Amazon.com, Inc. (AMZN.US) conference call: AI is a "once in a lifetime" business opportunity. Capital expenditure may reach $100 billion by 2025. AWS growth is constrained by supply chain limitations.

Amazon.com, Inc. (AMZN.US) announced its financial report after hours on Thursday, showing that while holiday season performance in the fourth quarter exceeded expectations, the outlook for the first quarter was not promising, leading to a more than 7% drop in the company's stock price after hours. At the same time, Amazon.com, Inc.'s capital expenditures in the fourth quarter were significantly higher than analysts' expectations and set a record high. The company's capital expenditures for 2025 are projected to reach $100 billion, indicating that Amazon.com, Inc. is increasing spending to support AI business development. Key points summarized from the financial report conference call: 1. Capital expenditures in 2025 are expected to reach $100 billion, the majority of which will be used for AI and cloud service AWS: The fourth quarter's capital expenditure level of $27.8 billion is expected to be "relatively representative" for 2025. Based on this trend, capital expenditures for 2025 are estimated to be around $100 billion. The "vast majority" of this capital expenditure will be used for AI and cloud service AWS. Similar to 2024, most of the expenditure will be used to meet the demands for AI services and support the technological infrastructure of North American and international businesses. 2. Cost effectiveness will drive AI demand: The cost of inference will decrease, which is a good thing for tech companies. When the costs of technology components like AI inference decrease, it does not mean that overall spending by companies on technology will decrease. Instead, there will be excitement about building new things that were previously costly, ultimately leading to increased spending on technology. This trend is expected to increase overall demand for AI. 3. Massive spending planned only when demand is seen: CEO Jeff Bezos stated that we will not make purchases without significant demand signals. When AWS expands its capital expenditures, especially in rare business opportunities like AI, I think this is a very good signal for the long-term development of the AWS business. In fact, I believe it is wise to invest this capital to pursue this opportunity because we believe that almost all applications we know today will be reinvented with AI integrated, and inference will become a core building block like computing, storage, and databases. AI is undoubtedly the biggest opportunity since cloud computing and possibly the biggest technological change and opportunity in business since the internet. 4. AWS growth limited by supply chain, expected to gradually ease by the second half of 2025: Limitations mainly include slower chip supply speeds from third-party partners than before, many midstream changes that need some time to produce the level of healthy and high-quality servers we expect. Secondly, our own hardware and the launch of Trainium 2 chips were just introduced at re:Invent, but the majority of the shipments will truly arrive in the coming quarters and months. Additionally, there are electricity limitations as the world is still restricted in electricity supply. It is expected that these limitations will gradually ease by the second half of 2025, and while our current growth is quite good, I think we may grow even faster. 5. Depreciation will impact profit: Reducing the lifespan of some servers and network equipment will impact profitability in 2025. Increasing the lifespan of heavy-duty equipment in fulfillment centers will add about $900 million to profit in 2025. Starting from January 2025, the lifespan of some servers and network equipment will be reduced from 6 years to 5 years. We expect this to reduce operating profit for the fiscal year 2025 by approximately $700 million. 6. Full application of AI: Using artificial intelligence in various business areas, including customer service, product shelving, inventory management, and Siasun Robot & Automation technology. New experiences: Offering a new shopping experience through features like Rufus (AI shopping assistant) and Amazon Lens. We are using it in our retail business, and every business we are in is the same. We have built or are building around 1,000 different generative AI applications. The following is the full text of the conference call (translated by AI): Host: Thank you for waiting. Hello, everyone, and welcome to Amazon.com, Inc.'s 2024 fourth-quarter financial performance conference call. Currently, all participants are in listen-only mode. After the presentation, we will have a question-and-answer session. Today's call is being recorded. Next, I will hand over the meeting to Vice President of Investor Relations, Dave Fildes, to deliver the opening remarks. Thank you, Mr. Fildes, please proceed. Dave Fildes, Vice President of Investor Relations: Hello, welcome to our 2024 fourth-quarter financial performance conference call. Today, our CEO, Andy Jassy, and CFO, Brian Olsavsky, will answer your questions. As you listen to today's call, we recommend that you have our press release ready, which contains our financial performance, as well as key metrics and comments on this quarter. Please note that unless otherwise stated, all comparisons in this call will be against the results of the same period in 2023. Our comments today and responses to your questions reflect the management's views as of February 6, 2025, and include forward-looking statements. Actual results may differ significantly. Additional information on factors that may affect our financial performance can be found in today's press release and in the documents we