# How Alphabet Slipped Ahead In The AI Race ## Метаданные - **Канал:** CNBC - **YouTube:** https://www.youtube.com/watch?v=_BngA7hLTv4 - **Дата:** 19.05.2026 - **Длительность:** 14:03 - **Просмотры:** 89,274 ## Описание 18 months ago, Google looked like it had missed the AI revolution. Now, Alphabet's stock is up 140% over the past year and Wall Street is betting it's one of the few companies positioned to profit from every layer of the generative AI boom. From Gemini to Google Cloud to its custom TPU chips, the company controls more of the AI stack than almost any of its rivals. This week's Google I/O is the next big test — investors want to see whether that confidence is backed by a real product roadmap. CNBC's MacKenzie Sigalos explains how Alphabet became the AI race’s unlikely frontrunner. Chapters: 0:00 Introduction 2:40 Chapter 1 - Google’s AI comeback 4:55 Chapter 2 - Monetizing AI 9:29 Chapter 3 - Google’s AI future Reporter: MacKenzie Sigalos Produced and Edited by: Andrew Evers Additional Editing: Matt Soto Senior Director of Video: Jeniece Pettitt Animations: Jason Reginato, Christina Locopo, Emily Park Additional Footage: Getty Images, Google » Subscribe to CNBC: https://cnb.cx/SubscribeCNBC » Subscribe to CNBC TV: https://cnb.cx/SubscribeCNBCtelevision About CNBC: From 'Wall Street' to 'Main Street' to award winning original documentaries and Reality TV series, CNBC has you covered. Experience special sneak peeks of your favorite shows, exclusive video and more. Want to get ahead at work? Then you need to learn how to make effective small talk. In CNBC’s new online course, How To Talk To People At Work, expert instructors teach you how to use everyday conversation to gain visibility, build meaningful relationships and advance your career. Sign up now: https://cnb.cx/4sGlSkh Connect with CNBC News Online Get the latest news: https://www.cnbc.com/ Follow CNBC on LinkedIn: https://cnb.cx/LinkedInCNBC Follow CNBC News on Instagram: https://cnb.cx/InstagramCNBC Follow CNBC News on Facebook: https://cnb.cx/LikeCNBC Follow CNBC on Threads: https://cnb.cx/threads Follow CNBC News on X: https://cnb.cx/FollowCNBC Follow CNBC on WhatsApp: https://cnb.cx/WhatsAppCNBC #CNBC How Google Quietly Became An AI Powerhouse ## Содержание ### [0:00](https://www.youtube.com/watch?v=_BngA7hLTv4) Introduction Alphabet stock is up 140% in the past year and even briefly surpassed Nvidia in early May as the world's most valuable company by market cap. A remarkable feat for a company that was seen as deeply at risk in the early days of the artificial intelligence boom. Now, the rally reflects a major shift in how Wall Street sees Google. A year ago, investors were asking whether generative AI would destroy its search business. And now it's whether Alphabet may be one of the few companies that can actually monetize AI across nearly every layer, models, chips, cloud, search, Android, and enterprise software. — Google, in my opinion, they're probably second best position when it comes to doing this full stack of AI. I think SpaceX is a step ahead of where Google is, but it's really coming down to those two companies. — A company like Alphabet with Google has been able to in a very integrated way show that technology integration within the entire ecosystem that Google has created really matters. — Now, that wasn't always the investor view. Despite helping invent the technology underpinning the current AI boom, Alphabet spent much of the last few years playing defense, OpenAI and others moved faster with consumerf facing models and in some cases released products that appeared to outperform Google's own offerings. — The company was not really innovating. Uh there was a question about what was happening with search, but also when it came to the underlying models, they hadn't done anything that was that spectacular. — But the company has made an impressive comeback. Its Gemini app has surpassed 750 million monthly active users. Apple is tapping Gemini to power a rebooted Siri and chose Google as its preferred cloud provider. Google Cloud revenue surged 63% year-over-year in Q1. 75% of customers now using Google's full stack from chips to Gemini models. — We exited the year with double the new customer velocity compared to Q1. The number of deals in 2025, over a billion dollars, surpassed the previous three years combined. We continue to deepen our relationships with existing customers who are outpacing their initial commitments by over 30%. — Google's AI comeback is definitely one of the most important strategic reversals that we've seen, I would say, ever. — The pace is relentless in inside the company. Sundar, our CEO, has been really helping prepare the company for over a decade uh for this moment uh being an AI first company. ### [2:40](https://www.youtube.com/watch?v=_BngA7hLTv4&t=160s) Chapter 1 - Google’s AI comeback Google had every reason to be ahead. It was an early mover in AI. It acquired DeepMind in 2014, released Google Assistant in 2016, and it actually developed the transformer architecture that became the foundation for modern generative AI. — Google was the first AI megate techch company. The position that Google got themselves in at the