← The AI Daily Brief: Artificial Intelligence News and Analysis
The AI Daily Brief: Artificial Intelligence News and Analysis
Are Markets Still Worried About an AI Bubble?
00:28:03
Loading summary
Transcript1 lines
- [00:01]
A
Today on the AI Daily Brief as Meta and Microsoft report earnings. Are markets still worried about an AI bubble? Before that in the headlines, SoftBank appears ready to double down on OpenAI with another $30 billion investment. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. Alright friends, quick announcements before we dive in. First of all, thank you to today's sponsors, zencoder, Rackspace, Robots and Pencils and Super Intelligent. To get an ad free version of the show go to patreon.com aidaily brief or you can subscribe on Apple Podcasts. If you are interested in sponsoring the show, send us a Note@ SponsorsideailyBrief AI now one more thing that you can find at AIDAILYBRIEF AI we have a new thing we're doing which is in conjunction with AIDB intel, which is a monthly Pulse survey to figure out how people are using AI and how it's changing. You can find a link on the main aidaily Brief AI website and it's a short survey that should take much less than five minutes. It basically asks things like which AI models did you use this month? Which AI model did you use the most? What was your most important use case? All in a very easy checkbox, multiple choice style format. It's part of our larger goal of helping people have up to date actual data about what's going on with AI and I would so appreciate it if you would take a couple minutes to fill it out. Anyone who contributes will get first access to the results at least a week before they are available for general consumption. And again, you can find everything about that at aidailybrief AI. Welcome back to the AI Daily Brief Headlines Edition. All the daily AI news you need in around five minutes. What's $30 billion between friends, right? Mazasun is in for another 30 billion as SoftBank appears to double down on OpenAI. The Wall Street Journal reports that SoftBank is in talks to invest another crisp 30 billion into OpenAI's next monster fundraising round. Now SoftBank is already one of OpenAI's largest shareholders with a roughly 11% stake after investing 30 billion last year. That funding seemed to stretch SoftBank's pockets, with the firm selling off Nvidia stock and taking loans against their AMD holdings to make it work. While rumors have been sparse, reporting from December suggested that OpenAI is targeting 100 billion in fundraising this year. The valuation is said to be 830 billion, which is another 66% jump from the $500 billion valuation struck in October. Frankly, it looks like they're well on their way with separate reports suggesting that Nvidia, Microsoft and Amazon are looking to participate to the tune of 60 billion between them. Now in my 2026 predictions I debated a bunch, but I ultimately came to the base case that I didn't think that anthropic or OpenAI would actually go public in 2026, with the reason being that a public markets are kind of a pain in the butt and b I thought that there was just going to be enough funding for them in private markets that they weren't going to be forced to go public. Now the capital needs are extreme here and so maybe they just have to take advantage of every option they have, including IPO. But with OpenAI well on their way to that hundred billion dollar round, they certainly are going to have their options Moving over to OpenAI competitor Anthropic, ServiceNow has signed another big AI deal, this time with Anthropic partnering with the lab along multiple angles. The multi year deal includes making Claude the default model across ServiceNow's platform. Claude will also drive their agent builder, which allows users to create custom workflows and vibe code apps. In addition, Claude access is being rolled out to all 29,000 ServiceNow employees, including Claude Code for the engineering team, ServiceNow CEO Bill McDermott said in a press release. ServiceNow with Anthropic is turning intelligence into action through AI native workflows for the world's largest enterprises. Together, we are proving that deeply integrated platforms with an open ecosystem are how the future is built. Now, details of the deal weren't released, so we don't know the length or monetary value. But the deal comes just a week after ServiceNow extended their other big deal with OpenAI, suggesting that they want to make sure that customers have access to AI from both companies. ServiceNow President Amit Zavri said, we don't view these partnerships as competitive or mutually exclusive. Enterprise customers want model choice. They want the right model for the right job keeping governance, security and auditability consistency on the ServiceNow AI platform. Each model brings different strengths and our role is to orchestrate them in ways that deliver the best outcomes for customers. And as much as that sounds like corporate speak, the reality is that one of the great frustrations for enterprise AI users is model lock in. So to the extent that they can work with partners like ServiceNow, where they have access to multiple models, for many people that's going to be desirable. Staying in the anthropic world for a minute, Claude Cowork has apparently raised the alarm at Microsoft. The information reports that the release of Anthropic's new productivity platform triggered emergency meetings. Product leaders at Microsoft told colleagues that Cowork seemed like a competitor for 365 copilot. They noted that Cowork seems more capable than Copilot when working on excel spreadsheets or PowerPoint slides. Staff were told to keep improving Copilot, and that seems to be exactly what's happened over the past two weeks. Sources said that multiple divisions at Microsoft are working on prototypes that mimic the functionality of Cowork, and they noted that Claude, which Microsoft now has access to, is powering many of these prototypes now. Obviously, Microsoft's relationship with OpenAI has been changing, but in many ways since the release of ChatGPT, Microsoft has been largely focused on keeping feature parity with OpenAI. Now it's pretty clear that keeping up with Anthropic has become their major concern. The reporting also noted a tension between Anthropic's ability to move with startup speed, I.e. they built Claude cowork in 10 days as compared to Microsoft's status as a lumbering tech behemoth. In addition to the speed of execution, there's also differences in what the companies can put out. Sources noted that Cowork is only available as a research preview and that Anthropic has warned users that it carries massive inherent security risks. Which is all well and good for Anthropic, but Microsoft wouldn't be able to release a version of Copilot that carries that same level of risk. Still, it sounds like all hands on deck at Microsoft. To keep up in the race, executives and product leaders are reportedly discussing possible new features in a teams