submitted to the U.S. Securities and Exchange Commission (SEC), including our recent annual report (10-K) and subsequent filings. In this call, we may discuss certain non-GAAP financial metrics. Additional disclosure information about these non-GAAP metrics, including reconciliations to comparable GAAP metrics, can be found in our press release, the slides from today's webcast, and the documents we submitted to the SEC (all of which are available on our Investor Relations website). Our guidance incorporates the order trends we have seen so far and what we believe today.The assumption is that our performance is essentially unpredictable and may be significantly impacted by many factors, including exchange rate fluctuations, global economic and political changes, customer demand and spending (including concerns about recession, inflation, interest rates, regional labor market constraints, world events, the internet, online commerce, cloud services, as well as the growth rate of emerging and new technologies) and various other factors detailed in the documents we submit to the SEC. Our guidance assumes, among other things, that we will not complete any additional business acquisitions, restructurings, or legal settlements. The demand for our products and services cannot be accurately predicted, so our actual performance may differ significantly from our guidance. Now, I will turn the meeting over to Andy.Andy Jassy, President and Chief Executive Officer: Thank you Dave. Today, we are reporting revenue of $187.8 billion, a year-over-year increase of 10%. Due to the continued strength of the US dollar this quarter, we faced larger currency headwinds than expected, reaching $700 million. Without this headwind, revenue would have grown by 11% year-over-year, surpassing the high end of our guidance. Operating profit was $21.2 billion, a 61% increase year-over-year; adjusted free cash flow after equipment finance leases was $36.2 billion, a $700 million increase year-over-year. We are pleased with our results in innovation, customer experience enhancement, and achievements in 2024, with more plans for 2025. Let me first talk about our retail business. In North America, our revenue grew by 10% year-over-year, and internationally, excluding the impact of exchange rates, revenue grew by 9% year-over-year. We continue to focus on expanding product selection, lowering prices, and enhancing convenience, which has driven strong unit sales growth, surpassing even our revenue growth. We continue to expand our product selection, offering customers choices at various price points. In 2024, we welcomed several well-known brands to our stores, including Clinique, Este Lauder Companies Inc. Class A, Ora Rings, and Giorgio Armani Beauty. We also continue to increase products from our selling partners, who accounted for 61% of the units sold in 2024, the highest annual percentage ever for third-party sellers. In the fourth quarter, we launched Amazon.com, Inc. Halls for US customers, providing a fun shopping experience that consolidates ultra-low priced items in a convenient destination. It had a very strong start. Customers always want Amazon.com, Inc. to be their trusted low-price shopping destination. In the fourth quarter, consumers saved over $15 billion through our low-priced everyday items and record-breaking events such as Prime Day in October, Black Friday during Thanksgiving, and Cyber Monday. Additionally, according to the annual federal price study, as the holiday shopping season began, Amazon.com, Inc. has had the lowest online prices among major retailers in the US for the eighth consecutive year, with an average price 14% lower than other leading retailers. Our delivery speed continued to increase in 2024, marking a record-breaking year for Prime members. We added more than 60% to the number of same-day delivery stations in 2024, now serving over 140 metropolitan areas. Overall, we achieved over 9 billion same-day or next-day deliveries worldwide. Our relentless pursuit of better selection, prices, and delivery speed has driven accelerated growth in Prime membership. Prime members pay only $14.99 per month to enjoy unlimited free shipping on over three billion items, usually same-day or next-day deliveries, exclusive shopping events like Prime Day, access to a vast selection of premium content on Prime Video and live sports events, ad-free listening to over 100 million songs and podcasts on Amazon Music for just $5 per month, unlimited grocery deliveries for orders over $35 from Whole Foods Market and Amazon Fresh at $9.99 per month, free Grubhub+ membership for unlimited deliveries, and our latest benefita $0.10 per gallon fuel discount at BP, Ameco, and AMPM gas stations. When you look at these as a whole and compare them with many other membership services that are priced equally or even higher, these benefits provide just one advantage, such as video, Prime is a streaming package, and we will bring more discounts to Prime members in 2025. We also focus on reducing the per unit service cost of our logistics network, which has been a key driver of our operating profit growth. We have talked about regionalizing the US network. We recently launched a redesigned US inbound network. Although still in the early stages, our inbound work has improved our inventory placements, bringing more products closer to the end customer. Before Black Friday in November, we increased the percentage of available order units in ideal buildings by over 40%. We have also spent significant time optimizing the number of packages we send to customers, reducing packaging for a more convenient and cost-effective solution for customers. As we build and optimize the last-mile network, our unit transportation costs continue to