. Now our focus for this year is going to be on growing the mobile version of Horizon as well as the VR One.

因此，這是一個非常好的跡象，表明我們能夠在 Meta 提供如此高品質的工作。 Horizo​​n 也在快速成長。它現已躋身 Quest 上十大最常用應用程式之列，我們的未來路線圖令人興奮。這是我之前討論過的人工智慧長期策略應用的另一個例子，但是是在另一個領域。我們花時間建立核心技術並調整體驗。當一切準備就緒時，我們就擅長種東西了。現在我們今年的重點是開發 Horizo​​n 的行動版本以及 VR One。

Ray-Ban Meta smart glasses are also off to a very strong start, both in sales and engagement. Our partner, EssilorLuxottica is already planning on making more than we both expected due to high demand. Engagement and retention are also significantly higher than the first version of the glasses. The experience is just a lot better with Meta AI in there as well as there's a higher resolution camera, better audio and more. And we also have an exciting road map of software improvements ahead, starting with rolling out multimodal AI and then some other really exciting new AI features later in the year.

雷朋 Meta 智慧眼鏡在銷售和參與度方面也取得了良好的開局。由於需求旺盛，我們的合作夥伴 EssilorLuxottica 已計劃生產超出我們預期的產品。參與度和保留度也明顯高於第一版眼鏡。有了 Meta AI，體驗就好多了，還有更高解析度的相機、更好的音訊等等。我們還制定了令人興奮的軟體改進路線圖，首先推出多模式人工智慧，然後在今年稍後推出其他一些真正令人興奮的新人工智慧功能。

I said this before, but I think that people are going to want new categories of devices that seamlessly engage with AIs frequently throughout the AI without having to take out your phone and press a button and point it at what you want to see. And I think that smart glasses are going to be a compelling form factor for this, and it's a good example of how our AI and metaverse visions are connected.

我之前說過，但我認為人們會想要新類別的設備，這些設備能夠在整個人工智能過程中頻繁地與人工智能無縫交互，而無需拿出手機按下按鈕並將其指向你想看到的內容。我認為智慧眼鏡將成為一個引人注目的外形因素，它是我們的人工智慧和元宇宙願景如何連結的一個很好的例子。

In addition to AI in the metaverse, we're continuing to improve our Apps and Ads businesses as well. Reels and our discovery engine remain a priority and major driver of engagement, and Messaging continues to be our focus for building the next revenue pillar of our business before our longer-term work reaches scale. But since I went a bit longer in the other areas today, I'm just going to mention a few highlights here.

除了虛擬宇宙中的人工智慧之外，我們還在繼續改進我們的應用程式和廣告業務。 Reels 和我們的發現引擎仍然是參與度的優先事項和主要驅動力，而訊息傳遞仍然是我們在長期工作達到規模之前建立業務下一個收入支柱的重點。但由於今天我在其他領域的討論時間較長，所以我只想在這裡提到一些亮點。

Reels continues to do very well across both Instagram and Facebook. People re-share Reels 3.5 billion times every day. Reels is now contributing to our net revenue across our apps. The biggest opportunity going forward is unifying our recommendation systems across Reels and other types of video that will help people discover the best content across our systems no matter what format it's in.

Reels 在 Instagram 和 Facebook 上繼續表現出色。人們每天轉發 Reels 35 億次。 Reels 現在為我們的應用程式淨收入做出了貢獻。未來最大的機會是統一我們跨 Reels 和其他類型影片的推薦系統，這將幫助人們發現我們系統中的最佳內容，無論其格式如何。

WhatsApp is also doing very well. And the most exciting new trend here is that it is succeeding more broadly in the United States, where there's a real appetite for a private, secure and cross-platform messaging app that everyone can use. And given the strategic importance of the U.S. and its outsized importance for revenue, this is just a huge opportunity.

WhatsApp 的表現也非常出色。最令人興奮的新趨勢是，它在美國取得了更廣泛的成功，美國對每個人都可以使用的私密、安全和跨平台訊息應用程式有著真正的興趣。考慮到美國的戰略重要性及其對收入的巨大重要性，這確實是一個巨大的機會。

Threads is also growing steadily with more than 130 million monthly actives. And I'm optimistic that we can keep the pace of improvements in growth going and show that a friendly, discussion-oriented app can be as widely used as the most popular social apps.

Threads 也穩定成長，每月活躍人數超過 1.3 億。我樂觀地認為，我們可以保持成長的步伐，並表明友好的、以討論為導向的應用程式可以像最受歡迎的社交應用程式一樣廣泛使用。

All right. That's what I wanted to cover today. Our communities are growing, and our business is back on track. Once again, a big thank you to all of our employees, partners, shareholders and everyone in our community for sticking with us and for making 2023 such a success. I'm looking forward to another exciting year ahead.

好的。這就是我今天想要介紹的內容。我們的社區正在發展，我們的業務也重回正軌。再次非常感謝我們所有的員工、合作夥伴、股東以及社區中的每個人，感謝你們對我們的支持，讓我們在 2023 年如此成功。我期待著又一個激動人心的一年。

And now here's Susan.

現在是蘇珊。

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. Q4 total revenue was $40.1 billion, up 25% or 22% on a constant currency basis. Q4 total expenses were $23.7 billion, down 8% compared to last year.

謝謝，馬克，大家下午好。讓我們從我們的綜合結果開始。除非另有說明，所有比較均按年比較。第四季總營收為 401 億美元，成長 25%，以固定匯率計算成長 22%。第四季總支出為 237 億美元，較去年下降 8%。

In terms of the specific line items, cost of revenue decreased 8%, 0driven mainly by lower restructuring costs that were partially offset by higher infrastructure-related costs. R&D increased 8% driven primarily by higher head count-related costs from Family of Apps and Reality Labs as well as higher non-head count-related Reality Labs operating expenses, which were partially offset by lower restructuring costs.

就具體項目而言，收入成本下降了 8%, 0，主要是由於重組成本下降，但基礎設施相關成本上升部分抵消了這種影響。研發增加了 8%，主要是由於應用程式系列和現實實驗室的人員數量相關成本增加以及非人員數量相關的現實實驗室營運費用增加，但重組成本的降低部分抵消了這些費用。

Marketing and sales decreased 29% due mainly to lower marketing spend and restructuring costs. G&A decreased 26% due mainly to lower restructuring expenses. We ended the fourth quarter with over 67,300 employees, down 22% from a year ago while up 2% from Q3 as our hiring efforts have resumed. Fourth quarter operating income was $16.4 billion, representing a 41% operating margin. Our tax rate for the quarter was 17%.

行銷和銷售額下降 29%，主要是由於行銷支出和重組成本減少。一般管理費用下降 26%，主要是由於重組費用減少。截至第四季末，我們的員工人數超過 67,300 名，比去年同期下降 22%，但隨著招聘工作的恢復，比第三季度增長 2%。第四季營業收入為 164 億美元，營業利益率為 41%。我們本季的稅率為 17%。

Net income was $14 billion or $5.33 per share. Capital expenditures, including principal payments on finance leases, were $7.9 billion, driven by investments in servers, data centers and network infrastructure. Free cash flow was $11.5 billion, which reflects a $1.6 billion payment of income taxes deferred from prior quarters of 2023. We ended the year with $65.4 billion in cash and marketable securities and $18.4 billion in debt.

淨利潤為 140 億美元，即每股 5.33 美元。受伺服器、資料中心和網路基礎設施投資的推動，資本支出（包括融資租賃本金支付）為 79 億美元。自由現金流為 115 億美元，反映了從 2023 年前幾季遞延的 16 億美元所得稅。到年底，我們的現金和有價證券為 654 億美元，債務為 184 億美元。

In the fourth quarter, we repurchased $6.3 billion of our Class A stock, bringing our total share repurchases for the full year to $20 billion. We had $30.9 billion remaining on our prior authorization as of December 31. And today, we announced a $50 billion increase in our stock repurchase authorization.

第四季度，我們回購了 63 億美元的 A 類股票，使全年股票回購總額達到 200 億美元。截至 12 月 31 日，我們的事先授權還剩 309 億美元。今天，我們宣布股票回購授權增加 500 億美元。

Moving now to our segment results. I'll begin with our Family of Apps segment. In Q4, we saw continued community growth across the Family of Apps as well as Facebook, specifically. We're sharing today that as previewed when we first introduced our family metrics, we are transitioning away from reporting Facebook-specific metrics.