beginning of 2025 should have never happened because they were uh years ahead of where their competition was. — And then came ChachiBT in late 2022. It turned Google's head start into a liability. Suddenly, the company looked slow, defensive, and exposed in the very category that it had helped create. — Google had really been disrupted within the category that they really innovated the best. So, you know, you had things like the bar demo mistake and the real concern that generative AI would cannibalize search. Gemini arrived a year later in December 2023, but instead of settling questions about Alphabet's place in AI, the rollout raised new ones. Early criticism centered on the product's uneven performance, restrictive guard rails, and a series of high-profile mistakes that reinforced the view that Google was still struggling to turn its AI research into a breakout consumer product. — They made some pretty big mistakes with Gemini initially, just in terms of how tight the rails were on it. And I think that they were quick to try to make some adjustments to make the model a little bit more appealing. — Investors priced that doubt into the stock. While other AI winners surged, Alphabet was still being valued like a company defending search, not leading the next platform shift. — The question is not whether Google will go away, it's whether the growth will go away. And that little margin uh was clearly at risk. But the narrative around Google began to turn when the company stopped trying to win the AI race on OpenAI's terms and started proving the value of the business that it already had in place. — I think Sundar was able to, I think, inject a sense of urgency within the company. — What we're seeing now is that Google is probably the best position company to monetize AI at scale because it controls almost every layer of the stack. ### [4:55](https://www.youtube.com/watch?v=_BngA7hLTv4&t=295s) Chapter 2 - Monetizing AI Now, the first phase of the AI trade was really about the chatbot moment. The next was about who could actually monetize AI across a full tech stack. — They recognized that their core asset was these almost three billion people that use Google search a day. We don't need to be as good as uh the chat GPT, but we don't want to be 10 times worse than chat GPT. And I think they successfully close that gap. The center of strategic gravity inside of Google without a doubt has shifted towards deep mind and frontier AI. After chat GPT, Google realized AI could no longer sit as a research adjunct inside the company. It had to become the organizing principle of the company. And the Gemini strategy effectively turned Deep Mind from a world-class research lab into a core operating engine of Google. — Gemini also got better giving Wall Street evidence that Google was not just defending its legacy business but building credible AI products. — We definitely saw uh I would say extraordinary period of growth in Q4 uh for Gemini app. Google has shown that their underlying model is getting progressively better. They've shown that the Gemini model can have a flex in terms of being able to do other things like nano banana and to be able to do more image generation and they have shown that the cloud business is something that is critical to the growth of AI. — Investors have rewarded that shift. Alphabet's market cap remains within striking distance of Nvidia's and its forward price toearnings multiple has expanded as Wall Street starts to value Google less like a search company under threat and a lot more like one of the few companies built to make AI pay its scale. That thesis got another boost after it was reported that Anthropic's cloud contract with Alphabet was valued at $200 billion. Alphabet briefly surpassed Nvidia in market value on that report, really reinforcing this idea that Google's not dependent on a single AI product. It has Gemini and DeepMind for models and research, Google Cloud for compute, TPUs as an alternative to Nvidia's chips, and search, YouTube, and Android is these massive distribution channels for AI features. — When you're building on your own custom silicon, for example, that's an advantage of speed. When you have access to power, you can get data centers up more quickly. That's a speed advantage, which is important. — Now, one of Google's clearest advantages is in custom silicon. In April, the company announced two new eighth generation TPUs, one built for training and one for inference. — The creation of a more efficient chip, the TPU. That innovation led to, you know, a complete integration across Google for infrastructure with their own chips that are customized for the specifics of the Google features and functionality. — For investors, that made Alphabet a rare kind of AI trade, a company with exposure to hardware, demand, cloud growth, and model development, all inside of the same business. Mizuho estimates roughly $61 billion of Google's cloud backlog through 2027 could come from TPU sales with most of that revenue likely recognized next year. That gives investors another way to play the AI infrastructure boom at a time when shares of AMD, Intel, and Micron have all more than doubled this year. — Their