decrease. Overall, we have reduced global per unit service costs for the second consecutive year, while increasing speed, improving safety, and expanding our selection. Looking into 2025 and beyond, we see opportunities to reduce costs through further optimizing inventory placement, expanding same-day delivery networks, and accelerating Siasun Robot & Automation and automation throughout the network. In terms of advertising, we are pleased with the strong growth on top of a solid base, achieving revenue of $17.3 billion this quarter, an 18% year-over-year increase. This is an annual revenue run rate of $69 billion, more than double the $29 billion from four years ago. Sponsored products represent the largest portion of advertising revenue, performing well, and we see more room for growth. We also have some newer streaming services that are beginning to be important new sources of revenue. In terms of streaming video, we completed the first year of Prime Video advertising and are satisfied with the early progress, entering the year with momentum. We are making omnichannel advertising easier to work with us. Omnichannel advertising starts with broad coverage advertisements at the top of the funnel, driving brand awareness, to the middle of the funnel where sponsored brands enable companies to specify particular keywords and audience types, getting people to visit their detailed Amazon.com, Inc. pages or brand stores. At the bottom of the funnel, sponsored products help advertisers display relevant product ads to customers at the point of purchase. Overall, we have seen strong growth in our advertising revenue, and we are excited about the progress we have made in various areas of our business. Thank you for your continued support and partnership with Amazon.com, Inc.We make it easy for brands to register and deploy these services in our growing advertisements. We also have unique audience capabilities, utilizing signals from billions of customers in Amazon.com, Inc.'s Marketing Cloud's secure clean room, to provide advertisers with analytical data, generate key marketing metrics, understand the performance of their marketing across different channels, as well as our new multi-touch attribution model, to help advertisers understand the effectiveness of their marketing. If advertisers use streaming television, display ads, sponsored products, and other types of advertising in their campaigns, multi-touch attribution will show the relative contribution of each to sales.Next is AWS. In the fourth quarter, AWS grew by 19% year-over-year, and the current annual revenue run rate is $115 billion. By most standards, AWS is already a significant business. While we expect growth to fluctuate in the coming years due to enterprise adoption cycles, capacity considerations, and technological advances, we are optimistic about the future of AWS customers and business. I have spent a lot of time thinking about the development in the upcoming years. While some may find it difficult to imagine a world where almost every application incorporates generative artificial intelligence, where reasoning becomes a core building block like computing, storage, and databases, and most companies have their own agents to perform various tasks and interact with each other. We have been thinking about such a world, and we continue to believe that this world will largely be built on the cloud, with much of it being built on AWS. To best help customers achieve this future, you need strong capabilities in all three layers of the stack. At the bottom, for those building models, unique and attractive chips are needed. Chips are a key element in driving training and inference computations. Most AI computations are powered by NVIDIA Corporation chips, and we clearly have a deep partnership with NVIDIA Corporation and will continue to do so for a long time to come. However, there are not yet many large-scale generative AI applications. When you reach a scale like ours, with applications like Alexa and Rufus, costs increase rapidly. Customers want better value for money, which is why we developed our own custom AI chips. Tranium 2 was just launched at the AWS re:Invent conference in December, and EC2 instances with these chips typically offer a 30% to 40% better value for money compared to other existing GPU-based instances. This advantage in large-scale applications is significant. Some technically strong companies, such as Adobe, Databricks, Poolside, and Qualcomm, have achieved impressive results in early tests of Tranium 2. This is also why Anthropic chose to build their future cutting-edge models on Tranium 2. We are collaborating with Anthropic to develop Project Raynier, a cluster of Tranium 2 super servers containing tens of thousands of training chips. The scale of this cluster will be five times that of the cluster Anthropic uses to train its current leading cloud models. We are already developing Tranium 3 and plan to preview it by the end of 2025, followed by defining Tranium 4. Developing chips with outstanding performance and leading value for money has become one of AWS's core advantages, starting from our core business with Nitro and Graviton chips, and now expanding to Tranium in the AI field, which sets AWS apart from other competing cloud providers. For model builders, another key component is the ability to simplify model building services. I won't go into too much detail about Amazon.com, Inc.'s SageMaker AI in these comments, as it has become the preferred service for AI model builders to manage AI data, build models, experiment, and deploy models. I just want to mention that SageMaker's Hyperpod feature automatically distributes training workloads to multiple AI accelerators, regularly saving checkpoints and automatically