現在轉向我們的細分結果。我將從我們的應用程式系列部分開始。在第四季度，我們看到整個應用程式家族以及 Facebook 的社群持續成長。今天我們要分享的是，正如我們首次推出家庭指標時所預覽的那樣，我們正在從報告特定於 Facebook 的指標轉變。

As part of this transition, we will no longer report Facebook daily and monthly active users or family monthly active people. In Q1, we will instead begin reporting year-over-year changes in ad impressions and the average price per ad at the regional level while continuing to report family daily active people.

作為過渡的一部分，我們將不再報告 Facebook 每日和每月活躍用戶或家庭每月活躍用戶。在第一季度，我們將開始報告廣告展示次數的同比變化以及區域級每個廣告的平均價格，同時繼續報告家庭日常活躍人數。

For full year 2023, Family of Apps total revenue was $133 billion and ad revenue was $131.9 billion, each up 16% year-over-year. The largest contributors to year-over-year ad revenue growth were the online commerce, CPG, entertainment and media and gaming verticals. The online commerce and gaming verticals benefited from strong demand by advertisers in China reaching people in other markets. In 2023, revenue from China-based advertisers represented 10% of our overall revenue and contributed 5 percentage points to total worldwide revenue growth.

2023 年全年，Family of Apps 總營收為 1,330 億美元，廣告收入為 1,319 億美元，較去年同期成長 16%。廣告收入較去年同期成長的最大貢獻者是線上商務、消費品、娛樂、媒體和遊戲垂直產業。線上商務和遊戲垂直產業受益於中國廣告商對其他市場受眾的強勁需求。 2023年，來自中國廣告主的收入占我們總收入的10%，對全球總收入成長貢獻5個百分點。

Turning back to Q4 results. Q4 total Family of Apps revenue was $39 billion, up 24% year-over-year. Q4 Family of Apps ad revenue was $38.7 billion, up 24% or 21% on a constant currency basis. Within ad revenue, the online commerce vertical was the largest contributor to year-over-year growth, followed by CPG and gaming.

回到第四季的業績。第四季應用程式系列總營收為 390 億美元，較去年同期成長 24%。第四季應用程式系列廣告營收為 387 億美元，成長 24%，以固定匯率計算成長 21%。在廣告收入中，線上商務垂直領域是年成長的最大貢獻者，其次是消費品和遊戲。

On a user geography basis, ad revenue growth was strongest in Europe and Rest of World at 33% and 32%, respectively, followed by Asia Pacific at 23% and North America at 19%. Foreign currency was a tailwind to advertising revenue growth in all international regions.

從用戶地理來看，歐洲和世界其他地區的廣告收入成長最為強勁，分別成長了 33% 和 32%，其次是亞太地區（23%）和北美（19%）。外匯是所有國際地區廣告收入成長的推動力。

In Q4, the total number of ad impressions served across our services increased 21%, and the average price per ad increased 2%. Impression growth was mainly driven by Asia Pacific and Rest of World. Pricing growth was driven by advertiser demand and currency tailwinds, which were partially offset by strong impression growth, particularly from lower-monetizing services and regions.

第四季度，我們服務中的廣告展示總數增加了 21%，每個廣告的平均價格增加了 2%。印象數成長主要由亞太地區和世界其他地區推動。定價成長是由廣告商需求和貨幣順風推動的，但強勁的印象成長（尤其是貨幣化程度較低的服務和地區的印象成長）部分抵消了廣告主需求和貨幣順風的推動。

Family of Apps other revenue was $334 million in Q4, up 82%, driven by business messaging revenue growth from our WhatsApp Business platform. We continue to direct the majority of our investments toward the development and operation of our Family of Apps. In Q4, Family of Apps expenses were $18 billion, representing approximately 76% of our overall expenses. Family of Apps expenses were down 13% due to lower restructuring expenses. Family of Apps operating income was $21 billion, representing a 54% operating margin.

第四季應用程式系列其他營收為 3.34 億美元，成長 82%，這主要得益於 WhatsApp Business 平台的商業訊息收入成長。我們繼續將大部分投資用於應用程式系列的開發和營運。第四季度，應用程式系列支出為 180 億美元，約占我們整體支出的 76%。由於重組費用減少，應用程式系列費用下降了 13%。 Family of Apps 營業收入為 210 億美元，營業利益率為 54%。

Within our Reality Labs segment, Q4 revenue was $1.1 billion, up 47%, driven by Quest 3 sales during the holiday season. Reality Labs expenses were $5.7 billion, up 14% year-over-year due primarily to higher head count-related expenses and non-head count-related R&D spend. Reality Labs operating loss was $4.6 billion.

在我們的 Reality Labs 部門中，受假日季 Quest 3 銷售的推動，第四季度營收為 11 億美元，成長 47%。 Reality Labs 支出為 57 億美元，年增 14%，主要是因為人員數量相關費用和非人員數量相關研發支出增加。 Reality Labs 營運虧損為 46 億美元。

Turning now to the business outlook. There are 2 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 remain pleased with engagement trends and have strong momentum across our product priorities. We're seeing sustained growth in Reels and Video overall as daily watch time across all video types grew over 25% year-over-year in Q4, driven by ongoing ranking improvements.

現在轉向業務前景。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，以及隨著時間的推移，我們透過這種參與獲利的有效性。首先，我們對參與趨勢仍然感到滿意，我們的產品優先事項具有強勁的勢頭。我們看到捲軸和影片整體持續成長，在排名持續提升的推動下，所有影片類型的每日觀看時間在第四季同比增長超過 25%。

In-feed recommendations of posts from accounts you're not connected to also continue to drive incremental engagement as we increasingly personalize recommendations, and we expect to deliver further improvements this year. On Instagram, we're making strong progress increasing the freshness of recommended content. And a big focus for Facebook this year will be unifying our video experiences and ranking across the app, which should create a more seamless user experience and deliver increasingly relevant video recommendations.

隨著我們越來越個人化的推薦，您未連結的帳戶的貼文的動態推薦也將繼續推動增量參與，並且我們預計今年將提供進一步的改進。在 Instagram 上，我們在提高推薦內容的新鮮度方面取得了巨大進展。 Facebook 今年的一大重點將是統一我們的影片體驗和整個應用程式的排名，這應該會創造更無縫的用戶體驗並提供越來越相關的影片推薦。

On WhatsApp, community growth remains strong, and we're seeing good traction with our Channels product, which has quickly scaled to over 500 million monthly active users since we rolled it out globally in September. We are also making good progress in areas that have the potential to grow people's engagement with our products in the longer term, such as Threads and generative AI.

在 WhatsApp 上，社群成長依然強勁，我們看到我們的 Channels 產品具有良好的吸引力，自 9 月在全球推出以來，該產品的每月活躍用戶已迅速擴大到超過 5 億。從長遠來看，我們在有潛力提高人們對我們產品的參與度的領域也取得了良好進展，例如線程和生成人工智慧。

On Threads, we're coming into 2024 with strong product momentum and remain focused on introducing additional valuable features and further scaling the community this year. With generative AI, we fully rolled out our Meta AI assistant and other AI chat experiences in the U.S. at the end of the year and began testing more than 20 GenAI features across our Family of Apps.

在 Threads 上，我們將帶著強勁的產品勢頭進入 2024 年，並將繼續專注於今年推出更多有價值的功能並進一步擴大社群規模。透過生成式 AI，我們在年底在美國全面推出了 Meta AI 助理和其他 AI 聊天體驗，並開始在我們的應用程式系列中測試 20 多項 GenAI 功能。

Our big areas of focus in 2024 will be working towards the launch of Llama 3, expanding the usefulness of our Meta AI assistant and progressing on our AI Studio road map to make it easier for anyone to create an AI. For the experiences in our apps, we'll follow our typical approach of making them great before we lean into growth and eventually monetization. But this continues to be an exciting area that has the potential to transform how people engage on our services in the years ahead.