tensor units are something that will be incremental. I think they will increasingly start to sell these to third parties. The rough numbers are this is running at a pace of about a 500 billion annual opportunity. So if they could gain five or 10% share within that could move the needle by a few percentage growth. — Now even if enterprises choose claude open AI or another model altogether the compute still has to run somewhere and increasingly that makes cloud capacity custom chips and data center access the scarce assets in the AI economy. If enterprises prefer Claude, then Google still wins in infrastructure because all of that activity has to live somewhere. That goes back to the $200 billion deal with Anthropic. Google still wins because of their TPUs. Google will win on cloud revenue. And then Google will still benefit because of their ownership stake in Anthropic. So, it's a really powerful hedge that Alphabet has with that relationship with Anthropic. ### [9:29](https://www.youtube.com/watch?v=_BngA7hLTv4&t=569s) Chapter 3 - Google’s AI future Looking ahead, Alphabet's cloud backlog is becoming central to the bullcase and increasingly tied to Anthropic. If Anthropic's reported $200 billion commitment is measured against Alphabet's cloud backlog, it could represent more than 40% of future contracted revenue. Google has also committed up to $40 billion to Anthropic, creating a loop where capital can flow into the AI startup and then back to Google through cloud and TPU spending. That raises the same concentration question now hanging over Oracle and Microsoft. Oracle stock soared after it reported a nearly 360% jump in backlog, much of it tied to OpenAI. Then investors started questioning how much of that future revenue depended on one customer and the stock lost roughly half of its value over five months. Microsoft faces a similar debate around open AAI and for Google the question is whether Anthropic becomes a durable infrastructure flywheel or customer concentration risk. — Google is a significant investor in anthropic. So there is a partnership embedded within that investment as Google thinks through all of these strategic options of okay, if we have this amazing infrastructure that's powered by our TPU chip, we can monetize that and how much do they fence that monetization off and decide who they want to let use it. — Now the risk for Alphabet is not weak fundamentals. It's whether the company has enough new catalysts to reset investor expectations after a 140% run in the last 12 months. Alphabet's task now is giving investors a reason to keep expanding the multiple. That puts all eyes on Google IO. Investors will be looking for clarity on three fronts. How Google monetizes Gemini, how it builds an agent strategy around search and commerce, and how much revenue it can capture from the broader AI ecosystem. A couple big areas that I'm hopeful that we'll hear things from is just related to how they're going to better monetize new advertising products within AI mode. Second is just how they talk about Aentic uh commerce. I think the uh the third piece is around personalized AI. — Google's already started laying that groundwork. It unveiled a new AI native laptop called the Google book that launches this fall and it's pushing Gemini even deeper into Android. Since 2024, we launched Gemini on Android devices and since then integrated Gemini with all the different form factors, car, TV, watch, and more. Um, and now in 2026, we're launching a completely next version of the experience with Gemini Intelligence — Enterprise. It is the next big test. Enthropic has shown how quickly that market can scale. And OpenAI is pushing deeper into the same category with codecs. Google has the models, distribution, and infrastructure to compete. But what investors still need to see is whether it can turn those advantages into a durable enterprise AI business. — They're going to be a player in enterprise. They have to be. They have to be recognizing what's going on right now and recognize that their models, their underlying models have made a lot of progress. Google's uniquely strong across multimodal systems because they have this experience with some of the largest applications that handle them in YouTube, Android, Maps, Search, Deep Mind, and TPUs. — Now, the other question is cost. Alphabet is projected to spend up to $190 billion on capital expenditures, more than double its 2025 level, which was already double from the year before that. Now, the bull case is not that Google wins every layer of AI. It's that nearly every AI winner may need Google somewhere in its supply chain. — I think five years from now, we're going to look back and recognize that in 2026, we were still early. A year ago, I thought we're in the third inning. I now think we're in the second inning. I do uh believe that intelligence at scale is going to create the most valuable companies in the world. And I think Google along with SpaceX are the two companies that are pursuing this in some of the most aggressive ways. And ultimately, I think that this is one of those trends where you think it's over and it just keeps going --- *Источник: https://ekstraktznaniy.ru/video/51524*