recovering from the last saved checkpoint in case of a fault instance, preventing interruptions and saving up to 40% of training time. This feature continues to be our differentiating advantage, and at re:Invent, we introduced several new exciting capabilities, including managing costs at the cluster level and prioritizing workload allocation when budgets are reached. It is increasingly being adopted by model builders. In the middle layer, for those who want to use cutting-edge models to build generative AI applications, Amazon.com, Inc. Bedrock is our fully managed service offering the widest selection of high-performance base models and the most attractive feature set to make building high-quality generative AI applications simple. We continue to iterate Bedrock rapidly and announced the addition of Luma AI, Poolside, and over 100 other popular emerging models to Bedrock at re:Invent. We also quickly added the DeepSeek-R1 model to Bedrock and SageMaker. Additionally, we introduced some exciting new Bedrock features at re:Invent, including prompt caching, intelligent prompt routing, and model distillation, all of which help customers achieve lower costs and latency in inference. Like SageMaker AI, Bedrock is growing rapidly and is popular among customers. In a related move, we have just launched Amazon.com, Inc.'s own cutting-edge model family, Nova, which matches the world's leading models in intelligence but offers lower latency and pricing compared to other models in Bedrock, about 75% lower, and integrates key Bedrock features such as fine-tuning, model distillation, Knowledge Library (RAG), and agent capabilities. Thousands of AWS customers have already begun to leverage the capabilities and value for money of Amazon.com, Inc. Nova models, including Palantir, Deloitte, SAP, Dentsu, Fortinet, Trellix, and Robinhood. We are just getting started. At the top layer of the stack, Amazon.com, Inc. Q is the most powerful generative AI-driven assistant supporting software development and leveraging your own data.In the last conference call, I shared a very practical use case where Q Transform helped the Amazon.com, Inc. team save 260 million dollars and 4500 developers in migrating over 30,000 applications to a new Java JDK version. This is real value, and the company is asking for more. We have met these demands with the recent delivery of Q Transform, which is capable of migrating Windows .NET applications to Linux, moving from VMware to EC2, and accelerating mainframe migrations. Early customer tests indicate that Q Transform can turn a mainframe migration that would have taken years into a project that only takes a few quarters, reducing the time for mainframe migration by over 50%. This is significant, and these transformations are a good example of practical AI.Despite AI continuing to be a prominent focus in our business, we have not lost sight of our core modern company technology infrastructure from local to cloud. We have signed new AWS agreements with several companies including Intuit, PayPal, Norwegian Cruise Line Holdings Ltd., Northrop Grumman Corporation, American Bankers Life Assurance Company, Reddit, Japan Airlines, Beckman Coulter, Hertz Global Holdings Inc., Redfin, Chime Financial, and Asana. The consistent customer feedback we received at the recent AWS re:Invent event is that we are still rapidly innovating in non-AI critical infrastructure areas such as storage, computing, databases, and analytics. Our functional leadership continues to expand, with several key releases that have excited customers, including Amazon.com, Inc. Aurora DSQL, our new serverless distributed SQL database that provides applications with high availability, strong consistency, PostgreSQL compatibility, and is 4 times faster than other popular distributed SQL databases; Amazon.com, Inc. S3 Tables, making S3 the first cloud object storage with fully managed support for Apache Iceberg analytics functionality; Amazon.com, Inc. S3 Metadata, which auto-generates queryable metadata, simplifying data discovery, business analytics, and real-time insights, helping customers unlock the value of data in S3; and the next generation Amazon.com, Inc. SageMaker, integrating all data, analytics services, and AI services into a single interface for easier large-scale analytics and AI operations. As we near the end of 2024, I want to thank our team members and partners for their significant contributions over the past year. From every perspective, it has been a very successful year. We are far from done and look forward to bringing more achievements to our customers in 2025. Next, I will hand over the meeting to Brian for the financial update. Brian T. Olsavsky, Chief Financial Officer: Thank you, Andy. First, let's take a look at our overall financial performance. Global revenue was $187.8 billion, an 11% year-over-year increase, excluding the impact of exchange rates. This equates to approximately $900 million headwinds due to exchange rates in the quarter, which is about $700 million higher than we expected in our fourth quarter guidance. Excluding this additional exchange rate headwind, our revenue would have exceeded the high end of our revenue guidance range. Global operating profit was $21.2 billion, our largest operating profit quarter ever, and $1.2 billion higher than the high end of our guidance range. Across all business segments, we continue to innovate for customers while operating more efficiently. In North America, revenue for the fourth quarter was $115.6 billion, a 10% year-over-year growth. International revenue was $43.4 billion, a 9% year-over-year growth, excluding the impact of exchange rates. Global paid units grew 11% year-over-year, as our focus on low prices, wide selection, and fast delivery