2024 年我們的重點領域將是致力於推出 Llama 3、擴展 Meta AI 助理的實用性以及推進 AI Studio 路線圖，讓任何人都可以更輕鬆地創建 AI。對於我們應用程式中的體驗，我們將遵循我們的典型方法，在我們致力於成長並最終實現貨幣化之前，讓它們變得更好。但這仍然是一個令人興奮的領域，有可能在未來幾年改變人們使用我們服務的方式。

Now to the second driver of our revenue performance, increasing monetization efficiency. There are 2 parts to this work. The first part is growing the level of ad inventory within organic engagement. Our approach to optimizing ad levels in our apps has become increasingly sophisticated over the years as we've developed a better understanding of the optimal place, time and person to show an ad, which has enabled us to adopt a more dynamic approach to serving apps. We expect to continue that work going forward, while services with relatively lower levels of monetization like video and messaging will serve as additional growth opportunities.

現在我們的收入表現的第二個驅動因素是提高貨幣化效率。這項工作有 2 個部分。第一部分是提高有機參與中的廣告庫存水準。多年來，我們優化應用程式中廣告層級的方法變得越來越複雜，因為我們對展示廣告的最佳地點、時間和人員有了更好的了解，這使我們能夠採用更動態的方法來為應用程式提供服務。我們預計未來將繼續這項工作，而視訊和訊息等貨幣化程度相對較低的服務將成為額外的成長機會。

The other part to increasing monetization efficiency is improving marketing performance. To do this, we're focused on 3 areas. The first is AI. We continue to leverage AI across our ad systems and product suite. We're delivering continued performance gains from ranking improvements as we adopt larger and more advanced models, and this will remain an ongoing area of investment in 2024.

提高貨幣化效率的另一部分是提高行銷績效。為此，我們重點關注 3 個領域。首先是人工智慧。我們繼續在廣告系統和產品套件中利用人工智慧。隨著我們採用更大、更先進的模型，我們將透過排名改進不斷獲得效能提升，這仍然是 2024 年持續投資的領域。

We're also building out our Advantage+ portfolio of solutions to help advertisers leverage AI to automate their advertising campaigns. Advertisers can choose to automate part of the campaign creation setup process, such as who to show their ad to with Advantage+ audience, or they can automate their campaign completely using Advantage+ shopping, which continues to see strong growth. We're also now exploring ways to apply this end-to-end automation to new objectives.

我們也建立了 Advantage+ 解決方案組合，以幫助廣告主利用人工智慧實現廣告活動自動化。廣告主可以選擇自動化部分行銷活動建立設定流程，例如向哪些人顯示其廣告，以及 Advantage+ 受眾，或者他們可以使用 Advantage+ 購物完全自動化其行銷活動，該服務持續強勁成長。我們現在也正在探索將這種端到端自動化應用於新目標的方法。

On the ads creative side, we completed the global rollout of 2 of our generative AI features in Q4, Text Variations and Image Expansion, and plan to broaden availability of our background generation feature later in Q1. Initial adoption of these features has been strong, and tests are showing promising early performance gains. This will remain a big area of focus for us in 2024.

在廣告創意方面，我們在第四季度完成了 2 個生成人工智慧功能的全球推廣：文字變化和圖像擴展，並計劃在第一季稍後擴大背景生成功能的可用性。這些功能的最初採用勢頭強勁，測試顯示出有希望的早期性能提升。這仍將是我們 2024 年關注的重點領域。

The second area of our work is helping businesses connect their marketing data and measure results. Here, we're continuing to invest in conversions API to make it easier for advertisers to not only adopt but maximize its impact. The last area is investing in new and engaging on-platform ad experiences.

我們工作的第二個領域是幫助企業連結行銷數據並衡量結果。在這裡，我們將繼續投資轉換 API，讓廣告主更輕鬆地採用並最大限度地發揮其影響力。最後一個領域是投資新的、有吸引力的平台廣告體驗。

Business messaging remains a priority here. Click-to-message ads continue to grow, and we're seeing advertisers increasingly use these ads to drive down funnel conversions. Paid messaging is growing quickly, and we are focused on broadening adoption in 2024 by making it easier for businesses to buy marketing messages and more easily adopt our Flows product to develop richer in-Thread experiences for consumers.

商業訊息傳遞仍然是這裡的首要任務。點擊訊息廣告持續成長，我們看到廣告主越來越多地使用這些廣告來降低管道轉換。付費訊息傳遞正在快速成長，我們致力於在 2024 年擴大採用範圍，讓企​​業更容易購買行銷訊息，並更輕鬆地採用我們的 Flows 產品，為消費者開發更豐富的線程內體驗。

We're also progressing with our early testing on WhatsApp and Messenger of AIs for businesses that can provide conversational support in chat. We'll take our time to get these experiences right, but we think this will be a compelling opportunity that we're well positioned to deliver on. Outside of business messaging, we're making strong progress with our shops' ads products, which crossed a $2 billion annual run rate in Q4.

我們也正在對 WhatsApp 和 Messenger of AI 進行早期測試，為可以在聊天中提供對話支援的企業進行測試。我們將花時間來打造正確的體驗，但我們認為這將是一個令人信服的機會，我們有能力抓住這個機會。除了商業訊息傳遞之外，我們商店的廣告產品也取得了巨大進展，第四季的年運行率突破了 20 億美元。

Next, I would like to discuss our approach to capital allocation. We are in the fortunate position of being able to invest in many organic opportunities to improve the performance of our core business in the near term while also making 2 significant longer-time horizon investments in AI and Reality Labs.

接下來我想討論一下我們的資本配置方法。我們很幸運，能夠投資許多有機機會，以在短期內提高我們核心業務的績效，同時也在人工智慧和現實實驗室中進行兩項重大的長期投資。

AI is a growing area of investment for us in 2024 as we hire to support our road map. We are also investing more in our AI infrastructure capacity this year. And given many of our ambitious forward-looking plans will rely on having sufficient compute capacity, we expect this to be an area we invest more aggressively in over the coming years. Reality Labs is the other long-term initiative we continue to invest heavily in as we progress on our efforts to develop the next computing platform.

到 2024 年，隨著我們聘請員工來支援我們的路線圖，人工智慧將成為我們不斷成長的投資領域。今年我們還將加大對人工智慧基礎設施能力的投資。鑑於我們許多雄心勃勃的前瞻性計劃將依賴於擁有足夠的運算能力，我們預計這將是我們在未來幾年更加積極投資的領域。隨著我們在開發下一代運算平台方面取得進展，現實實驗室是我們繼續大力投資的另一個長期計劃。

Aside from organic investments, returning capital to shareholders remains important to us. We believe our strong financial position and performance will enable us to invest in the business while also continuing to return capital to investors over time. We've historically done so through share repurchases. And while we expect to maintain an active share repurchase program, we are modestly evolving our approach going forward by returning a portion of capital through a regular dividend subject to quarterly Board approval.

除了有機投資之外，向股東返還資本對我們來說仍然很重要。我們相信，我們強大的財務狀況和業績將使我們能夠投資該業務，同時隨著時間的推移繼續向投資者返還資本。我們歷來都是透過股票回購來做到這一點的。雖然我們預計將維持積極的股票回購計劃，但我們正在適度改進我們的方法，透過定期股息返還部分資本，但須經董事會季度批准。

In addition, we continue to monitor the active regulatory landscape, including the increasing legal and regulatory headwinds in the EU and the U.S. that could significantly impact our business and our financial results. Of note, the FTC is seeking to substantially modify our existing consent order and impose additional restrictions on our ability to operate. We are contesting this matter, but if we are unsuccessful, it would have an adverse impact on our business.

此外，我們繼續監控活躍的監管環境，包括歐盟和美國日益增加的法律和監管阻力，這可能會對我們的業務和財務表現產生重大影響。值得注意的是，聯邦貿易委員會正在尋求大幅修改我們現有的同意令，並對我們的營運能力施加額外的限制。我們正在就此事提出異議，但如果不成功，將對我們的業務產生不利影響。

Turning now to the revenue outlook. We expect first quarter 2024 total revenue to be in the range of $34.5 billion to $37 billion. Our guidance assumes foreign currency is neutral to year-over-year total revenue growth based on current exchange rates.

現在轉向收入前景。我們預計 2024 年第一季總營收將在 345 億美元至 370 億美元之間。我們的指導假設根據當前匯率，外幣對年比總收入成長是中性的。

Turning now to the expense outlook. We expect full year 2024 total expenses to be in the range of $94 billion to $99 billion, unchanged from our prior outlook. We continue to expect a few factors to be drivers of total expense growth in 2024.