continues to resonate with customers. In terms of profitability, North America operating profit was $9.3 billion, a $2.8 billion year-over-year increase. Operating profit margin was 8%, up 190 basis points year-over-year. In the International segment, operating profit was $1.3 billion, a $1.7 billion year-over-year increase. Operating profit margin was 3%, up 400 basis points year-over-year. This marks the eighth consecutive quarter of year-over-year growth in operating profit margin for both North America and International segments. In 2024, we also reduced global unit service costs for the second consecutive year. In the fourth quarter, we achieved strong productivity in our transportation network, driven by improvements in inventory placement, an increase in units per package, and reduced transportation distances. We also saw productivity improvements in our fulfillment centers. Overall, our team performed exceptionally well throughout the quarter, especially during the peak season. I want to thank them for everything they have done for our customers. Looking ahead, we have multiple opportunities to continue lowering costs through better inventory placement, allowing us to deliver goods to customers more quickly. In the U.S., we are adjusting our inbound network and expanding our same-day delivery network. Globally, we are increasing automation and Siasun Robot & Automation technology throughout the network. While these efforts take time to implement and progress may not be linear, we have a solid plan to continue driving improvements in our cost structure. Advertising remains a significant contributor to profitability in North America and International segments. This quarter, we achieved strong advertising revenue growth on a continually expanding base. We will also continue to invest in areas that are long-term important to Amazon.com, Inc. customers such as Alexa, healthcare, groceries, and Kuiper, including the production satellites we plan to launch in the coming months. It is worth noting that we currently capitalize most of the costs associated with our satellite network development. Once the service becomes commercially viable, including selling to customers, we will capitalize certain costs. Next is our AWS business segment, with revenue of $28.8 billion, a 19% year-over-year increase. AWS currently has an annual revenue run rate of $115 billion. In the fourth quarter, we continue to see growth in both generative AI and non-generative AI products as companies focus on new projects, shift more workloads to the cloud, restart or accelerate existing migrations from on-premises to the cloud, and leverage the powerful capabilities of generative AI. Customers recognize that to fully leverage the advantages of generative AI, they must move to the cloud. AWS reported operating profit of $10.6 billion, a year-over-yearAn increase of $3.5 billion. This is due to strong growth, our innovation in software and infrastructure driving efficiency improvements, and our continued focus on cost control throughout the business. As we have said in the past, we expect the operating profit margin of AWS to fluctuate over time, in part due to the level of investments we are making.In addition, starting from 2024, we will extend the expected lifespan of servers, which contributed about 200 basis points to AWS profit margin growth year-over-year in the fourth quarter. Now let's talk about our capital investments. Just a reminder, we define it as the sum of cash capital expenditures and equipment financing leases. Capital investments for the fourth quarter amounted to $26.3 billion, and we believe this run rate will reasonably represent our capital investment rate for 2025. Similar to 2024, the majority of the spending will be used to support the growing demand for technology infrastructure. This is primarily related to AWS, including supporting our AI service needs and the technology infrastructure for both North American and international business. Additionally, we continue to invest in our logistics and transportation network capacity to support future growth. We are investing in building more same-day delivery facilities, optimizing our inbound logistics network, and increasing investments in Siasun Robot & Automation technology and automation to improve delivery speed and reduce unit service costs. These capital investments will provide long-term support for future growth. Next, let's talk about the revenue guidance for the first quarter. We expect net sales to be between $151 billion and $155.5 billion. I want to emphasize two factors that impact our first-quarter revenue guidance. First, the estimated impact of foreign exchange changes on year-over-year growth based on current exchange rates is expected to create headwinds of approximately $2.1 billion in the first quarter, equivalent to a 150 basis point year-over-year decrease. Please note that global currency exchange rates may fluctuate during the quarter, as we saw the U.S. dollar strengthen against most other currencies in the fourth quarter. Secondly, a reminder that we are comparing against the leap year effect from last year. The extra day in the first quarter of 2024 brought approximately $1.5 billion in additional net sales to our business, equivalent to a decrease of about 120 basis points in year-over-year growth rate, impacting all business segments. Operating profit for the first quarter is expected to be between $14 billion and $18 billion. This guidance includes the estimated impact of adjustments to the useful life of certain fixed assets. I will provide more details later, but overall, we expect these adjustments to reduce the operating profit of assets on the balance sheet as of December 31, 2024, by approximately $400 million in the 2025 fiscal year. First, in the fourth quarter, we completed a study on the useful life of servers and network equipment and observed the accelerated pace of technological developments, particularly in the areas of artificial intelligence and machine learning. Therefore, we decided to shorten the useful life of some servers and network equipment from 6 years to 5 years starting in January 2025. We expect this to reduce the operating profit for the 2025 fiscal year by approximately $700 million. Additionally, we also accelerated the scrapping of some servers and network equipment, resulting in accelerated depreciation and related expenses of approximately $920 million recorded in the fourth quarter of 2024, which is expected to further reduce the operating profit for the 2025 fiscal year by about $600 million. These two adjustments in the useful life of servers and network equipment primarily affect our AWS business segment. Lastly, we have completed a study on the useful life of certain heavy equipment used in our fulfillment centers and have decided to extend the useful life of these equipment from 10 years to 13 years starting in January 2025. We expect this to increase the operating profit for the 2025 fiscal year by approximately $900 million. As we move towards 2025, we are excited about the achievements of our team. We remain focused on creating better experiences for our customers, and we believe that putting the customer at the center is the only reliable way to create lasting value for our shareholders. Now, let's move on to the question-and-answer session. Q&A Session Host: Now we will begin the question-and-answer session. Please follow the instructions provided. Our first question comes from Mr. Mark Mahaney from Evercore ISI, please go ahead with your question. Mark Mahaney, Analyst: Thank you. I have two questions. First, Brian, should we consider capital expenditures of approximately $100 billion for 2025? Second, Andy, do you think that AWS growth is currently being impacted by supply chain constraints? Do you believe these constraints are widespread across the industry, or do they have a substantive impact on AWS specifically? Thank you very much. Andy Jassy, President and CEO: Let me address these two questions. This is Andy speaking. Regarding capital expenditures, as Brian mentioned earlier, our capital expenditures for the fourth quarter were $26.3 billion. I believe this can reasonably represent the annual capital expenditure rate for 2025. Most of the capital expenditures are used in the AI field of AWS. The way AWS operates and the cash flow cycle is, the faster we grow, the higher the capital expenditures because we need to procure data centers, hardware, chips, and network equipment in advance, before we can monetize them. We don't make purchases without significant demand signals. So when AWS expands its capital expenditure, particularly in rare commercial opportunities like AI, I think this is a fairly good signal for the medium to long-term development of AWS business. In fact, I think it's wise to invest this capital in pursuit of this opportunity because we believe almost all applications we know today will be reinvented with AI integrated, and reasoning will become a core building block just like computing, storage, and databases. If you believe this, along with the entirely new experiences that we could only dream of before now becoming possible through AI, AI undoubtedly presents the greatest opportunity since cloud computing and perhaps the largest technological transformation and opportunity in business since the internet. So I think everyone, whether it's our business, our customers, or our shareholders, will be satisfied with our pursuit of capital opportunities and business opportunities in the field of AI in the mid to long term. We are also investing capital in our retail business, primarily to continue to improve delivery speed and reduce unit service costs. So you will see us increase the number of same-day delivery facilities from the current level.You will also see us increasing the number of distribution points in rural areas so that we can deliver goods to people living in rural areas faster. In addition, we will also make significant investments in Siasun Robot & Automation technology and automation to reduce our unit service costs and continue to improve our productivity.Therefore, this is the part of capital expenditure. I think your second question, Mark, is about whether AWS's growth is being impacted by supply chain constraints? When you have a business like ours with billions of dollars in annual revenue in the AI field and it's growing in the triple digits year-over-year, it's hard to complain. However, indeed, if there weren't some capacity limits, we might be growing even faster. These limits are mainly reflected in a few areas: firstly, the chip supply speed from our third-party partners has slowed down a bit, changes in the intermediate process take some time to bring the hardware to the level of health and high-quality servers we expect. Secondly, this is also reflected in the launch of our own hardware and chip Trainium 2, which we just announced at re:Invent but most of the shipments will truly arrive in the next few quarters, in the coming months. Additionally, there are power constraints, I think the world is still limited in terms of power supply, and I think if we weren't limited, we could offer more services to our customers. There are also some components in the supply chain, such as motherboards, that are slightly tight in supply for various types of servers. So, I think the team is doing well in this regard, by piecing together and providing capacity, allowing our customers to continue to grow. As I mentioned earlier, our current