現在轉向費用前景。我們預計 2024 年全年總支出將在 940 億美元至 990 億美元之間，與我們先前預期的相同。我們仍預期一些因素將成為 2024 年總費用成長的驅動因素。

First, we expect higher infrastructure-related costs this year. Given our increased capital investments in recent years, we expect depreciation expenses in 2024 to increase by a larger amount than in 2023. We also expect to incur higher operating costs from running a larger infrastructure footprint.

首先，我們預期今年基礎建設相關成本將會上升。鑑於我們近年來增加的資本投資，我們預計 2024 年的折舊費用將比 2023 年增加更多。我們也預計，運作更大的基礎設施足跡將產生更高的營運成本。

Second, we anticipate growth in payroll expenses as we work down our current hiring underrun and add incremental talent to support priority areas in 2024, which we expect will further shift our workforce composition toward higher-cost technical roles. Finally, for Reality Labs, we expect operating losses to increase meaningfully year-over-year due to our ongoing product development efforts in AR/VR and our investments to further scale our ecosystem.

其次，隨著我們解決當前的招聘不足問題並增加增量人才以支持 2024 年的優先領域，我們預計工資支出將會增長，我們預計這將進一步將我們的勞動力結構轉向成本更高的技術職位。最後，對於 Reality Labs 來說，由於我們在 AR/VR 方面持續進行的產品開發工作以及為進一步擴大生態系統而進行的投資，我們預計營運虧損將比去年同期大幅增加。

Turning now to the CapEx outlook. We anticipate our full year 2024 capital expenditures will be in the range of $30 billion to $37 billion, a $2 billion increase of the high end of our prior range. We expect growth will be driven by investments in servers, including both AI and non-AI hardware, and data centers as we ramp up construction on sites with our previously announced new data center architecture.

現在轉向資本支出前景。我們預計 2024 年全年資本支出將在 300 億至 370 億美元之間，比先前範圍的上限增加 20 億美元。我們預計，隨著我們利用先前宣布的新資料中心架構加大站點建設力度，對伺服器（包括人工智慧和非人工智慧硬體）以及資料中心的投資將推動成長。

Our updated outlook reflects our evolving understanding of our AI capacity demands as we anticipate what we may need for the next generations of foundational research and product development. While we are not providing guidance for years beyond 2024, we expect our ambitious long-term AI research and product development efforts will require growing infrastructure investments beyond this year.

我們更新的展望反映了我們對人工智慧能力需求不斷變化的理解，因為我們預測下一代基礎研究和產品開發可能需要什麼。雖然我們不會提供 2024 年後的指導，但我們預計，我們雄心勃勃的長期人工智慧研究和產品開發工作將需要今年之後不斷增加的基礎設施投資。

On to tax. Absent any changes to U.S. tax law, we expect our full year 2024 tax rate to be in the mid-teens. In closing, this was a pivotal year for our company. We increased our operating discipline, delivered strong execution across our product priorities and improved advertising performance for businesses who rely on our services. We will look to build on our priorities in each of those areas in 2024 while advancing our ambitious longer-term efforts in AI and Reality Labs.

到稅收。如果美國稅法沒有任何變化，我們預計 2024 年全年稅率將在十幾歲左右。最後，今年對我們公司來說是關鍵的一年。我們加強了營運紀律，在產品優先事項中提供了強有力的執行力，並為依賴我們服務的企業提高了廣告效果。我們將著眼於 2024 年在每個領域的優先事項基礎上再接再厲，同時推進我們在人工智慧和現實實驗室方面雄心勃勃的長期努力。

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) And your first question comes from the line of Brian Nowak from Morgan Stanley.

（操作員說明）您的第一個問題來自摩根士丹利的布萊恩·諾瓦克（Brian Nowak）。

I have 2. I appreciate, Mark, the color about sort of the learnings and the data feedback loop. Let me ask you one about sort of the advertising business and all of the new machine learning and new GenAI ad tools that you built and rolled out the last year or so. Can you just talk about some of the largest areas of improvement your advertisers say is really driving this inflection of advertising? And what are the areas where you still get the most feedback on for further areas of improvement in the ad business that we should think about?

我有 2。馬克，我很欣賞關於學習和數據回饋循環的顏色。讓我問您有關廣告業務的問題，以及您在去年左右建立和推出的所有新機器學習和新 GenAI 廣告工具。您能否談談您的廣告商所說的真正推動廣告這種變化的一些最大的改進領域？對於我們應該考慮的廣告業務進一步改進領域，您仍然獲得最多回饋的領域是什麼？

And then, Susan, maybe, one, the first quarter guide, specifically on the acceleration implied. Anything on -- more color on which surfaces or ad units or regions are sort of giving you confidence in that type of acceleration at this point in the quarter?

然後，蘇珊，也許，第一季指南，特別是關於暗示的加速。任何表面、廣告單元或區域上的更多顏色都可以讓您對本季此時的這種類型的加速充滿信心嗎？

Thank you, Brian. I can take both of those. On your first question, what is driving strength in the Ads business, we certainly believe that we've been continuing to drive ad performance improvements and year-over-year conversion growth remains strong. And we see this reflected in the feedback that we hear from advertisers also.

謝謝你，布萊恩。這兩樣我都可以接受。關於你的第一個問題，廣告業務的成長動力是什麼，我們當然相信我們一直在繼續推動廣告效果的改進，並且同比轉換率成長仍然強勁。我們從廣告商那裡聽到的回饋中也看到了這一點。

There are really 3 areas of our work here that we're very focused on and have been and will continue to be in 2024. First, just creating engaging on-platform ad experiences. We see that with formats like click-to-message ads, where we continue to see good momentum. Shops ads, we talked about the $2 billion annual run rate in Q4 after we just opened availability to all U.S. advertisers in Q2. Lead generation ads, another example. So we're really focused on creating engaging on-platform ad experiences.

我們實際上非常關注 3 個工作領域，並且一直並將在 2024 年繼續關注。首先，創造引人入勝的平台內廣告體驗。我們透過點擊訊息廣告等格式看到了這一點，我們繼續看到良好的勢頭。商店廣告，我們在第二季度剛剛向所有美國廣告商開放後，我們談到了第四季度 20 億美元的年運行率。另一個例子是潛在客戶開發廣告。因此，我們真正專注於創造引人入勝的平台廣告體驗。

Second, we're really making it easier for advertisers to connect with their marketing data. We're continuing to invest in features like conversions API, like AEM and making these features easier to adopt, enhancing reporting and other performance. And we've seen, again, that those are -- we get very positive feedback on advertisers who use those tools.

其次，我們確實讓廣告主更輕鬆地連接他們的行銷數據。我們將繼續投資 AEM 等轉換 API 等功能，並使這些功能更易於採用，從而增強報告和其他效能。我們再次看到，我們從使用這些工具的廣告商那裡得到了非常正面的回饋。

And then finally, continuing to really use AI in important ways across our ads platform. For a long time, we have invested in building larger and more advanced models that have resulted in more Accurate predictions of relevant ads for people and improved performance for advertisers.

最後，繼續在我們的廣告平台上以重要方式真正使用人工智慧。長期以來，我們一直投資於建立更大、更先進的模型，從而為人們提供更準確的相關廣告預測，並提高廣告商的效果。

And then, of course, we're investing a lot in AI-powered tools and products. So we're really scaling our Advantage+ suites across all of the different offerings there, which really helped to automate the ads creation process for different types of advertisers. And we're getting very strong feedback on all of those different features, advantage+ Shopping, obviously, being the first, but Advantage+ Catalog, Advantage+ Creative, Advantage+ Audiences, et cetera. So we feel like these are all really important parts of what has continued to grow improvements in our Ads business and will continue to going forward.

當然，我們在人工智慧驅動的工具和產品上投入了大量資金。因此，我們確實在所有不同的產品中擴展我們的 Advantage+ 套件，這確實有助於為不同類型的廣告商實現廣告創建流程的自動化。我們從所有這些不同的功能中得到了非常強烈的回饋，優勢+購物顯然是第一個，但優勢+目錄、優勢+創意、優勢+受眾等等。因此，我們認為這些都是我們廣告業務持續改進的非常重要的部分，並將繼續向前發展。

On your second question, which is about the Q1 year-over-year -- sorry, the Q1 revenue guide. This really just reflects a lot of the trends we saw in Q4, which is strong, broad-based advertising demand across verticals, particularly within online commerce and gaming. I'll also note that we get the benefit from having February 29 in this quarter. And again, just the improvements that we continue to accrue to the business from all of our investments in improving ad performance over time.