growth rate is quite good, but I think we could grow even faster without these constraints. I expect these constraints to gradually ease by the second half of 2025. As I said, although our current growth is quite good, I think we could grow even faster. Host: The next question comes from Mr. Eric Sheridan, Goldman Sachs Group, Inc., please go ahead with your question. Eric Sheridan, Analyst: Thank you for taking my question. I will continue with Mark's previous question. Andy, considering some recent news from China in the past few weeks, and in the long term, reducing unit costs through AI, I understand your comments on capital expenditure for 2025. But from your position in the industry, with trends like open source and custom chips, how do you view reducing unit costs through AI, or speeding up or amplifying market deployment time, or possibly increasing return on capital? Thank you very much. Brian T. Olsavsky, Chief Financial Officer: Yes, I'd like to say a few things because there are some questions here. Firstly, like many others, we are impressed by what DeepSeek is doing. I think part of it is some of their training techniques, primarily the reversal of order in reinforcement training, reinforcement learning happening earlier and without human intervention. We find it interesting before supervised fine-tuning. We also find their work on reasoning optimization interesting. For those of us building cutting-edge models, we are all doing similar things and learning from each other. I think you'll see and we will continue to see mutual leaps among us, there's a lot of innovation in the future. If you run a business like AWS and believe, as we do, that nearly all large generative AI applications will use a variety of model types, different customers will use different models to handle different workloads, then you want to offer as many leading-edge models for customers to choose from as possible. That's what we are doing with the Bedrock service through Amazon.com, Inc. This is also why we are moving quickly to ensure that DeepSeek is available faster in Bedrock and SageMaker than others, and we already have customers starting to try it out. I think an interesting phenomenon in the past few weeks is that sometimes people assume that if you can reduce the cost of a technology component, in this case mainly inference, then that will somehow lead to a decrease in total technology spend. We have never seen this happen. We did the same thing in cloud computing. When we launched AWS in 2006, we offered S3 object storage for $0.15 per GB and compute services for $0.10 per hour, of course, these prices have since dropped significantly after many years. People thought that businesses would spend less money on infrastructure technology. But the reality is that companies will spend less money on the infrastructure per unit, which is very useful for their business, but they will then get excited about projects they previously found too expensive and end up spending more money on technology. I think the same thing will happen in the field of AI. The cost of inference will significantly decrease. DeepSeek, which everyone has heard about in the past few weeks, is just a part of it, but everyone is working on it. I believe the cost of inference will decrease significantly. This will make it easier for companies to integrate inference and generative AI into all their applications. I think if you run a business like ours, we want to make it as easy as possible for customers to build customer experiences on top of our various infrastructure services, so the reduction in inference costs will be very beneficial to customers and our business. Host: The next question comes from Mr. Doug Anmuth, from JPMorgan, please go ahead with your question. Doug Anmuth, Analyst: Thank you for taking my question. I'll start with AWS. Brian, can you further discuss the issue of AWS profitability? In the past two years, the profit margin has fluctuated from over 20% to over 30%. With significant investments in generative AI, how should we think about a more normalized profit margin level? Also, in the store business, could you talk about the impact of the reduction in volume of goods transported through UPS in the future? Can you manage the additional shipping volume required? Thank you. Brian T. Olsavsky, Chief Financial Officer: Of course, Doug. Thank you for your question. Firstly, regarding AWS, yes, we have indeed seen significant fluctuations in the operating profit margin of AWS. We have previously stated that as weWith the passage of time, these profit margins will be unpredictable. Currently, AI is still in its early stages. It does indeed initially come with lower profit margins and heavy investment burdens, as we have discussed before. In the short term, this may put some pressure on profit margins.But in the long run, we believe that the profit margin will be comparable to non-AI businesses. Therefore, we are very satisfied with the strong growth of the AWS team, their focus on improving efficiency and saving power in all data centers, reusing power in new generative AI applications, and overall cost reduction. So, we are very pleased with the performance of the AWS team and look forward to a strong performance in 2025. Andy Jassy, President and CEO: I'll answer the UPS question. UPS has been our partner for many years, and we expect to continue our partnership with them in the future. As you know, in the past few years, especially during the pandemic, we have been shipping more and more packages through our own logistics network, which has accelerated this trend. We did this partly because we needed to quickly scale during the pandemic when everything was shut down, and we needed to serve a larger share of the retail market at that time and with a low-cost structure because our customers expect low prices, which is the nature of our business. UPS may think that providing services to Amazon.com, Inc. has lower profit margins. Therefore, they chose to forego some of the business volume that they could have gained from our partnership. We can handle this business through our own logistics capabilities, and we will see how this partnership evolves. Host: Next question comes from Mr. Brian Nowak from Morgan Stanley, please go ahead with your question. Brian Nowak, Analyst: Thank you for taking my question. Andy, could you elaborate on the accelerated application of Siasun Robot&Automation technology that you mentioned? Can you share some new data or learnings from Shreveport? How should we think about the scalability of Siasun Robot&Automation technology and when we might see substantial impact on profitability? Furthermore, from a broader perspective, with the changes driven by generative AI and GPU, could you provide more insight into how you see the Amazon.com, Inc. retail shopping experience evolving in 2025 through better utilization of generative AI or GPU-driven machine learning? Andy Jassy, President and CEO: Sure. Regarding Siasun Robot&Automation technology, I want to say that we have been integrating Siasun Robot&Automation technology into our fulfillment network at a large scale for many years, and we have seen cost savings, productivity improvements, and safety enhancements. Therefore, we have already derived significant value from Siasun Robot&Automation technology. Recently, I think what you were referring to with Shreveport is that we have seen the next wave of Siasun Robot&Automation technology projects begin production, and we have fully integrated them into the Shreveport facility experience for the first time. We are very encouraged by the speed increase, productivity improvement, and reduction in unit service costs that we have seen there. It's still early days, these technologies are only applied in Shreveport at the moment, but we plan to start expanding and deploying them to other facilities in the network, some of which will be our new facilities, and others will be existing facilities transformed to use the same Siasun Robot&Automation technology. I want to tell you that this group of about six new Siasun Robot&Automation technology projects is far from reaching the end of the possibilities of using Siasun Robot&Automation technology to improve productivity, unit service costs, and safety in our fulfillment network. We are moving on to the next wave of innovation, but I think this will be a multi-year effort as we continue to adjust different parts of the fulfillment network to utilize Siasun Robot&Automation technology. In fact, we don't think there's much that we can't improve through Siasun Robot&Automation technology. Regarding your other question, I think this is actually about how we better utilize AI in other areas of our business, particularly in AWS. Maybe you're asking about our retail business. I think you can think about it like this: currently, we and other companies using AWS mainly derive value from AI in two macro ways. The first macro category revolves around productivity and cost savings. In many ways, this is the low-hanging fruit of AI, and you can see it everywhere in our retail business. For example, if you look at customer service and our chat Siasun Robot&Automation that we built, we completely re-architected it using generative AI. It already has quite high customer satisfaction, and the new generative AI chat Siasun Robot&Automation has increased customer satisfaction by 500 basis points. If you look at our millions of third-party seller partners, one of their biggest pain points is that due to our heavy organization of our marketplace, making it easy to find products, there are many different fields to fill in when creating a new product detail page, but we built a generative AI application for them, where they can just fill in a few lines of text, or take a picture, or point to a URL, and the generative AI application will automatically fill in most of the other information, speeding up the time to list products on the website and making it easier for sellers. If you look at how we manage inventory, trying to understand when and where to place inventory in facilities, our generative AI application has improved our forecasting accuracy by 10%, regionally predicting accurately.Accuracy has increased by 20%. In our Siasun Robot & Automation technology, many of the "brains" we just talked about are integrated with generative AI, which can tell the robotic arm what is in the box, what to grab, how to move it, and where to place it in another box. Therefore, generative AI is actually the "brains" of many of our Siasun Robot & Automation.RufusRufus If you look at the personalized function, especially now, you can enter Rufus and inquire about the status of orders, or what I just ordered, or can you pull out the product I ordered two months ago? The personalized function is getting better and better. Therefore, we expect that by 2025, the situation where you are unsure of what to buy and need Rufus' help will continue to increase, and be increasingly helpful to customers.Host: Thank you for participating in our conference call today and for your questions. A recording of the meeting will be available on the investor relations website for at least three months. We appreciate your interest in Amazon.com, Inc. and look forward to speaking with you again next quarter. Ladies and gentlemen, that concludes our conference call for today. You may now disconnect. Thank you for your participation. This article is reprinted from Wall Street View, GMTEight editor: Chen Xiaoyi.