關於你的第二個問題，這是關於第一季的同比情況——抱歉，第一季的收入指南。這實際上反映了我們在第四季度看到的許多趨勢，即跨垂直行業的強勁、廣泛的廣告需求，特別是在線商務和遊戲領域。我還要指出的是，我們從本季的 2 月 29 日中獲益匪淺。再說一遍，隨著時間的推移，我們在提高廣告效果方面的所有投資不斷為業務帶來改進。

Your next question comes from the line of Eric Sheridan from Goldman Sachs.

您的下一個問題來自高盛的艾瑞克·謝裡丹。

Maybe 2 questions, if I can. Mark, coming out of a year like you had with the year of efficiency, what do you see as the key messages you want to share with investors about what you learned during the year of efficiency and how it might inform running an organization with the scale of ambition and just your organizational scale that Meta has on a long-term basis? That would be number one.

如果可以的話，也許有兩個問題。馬克，就像你經歷的效率年一樣，你認為你想與投資者分享的關鍵信息是什麼，關於你在效率年中學到的東西，以及它如何為運營一個規模化的組織提供信息Meta 的長期目標和組織規模？那將是第一。

And then Susan, I went back and looked, I don't think you've ever raised the high end of your CapEx guidance before, unless I'm mistaken. But with a wide range like that in CapEx, what should we be monitoring for what would push you towards either the lower end or the higher end of that CapEx guidance as you move through 2024?

然後蘇珊，我回頭看了看，我認為你以前從未提高過資本支出指導的上限，除非我弄錯了。但是，隨著資本支出的範圍如此廣泛，我們應該監控什麼，以便在 2024 年推動您走向資本支出指導的低端或高端？

I'll take the first one. So the themes for the year of efficiency were to make us a stronger technology company by becoming leaner and more balanced towards our engineering work and more streamlined and to improve our financial performance, primarily with the goal of providing stability so we can invest in these long-term, ambitious visions around AI and metaverse over what we see as the coming decade or more as these things play out.

我就拿第一個。因此，效率年的主題是透過使我們的工程工作更加精簡、更加平衡，更加精簡，使我們成為一家更強大的科技公司，並提高我們的財務業績，主要目標是提供穩定性，以便我們可以投資於這些長期項目。- 術語，圍繞人工智能和元宇宙的雄心勃勃的願景，我們認為未來十年或更長時間，隨著這些事情的發展。

And I think a lot of people looked at what we were doing as if it might have been some kind of short-term thing, which that's never really our focus. But the part about making the company leaner, I think, is the more important part to take forward, right? Because obviously, we're in a place now where the business is performing well. And I think the obvious question would be, okay, well, given that, should we just -- should we invest a lot more in things?

我認為很多人看待我們正在做的事情就好像這可能是某種短期的事情，而這從來都不是我們真正關注的焦點。但我認為，讓公司變得更精簡的部分是更重要的部分，對吧？因為顯然，我們現在的業務表現良好。我認為顯而易見的問題是，好吧，考慮到這一點，我們是否應該——我們是否應該在事物上投入更多？

And the biggest thing that's holding me back from doing that is that at this point, I feel like I've really come around to thinking that we operate better as a leaner company. So even though it's always -- there's always questions about like adding a few people here or there to do something. And I guess I just have more of an appreciation about how all of that adds up and in the near term, maybe makes you go a little bit faster. But over the long term, the discipline to kind of hold things to a more streamlined level actually improves the overall company performance. So I'm really focused on that.

阻礙我這樣做的最大的事情是，在這一點上，我覺得我真的開始認為我們作為一家更精簡的公司運作得更好。因此，儘管總是存在這樣或那樣的問題，例如在這裡或那裡添加一些人來做某事。我想我只是更欣賞所有這些如何加起來，在短期內，也許會讓你走得更快一點。但從長遠來看，將事情保持在更精簡的水平的紀律實際上可以提高公司的整體績效。所以我真的很專注於此。

In 2024, we do have a big recruiting backlog from last year because part of the layoffs that we did included teams basically swapping out certain talent profiles for others. And we still need to hire some of the other talent profiles that we swapped people out for, and that will be ongoing through this year. But in terms of new head count that we added to the plan, it's relatively minimal compared to what we would have done historically.

到 2024 年，我們確實有去年的大量招募積壓，因為我們所做的部分裁員包括團隊基本上用其他人才取代了某些人才。我們仍然需要聘請一些我們換掉的其他人才，這將持續到今年。但就我們添加到計劃中的新員工人數而言，與我們歷史上所做的相比，這相對較少。

And I sort of expect that for the next period of time going forward, Even beyond 2024. My operating assumption is that we will also try to keep it relatively minimal because I think that -- until we reach a point where we're just really underwater on our ability to execute, I kind of want to keep things lean because I think that's the right thing for us to do culturally. So that's kind of the best window into how I'm thinking about this.

我預計在接下來的一段時間內，甚至在 2024 年之後。我的營運假設是，我們也會盡力將其保持在相對最低限度，因為我認為 - 直到我們真正達到一個點在我們執行能力的水下，我有點想保持精簡，因為我認為這對我們的文化來說是正確的事情。所以這是了解我如何思考這個問題的最佳窗口。

The other piece that, I guess, somewhat dovetails into what Susan is going to talk to you next is that a big part of why I wanted to improve our profitability is to give ourselves the ability to go through what is a somewhat unpredictable and volatile period over the next 5 or 10 years. There are different risk factors that are geopolitical or regulatory or different things, but also the technology landscape is somewhat unknown. And we want the ability to be able to surge investment on things like building out larger training clusters or just making different investments where that's necessary.

我想，與蘇珊接下來要和你談論的內容有些吻合的是，我想提高我們的盈利能力的一個重要原因是讓我們自己有能力度過一個有點不可預測和不穩定的時期未來5年或10年。存在地緣政治或監管或不同因素等不同的風險因素，而且技術前景也有些未知。我們希望能夠增加對諸如建立更大的培訓集群或在必要時進行不同投資等方面的投資。

And in order to make sure that we have the flexibility to do that, we want to make sure that we keep our cost structure to a point where we sort of have some extra space built in. So those are really the 2 points. That was the theme that I laid out at the beginning of the year of efficiency last year: make us a stronger technology company and give us the flexibility and stability to execute the long-term goals. And those remain, I think, the big focuses going forward on that.

為了確保我們能夠靈活地做到這一點，我們希望確保我們的成本結構保持在一個可以內建一些額外空間的程度。所以這實際上是兩點。這就是我在去年效率年初提出的主題：讓我們成為一家更強大的科技公司，並賦予我們執行長期目標的靈活性和穩定性。我認為，這些仍然是未來的重點。

And I can take the second question, Eric. So we're really continuing to evolve our AI road maps and ambitions and our understanding of the capacity demands that we might have as we train next generations of foundation models and to support all of the associated product development going forward. So the increase in the top end of the range really reflects that evolving understanding of how much demand we may need. And we're also continuing to keep a close eye on supply availability.

我可以回答第二個問題，艾瑞克。因此，我們確實在繼續發展我們的人工智慧路線圖和雄心，以及我們對在訓練下一代基礎模型並支援未來所有相關產品開發時可能擁有的能力需求的理解。因此，該範圍高端的增加確實反映了我們對可能需要多少需求的不斷變化的理解。我們也將繼續密切關注供應情況。

Where we land in that range is a function of both the supply and demand factors I mentioned. And this continues to be a pretty dynamic planning process for us. And there are also certainly other factors that drive uncertainty. How quickly can we execute on the new data center architecture? How the supply chain turns out to unfold over the course of the year? But our expectation is, generally, that we will need to invest more to support our AI work in the years ahead, and we're seeing some of that reflected in 2024.

我們在這個範圍內的位置是我提到的供給和需求因素的函數。對我們來說，這仍然是一個非常動態的規劃過程。當然還有其他因素會帶來不確定性。我們能夠以多快的速度執行新的資料中心架構？一年來供應鏈的發展如何？但總體而言，我們的預期是，未來幾年我們需要投入更多資金來支持我們的人工智慧工作，我們將在 2024 年看到其中的一些體現。

Your next question comes from the line of Mark Shmulik from AllianceBernstein.

您的下一個問題來自 AllianceBernstein 的 Mark Shmulik。

Mark, you mentioned at the top of the call this vision, kind of everyone having a Meta AI assistant to kind of get things done. Previously, when we were talking about the metaverse from a time horizon perspective, we were looking at a decade to kind of get there. But given kind of the pace of innovation around AI that you've seen, has that time line changed? And like do you think we'll get there quicker?

馬克，你在電話會議的頂部提到了這個願景，即每個人都擁有一個元人工智慧助理來完成工作。在此之前，當我們從時間範圍的角度討論虛擬宇宙時，我們考慮的是十年後才能實現這一目標。但考慮到您所看到的人工智慧創新步伐，時間軸是否發生了變化？你認為我們會更快到達那裡嗎？

And then second question for Susan. Just quickly on shop ad. I appreciate the color. Obviously, we've had the Amazon partnership news intra-quarter. We know Meta has made a few attempts that are getting kind of more integrated into shopping over the last few years. Can you hear some of the learnings and, I guess, what's different this time around?

然後是蘇珊的第二個問題。就快點上商店廣告吧。我很欣賞它的顏色。顯然，我們已經在季度內收到了亞馬遜合作的消息。我們知道 Meta 已經做出了一些嘗試，在過去的幾年裡，這些嘗試越來越融入購物。您能聽到一些經驗教訓嗎？我想，這次有什麼不同？

Sure. I can take the first one. I do think that AI is going to make all of the products and services that we use and make better. So it's hard to know exactly how that will play out. But for the work in Reality Labs, specifically, there's a bunch of areas. Like if you take smart glasses, before, we thought that we would have to build like full displays and holograms and deliver the sense of presence before that became a mainstream product.

當然。我可以拿第一個。我確實認為人工智慧將使我們使用和製造的所有產品和服務變得更好。所以很難確切地知道結果會如何。但對於現實實驗室的工作來說，具體來說，有許多領域。就像智慧眼鏡一樣，之前我們認為我們必須建造完整的顯示器和全像圖，並在成為主流產品之前提供臨場感。

And now it seems quite possible that smart glasses that have AI assistance built in will be the killer app, and that the holograms and sense of presence will come later as a -- maybe on the same time horizon we were talking about before but could end up being just as important as we expected. But there could be a big market here even before that. So I think we'll figure that out over the next few years. But yes, I mean, overall, I think the 2 go together pretty hand-in-hand. I would have predicted that a lot of the parts of Reality Labs -- I guess the sequence in technology is sometimes surprising.

現在看來，內建人工智慧輔助的智慧眼鏡很有可能成為殺手級應用，而全息圖和存在感將在稍後出現——也許與我們之前討論的時間範圍相同，但可能會結束正如我們預期的那樣重要。但在此之前，這裡可能會有一個很大的市場。所以我認為我們會在未來幾年內解決這個問題。但是，是的，我的意思是，總的來說，我認為兩者是緊密結合在一起的。我本來可以預測現實實驗室的許多部分——我想技術的順序有時會令人驚訝。

We always kind of expected that as part of building glasses or any of these platforms that having an AI assistant would be a foundational part of it. But the fact that that's -- that there have just been big strides in that over the last year or 2 is just -- is a big opportunity for these products sooner. So I think it's still unknown exactly how that will play out, but I think we'll know more over the next few years. It's very exciting, though.

我們總是期望，作為建造眼鏡或任何這些平台的一部分，人工智慧助理將成為其基礎部分。但事實上，在過去的一兩年裡，這方面剛剛取得了重大進展，這對這些產品來說是一個巨大的機會。所以我認為目前還不清楚到底會如何發展，但我想我們在未來幾年會知道更多。不過，這非常令人興奮。

And Mark, I can take the second question. So first of all, online commerce continues to be one of our strongest verticals overall. And a lot of the work that we do to improve ads performance generally really helps do things like grow conversions year-over-year, improves the performance of direct response ads. We've invested a lot in the Advantage+ set of tools, tools like Advantage+ Shopping. So I would say that our offerings for e-commerce advertisers overall are very strong, and we continue to invest in making them better and more performant.

馬克，我可以回答第二個問題。首先，線上商務仍然是我們整體上最強大的垂直行業之一。我們為提高廣告效果所做的許多工作通常確實有助於實現逐年增加轉換率、提高直接回應廣告的效果等。我們在 Advantage+ 工具集上投入了大量資金，例如 Advantage+ Shopping 等工具。因此，我想說，我們為電子商務廣告商提供的產品總體而言非常強大，我們將繼續投資，使它們變得更好、更有效能。

On Shops ads, specifically, this is an area where we are really, I would say, focusing on introducing improvements that continue to make it easier for businesses to onboard 2 shops. So for example, eligible Shopify businesses can now onboard to Shops on Facebook and Instagram very seamlessly. And we're making it easier for advertisers to turn their existing ads into Shops ads. And we'll continue to focus on deepening integrations with partners and leveraging AI to make Shops ads even more performant.

具體來說，在商店廣告方面，我想說的是，我們確實專注於引入改進，並繼續讓企業更輕鬆地開設 2 家商店。例如，符合資格的 Shopify 企業現在可以非常無縫地加入 Facebook 和 Instagram 上的商店。我們正在幫助廣告客戶更輕鬆地將其現有廣告轉變為商店廣告。我們將繼續專注於深化與合作夥伴的集成，並利用人工智慧來提高商店廣告的效果。

You mentioned the Amazon ads pilot, and that's a place where we're partnering with Amazon to make Buy with Prime, to create a more seamless shopping experience on Facebook and Instagram so that it's easier for people to purchase directly from an Amazon ad. This is really early, and we're testing this experience with them. But it's another avenue for us to explore how can we make these shopping experience easier for folks on our platform.

您提到了亞馬遜廣告試點，我們正在與亞馬遜合作推出「Buy with Prime」服務，在 Facebook 和 Instagram 上打造更無縫的購物體驗，以便人們更輕鬆地直接透過亞馬遜廣告進行購買。現在還為時過早，我們正在與他們一起測試這種體驗。但這是我們探索如何為我們平台上的人們提供更輕鬆的購物體驗的另一個途徑。

Your next question comes from the line of Justin Post from Bank of America.

您的下一個問題來自美國銀行的賈斯汀·波斯特。

I was wondering if you could help us think about the messaging run rate for revenues and maybe the long-term opportunity and how you think about that going forward on a multiyear basis. And then secondly, the guidance for Q1 suggests stable growth at the midpoint despite comps getting tougher. So I'm wondering if you could help us at all think about -- the comps get increasingly tougher as we go through the year, how you're thinking about those comps and how you could grow in the second half.

我想知道您是否可以幫助我們考慮收​​入的消息傳遞運行率，也許還有長期機會，以及您如何看待未來多年的發展。其次，第一季的指導表明，儘管競爭變得更加艱難，但中點仍將保持穩定成長。所以我想知道你是否可以幫助我們思考——隨著這一年的過去，競爭變得越來越艱難，你如何看待這些競爭以及下半年如何成長。

Thanks, Justin. I can take both of those. So on your first question about messaging monetization, right now, the 2 primary avenues that we have there are click-to-messaging ads and paid messaging, which we are investing a lot in both to facilitate the full cycle of customer engagement. We saw that revenue growth from click-to-message ads remained strong in Q4. We see broadening advertiser adoption. And there, I would say that our focus in terms of where we're investing is really around enhanced optimization and reporting. So we're really trying to drive down funnel performance, enabling businesses to optimize for purchase.

謝謝，賈斯汀。這兩樣我都可以接受。因此，關於您關於訊息傳遞貨幣化的第一個問題，目前我們擁有的兩個主要途徑是點擊訊息傳遞廣告和付費訊息傳遞，我們在這兩個方面都進行了大量投資，以促進客戶參與的整個週期。我們看到第四季點擊訊息廣告的營收成長仍然強勁。我們看到廣告商的採用範圍不斷擴大。在那裡，我想說，我們投資的重點實際上是圍繞著增強的優化和報告。因此，我們確實在努力降低通路績效，使企業能夠優化購買。

And we've seen strong growth in purchase optimization revenue on click-to-Messenger ads since we've rolled that out. And we're planning to bring purchase optimization to click-to-WhatsApp campaigns this year.

自從我們推出點擊信差廣告以來，我們發現其購買優化收入出現了強勁成長。我們計劃今年為點擊 WhatsApp 活動帶來購買優化。

And across the click-to-messaging ads, we're introducing more robust reporting to help advertisers kind of understand how people are engaging with them via click-to-messaging ads and what is the value of that engagement. And then, of course, in the longer term, this is a place where we're excited about the opportunities for increased automation to really help businesses scale their ability to have conversations with their consumers.

在點擊訊息廣告中，我們引入了更強大的報告，以幫助廣告商了解人們如何透過點擊訊息廣告與他們互動，以及這種互動的價值是什麼。當然，從長遠來看，我們對提高自動化的機會感到興奮，這些機會可以真正幫助企業擴大與消費者對話的能力。

On the paid messaging side, this is much earlier, of course, but we're seeing good momentum, driven by particularly strong growth from marketing messages. And we've also seen good results from our updated pricing model that we introduced in June. And here, we're really focused on making paid messaging easier to buy. So we're now testing the ability for businesses to send paid messaging directly on Meta Ads Manager.

當然，在付費消息方面，這要早得多，但在行銷消息特別強勁的成長推動下，我們看到了良好的勢頭。我們也從 6 月推出的更新定價模式中看到了良好的結果。在這裡，我們真正專注於讓付費訊息更容易購買。因此，我們現在正在測試企業直接在 Meta Ads Manager 上發送付費訊息的能力。

And then, again, similar to click-to-messaging, we are investing in ways to make it easier to get more done within the app. So here, we launched Flows globally in October, which enables businesses to offer richer user experiences within WhatsApp, things like selecting a seat on a flight or booking an appointment. We're seeing good early traction, and we're focused on just introducing more ways to help businesses easily create flows, and we'll plan on introducing more capabilities here. So there's a lot, I would say, going on in the portfolio of messaging monetization work that we're doing.

然後，與點擊訊息傳遞類似，我們正在投資各種方法，以便更輕鬆地在應用程式內完成更多工作。因此，我們在 10 月在全球推出了 Flows，使企業能夠在 WhatsApp 中提供更豐富的用戶體驗，例如選擇航班座位或預訂約會。我們看到了良好的早期吸引力，我們專注於引入更多方法來幫助企業輕鬆創建流量，我們將計劃在這裡引入更多功能。因此，我想說，我們正在做的消息貨幣化工作組合中正在進行很多工作。

Now your second question was around the 2024 outlook. And we obviously have not shared any guidance beyond Q1. But our revenue for the full year is going to be influenced by a number of factors, including macro conditions that are certainly harder to predict the further out you go. And over the course of 2024, as you said, we'll also be lapping periods of increasingly strong demand. So we don't provide guidance beyond Q1. There's certainly a range of outcomes, and we'll have a better read on how we will compare to future quarters as we progress through the year.

現在你的第二個問題是關於 2024 年的展望。顯然，我們沒有分享第一季之後的任何指導。但我們全年的收入將受到許多因素的影響，包括宏觀條件，隨著時間的推移，宏觀條件肯定更難預測。正如您所說，在 2024 年，我們也將經歷需求日益強勁的時期。因此，我們不會提供第一季之後的指導。當然會有一系列結果，隨著今年的進展，我們將更好地了解我們將如何與未來幾季進行比較。

Your next question comes from the line of Doug Anmuth from JPMorgan.

您的下一個問題來自摩根大通的 Doug Anmuth。

One for Mark, one for Susan. Mark, I believe you'd shifted FAIR out of Reality Labs into the Family of Apps. Can you just talk about some of the benefits there of moving that group as you pursue general intelligence? And then Susan, can you just walk us through your thought process on adding a dividend to your capital returns at this stage beyond the share repurchases?

一份給馬克，一份給蘇珊。馬克，我相信您已將 FAIR 從現實實驗室轉移到應用程式家族。您能否談談在追求通用智慧時移動該團隊的一些好處？蘇珊，您能否向我們介紹一下您在現階段除股票回購之外向資本回報增加股息的思考過程？

I can talk to the first one. Yes, the whole reason why we moved FAIR is basically to be closer to the GenAI group. They're both research groups. The GenAI group basically builds our Llama launch vehicles and products, but also conducts a fair amount of research, too, especially things that are going to be coming into the upcoming versions of Llama. FAIR is focused on more foundational work and longer-term work. So it's, I'd say, more things that might be a couple of years out to 10-plus years out. And as we -- I guess, here's one way to think about it.

我可以和第一個交談。是的，我們移動 FAIR 的全部原因基本上是為了更接近 GenAI 團隊。他們都是研究小組。 GenAI 小組主要建造我們的 Llama 運載火箭和產品，但也進行了大量的研究，特別是即將推出的 Llama 版本中的內容。 FAIR專注於更基礎的工作和更長期的工作。所以，我想說，更多的事情可能需要幾年甚至十幾年的時間才能完成。我想，我們可以用一種方式來思考這個問題。

A lot of last year and the work that we're doing with Llama 3 is basically making sure that we can scale our efforts to really produce state-of-the-art models. But once we get past that, there's a lot more kind of different research that I think we're going to be doing that's going to take our foundation models in potentially different directions than other players in the industry are going to go in because we're focused on specific vision for what we're building.

去年我們在 Llama 3 上所做的許多工作基本上都是為了確保我們能夠擴大我們的努力，真正生產出最先進的模型。但一旦我們克服了這一點，我認為我們將進行更多不同的研究，這將使我們的基礎模型朝著與行業其他參與者可能不同的方向發展，因為我們」我們專注於我們正在建立的具體願景。

So it's really important as we think about what's going to be in Llama 5 or 6 or 7 and what cognitive abilities we want in there and what modalities we want to build into future multimodal versions of the models. We need to be doing that work in advance and to research those things. And it helps to -- and even though FAIR and GenAI will continue to be 2 kind of separate groups on different time horizons.

因此，當我們思考 Llama 5、6 或 7 中將包含什麼以及我們想要其中的認知能力以及我們希望在模型的未來多模式版本中建立哪些模式時，這一點非常重要。我們需要提前做好這項工作並研究這些事情。儘管 FAIR 和 GenAI 在不同的時間範圍內仍將是兩種不同的團體，但它還是有幫助的。

I think to have some level of alignment between -- on the vision of what we're building between the 2 of them, so that way, the FAIR team can have in mind, hey, if we research this, then maybe it can intercept Llama 6 or something. Then that, I think, is just going to be more helpful for increasing the ambition and focus of all of the work that we do.

我認為在我們兩者之間建立的願景上有一定程度的一致性，這樣，FAIR 團隊就可以記住，嘿，如果我們研究這個，那麼也許它可以攔截駱駝6什麼的。我認為，這將更有利於增強我們所做的所有工作的雄心和重點。

It's one of the reasons that I talked about. We did open-ended research on AI for a while. But having a clear product target with these AI agents, I think, is really going to help focus the work and give us a feedback loop that's going to increase the productivity and output that we get dramatically.

這也是我剛才談到的原因之一。我們對人工智慧進行了一段時間的開放式研究。但我認為，為這些人工智慧代理商制定明確的產品目標確實有助於集中工作，並為我們提供反饋循環，從而大幅提高我們的生產力和產出。

Doug, on your second question about why now in terms of initiating a dividend, returning capital to shareholders remains an important priority for us. And we believe introducing a dividend really just serves as a nice complement to the existing share repurchase program. The dividend doesn't change how much we will be -- or it doesn't change the way we determine the total amount of capital we return. And we expect that share repurchases will continue to be the primary way that we return capital to shareholders. But introducing a dividend just gives us a more balanced capital return program and some added flexibility in how we return capital in the future.

道格，關於你的第二個問題，為什麼現在在啟動股息方面，向股東返還資本仍然是我們的重要優先事項。我們相信引入股息實際上只是對現有股票回購計劃的一個很好的補充。股利不會改變我們將獲得的金額，也不會改變我們確定返還資本總額的方式。我們預期股票回購將繼續成為我們向股東返還資本的主要方式。但引入股息只是為我們提供了更平衡的資本回報計劃，並為我們未來的資本回報方式增加了一些靈活性。

Your next question comes from the line of Ron Josey from Citi.

您的下一個問題來自花旗集團的 Ron Josey。

Mark, I hope your knee is getting better. I wanted to ask you about Apple opening up its App Store in Europe and maybe what role that might play for Meta in further building out the App Store app install ads and things along those lines. And Susan, as a follow-up to maybe a prior question, you mentioned the potential around business messaging and the team -- what the team is doing around really testing with WhatsApp and Messenger. Just talk just how you might see this unfold maybe with AI Studio and maybe timing would be helpful.

馬克，我希望你的膝蓋好起來。我想問一下蘋果在歐洲開放 App Store 的情況，以及 Meta 在進一步建立 App Store 應用程式安裝廣告等方面可能發揮什麼作用。蘇珊，作為先前問題的後續，您提到了業務訊息和團隊的潛力 - 團隊正在圍繞 WhatsApp 和 Messenger 進行實際測試。只需談談您可能如何透過 AI Studio 看到這一點，也許時機會有所幫助。

I don't think that the Apple thing is going to have any difference for us because I think that the way that they've implemented it, I would be very surprised if any developer chose to go into the alternative App Stores that they have. They've made it so onerous, and I think so at odds with the intent of what the EU regulation was that I think it's just going to be very difficult for anyone, including ourselves, to really seriously entertain what they're doing there.

我不認為蘋果的事情對我們有任何區別，因為我認為他們實現它的方式，如果任何開發人員選擇進入他們擁有的替代應用商店，我會感到非常驚訝。他們把它變得如此繁重，我認為這與歐盟法規的意圖不一致，我認為任何人，包括我們自己，都很難真正認真地對待他們在那裡所做的事情。

Susan, did you want to add anything?

蘇珊，你想補充什麼嗎？

No. I was going to take the second question, I think, which was around the rollout of some of the GenAI features from a monetization perspective. And we're -- first of all, I would say that we don't expect our GenAI products to be a meaningful 2024 driver of revenue. But we certainly expect that they will have the potential to be meaningful contributors over time.

不，我想我要回答第二個問題，這是從貨幣化角度推出一些 GenAI 功能的問題。首先，我想說的是，我們預計我們的 GenAI 產品不會成為 2024 年有意義的收入驅動因素。但我們當然希望隨著時間的推移，他們將有潛力成為有意義的貢獻者。

Right now, the most near-term monetization opportunity is with our ad creative tools. We've talked about some of the places that we've been rolling those features out broadly with Text Variations and Image Expansion now globally available. And we're seeing those tools receive adoption and provide real value for advertisers even at this early stage. And we're going to keep investing in making those more useful and performant, which then leads to more advertiser adoption, and that can be sort of a virtuous feedback loop.

目前，最近的獲利機會是我們的廣告創意工具。我們已經討論了我們已經在全球廣泛推出這些功能的一些地方，其中包括文字變體和圖像擴展。我們看到這些工具得到了採用，即使在早期階段也為廣告商提供了真正的價值。我們將繼續投資，使這些技術更有用、更有高效，從而吸引更多廣告商採用，這可能是一種良性回饋循環。

Over a longer time frame, the GenAI features that we're bringing to business messaging, which I think was where your question originated, we think really represents a compelling opportunity. We're testing AI chats on a very small scale today with a few businesses, but it will take time to continue making those AIs increasingly useful. We're hearing good feedback from the businesses testing them, and we expect to make progress this year as we expand our tests further.

從長遠來看，我們為商業訊息傳遞帶來的 GenAI 功能，我認為這就是您的問題的起源，我們認為這確實代表了一個引人注目的機會。今天，我們正在與一些企業進行小規模的人工智慧聊天測試，但要繼續讓這些人工智慧越來越有用還需要時間。我們從測試它們的企業那裡聽到了良好的反饋，隨著我們進一步擴大測試，我們預計今年會取得進展。

I'd say one other thing is just on the consumer side, again, we think GenAI can make it easier for people to produce compelling content across all of our apps, including our messaging apps. And our AI assistant certainly provides also added utilities. So in all of these areas, I would say, we're following our typical playbook of testing and tuning the experiences until they're very good, and then we'll invest in growth and eventually explore monetization.

我想說的另一件事是在消費者方面，我們認為 GenAI 可以讓人們更輕鬆地在我們的所有應用程式（包括我們的訊息應用程式）中製作引人注目的內容。我們的人工智慧助理當然也提供了額外的實用程式。因此，我想說，在所有這些領域，我們都遵循測試和調整體驗的典型策略，直到它們非常好，然後我們將投資於成長並最終探索貨幣化。

Your last question comes from the line of Ross Sandler from Barclays.

你的最後一個問題來自巴克萊銀行的羅斯桑德勒。

Mark, just curious what you're seeing with Meta AI at this stage. You talked about rapid iteration of that, given your large audience. So yes, I guess, what's going on, thus far? Which of the apps in the Family of Apps have seen engagement increase from Meta AI? And do you think that there could be increased commercial activity as either your AI agent or some of the other ones out there get more usage within your apps?

馬克，只是好奇你現階段在 Meta AI 上看到了什麼。您談到了考慮到大量觀眾的快速迭代。所以是的，我想，到目前為止發生了什麼？應用程式系列中哪些應用程式的參與度因 Meta AI 而有所提高？您是否認為隨著您的人工智慧代理或其他一些代理在您的應用程式中得到更多使用，商業活動可能會增加？

Yes. I think having a good assistant is going to be one of the real values that this generation of AI creates as well as giving every creator an opportunity to have an assistant or agent that people can engage with and every business and agent and also allowing people to create a bunch of quirky and fun things as well. But yes, I think Meta AI is going to be very important across the product.

是的。我認為擁有一個好的助手將成為這一代人工智慧創造的真正價值之一，同時也讓每個創造者都有機會擁有一個人們可以接觸的助手或代理，每個企業和代理也讓人們能夠還創造了一堆古怪而有趣的東西。但是，是的，我認為元人工智慧在整個產品中將非常重要。

It currently is available in some countries in WhatsApp, Messenger and Instagram. It's at the phase where we're not really pushing it super proactively. We've sort of made it available, and as people use it, we're learning from how -- what are the basic ways that people want to engage with it. We're currently in the tuning phase on this. I mean, we started rolling this out, I think, it was in November or so, maybe October.

目前，該功能可在部分國家的 WhatsApp、Messenger 和 Instagram 中使用。目前我們還沒有真正超級積極地推動它。我們已經提供了它，當人們使用它時，我們正在學習人們想要使用它的基本方法。我們目前正處於對此的調整階段。我的意思是，我們開始推出這個，我想，是在 11 月左右，也許是 10 月。

And I would expect that, over the course of this year, we are going to start rolling this out much more prominently across our apps. And that was sort of what I was saying in my opening remarks about the analogy to Stories, where we ran a bunch of experiments, put it out there, got feedback and then eventually did a lot of integrations, even to the point where people were making jokes about us putting Stories in settings, which, for the record, never happened, but I found funny.

我預計，在今年，我們將開始在我們的應用程式中更加突出地推出這項功能。這就是我在開場白中所說的關於故事的類比，我們進行了一系列實驗，將其放在那裡，獲得反饋，然後最終進行了大量集成，甚至達到了人們所關注的程度。開玩笑說我們把故事放在場景中，根​​據記錄，這從未發生過，但我覺得很有趣。

So I think that we're going to be on that journey this year. We're basically in the learning and tuning phase now. I'm happy with how that's going. We're currently working on and planning the next set of integrations and places where this is going to be available to people across the products, and I'm really looking forward to that. And I think that's going to be one of the big themes for '24 for us.

所以我認為今年我們將踏上這趟旅程。我們現在基本上處於學習和調整階段。我對事情的進展感到滿意。我們目前正在研究和規劃下一組整合以及可供跨產品的人使用的位置，我真的很期待這一點。我認為這將成為我們 24 世紀的一大主題。

Great. Thank you for joining us today. We appreciate your time, and we look forward to speaking with you again soon.

偉大的。感謝您今天加入我們。感謝您抽出寶貴的時間，我們期待很快再次與您交談。

This concludes today's conference call. Thank you for your participation, and you may now disconnect.

今天的電話會議到此結束。感謝您的參與，您現在可以斷開連接。