channel called AI Accelerator. CEO Satya Nadella was even reportedly dabbling with claudebot, now called Moltbot, over recent days. Nadella encouraged staff to also test the automation tool and figure out how its features could be applied to Copilot, the information writes. The conversations are part of Nadella's effort to pressure deputies to speed up the company's use of AI in its products, and they resemble how other AI developers, including Google and OpenAI, have reacted with urgency to competitive products in the fast paced field. He has sounded alarms about the urgency of the situation, despite his company's position of strength as the world's fourth most valuable firm. Nadella reportedly now has a standing weekly meeting where staff can demo new features from competing labs. Sources said that many of these meetings have featured Claude code driven by Opus 4.5 and they mentioned that the release of Cowork has ramped up the urgency of the meetings in recent weeks. At this stage the reporting doesn't include the term code red, but you have to think we're not far from an all out response from Microsoft. Twitter was of course ablaze with jokes. Kyle Russell wrote Microsoft Office oh, you mean the Claude wrapper. Meanwhile, a couple of weeks ago, even before this, Gavin Baker said Claude Cowork is what Copilot should have been, evidently built in 10 days with Claude code while Microsoft has been working on Copilot for years. Although I have to say one really insightful comment is in response to Gavin came from repl.CEO Amjad Massad who said to make a bit of an excuse for Microsoft, the world is just waking up to the fact that coding agents are general agents. It's bitter lesson adjacent writing and executing code will likely outperform years of handcrafting vertical specific agents with expert knowledge. Actually, it might exactly map in bitter lesson. Program synthesis is a form of scalable search. Given the debates I've been having with you guys around my thesis that code AGI is functional AGI. I thought that was an interesting comment Moving over to Google's side of the house Google is getting in the AI browser game with a big agentic upgrade for Chrome. Google first added Gemini to Chrome in September, but that iteration kept the AI sandboxed in its own window. With this update, Gemini will be able to use open tabs as context and function as a web agent. That feature set is of course similar to agentic browsers from OpenAI, Perplexity and the browser company, which were all released last year. Google is even borrowing the ux, placing Gemini in a sidebar for split screen agentic browsing. One interesting quality of life feature that Google showed off was Gemini's ability to understand multiple tabs from the same website as a context group. For example, if you're shopping multiple items against each other, Gemini will understand that context when giving advice. Gemini in Chrome also leverages Google's recently released personal intelligence feature, allowing the agent to draw context from your Gmail search history and photos. The feature will initially be limited to Pro and Ultra subscribers, so this isn't the free web agent that takes the technology mainstream. Still, Google is shipping quickly and making a lot of progress with these functional consumer agents. Lastly, today, another one that harkens back to one of the discussions in my 2026 predictions. Tesla has made a $2 billion investment into Xai in defiance of a shareholder vote. Elon Musk has been toying with the idea of cross investment between his two companies over the past year. The matter went to a shareholder vote in November with over a billion votes in favor and 916 million against. A significant number of shareholders abstain from voting, which counts as a vote against under Tesla's bylaws, the meaning the vote technically failed. However, it seems that near enough was good enough for Musk. Tesla disclosed in a shareholders letter on Wednesday that they made a $2 billion investment in Xai's recent fundraising round and the investment is apparently all part of the plan. With Tesla writing as set forth in Master Plan Part 4, Tesla is building products and services that bring AI into the physical world. Meanwhile, XAI is developing leading digital AI products and services such as its LLM grok. In that context, and as part of Tesla's broader strategy under Master Plan Part 4, Tesla and Xai also entered into a framework agreement in connection with the investment. Now, Tesla has already collaborated with XAI by providing them batteries for power redundancy at their data centers, and XAI has also provided GROK in some Tesla vehicles. And there's plans to use XAI models to power the Optimus robots. In the shareholders letter, Musk wrote, if there are things XAI can help accelerate our progress, then why should we not do that? And that is the reason why we've gone ahead with such an investment, because this is part of the strategic initiative. Now the investment comes as Tesla hits a pretty rough patch. During their Wednesday night earnings call, Tesla disclosed a 61% drop in profits year over year. They announced that the Model 3 and Model X will be discontinued with the production line repurposed for Optimus robots, musk told investors on the call. This year for Tesla is the first major steps. As we increase vehicle autonomy and begin to produce Optimus robots at scale, we're making very, very big investments. So this is going to be a very big capex here. That is deliberate because we're making big investments for an epic future. Back in July of last year, 0xmo on Twitter wrote, Tesla will acquire Xai. There's no way around it. GROK will be the brain for Optimus. So you can't have the brain and the body made by different teams. It doesn't get you the best product and it's not Elon style. I give it around 12 months. Reflecting on the recent news, the same account wrote, tesla is investing in XAI and collaborating even closer. It's happening now. That is going to do it for our headlines, but for the rest of Today we are going to stay on public market themes. Moving now to ask, are markets still worried about an AI bubble? All right friends, quick break to talk about a question I hear constantly. How do you actually move from AI experimentation to production without getting buried in infrastructure decisions? That's where Rackspace AI Launchpad comes in. It's a fully managed service designed to help enterprises build, test and scale AI workloads through a guided phased approach. With AI Launchpad, Rackspace manages the infrastructure, GPUs and core tooling systems so teams can focus on validating use cases instead of building environments from scratch. You start with a proof of concept, move into a real pilot, and then scale into production on managed enterprise grade GPU infrastructure. Whether you're testing inference at the edge, fine tuning foundation models or standing up a production pipeline, the goal is the faster progress with less operational friction. If you're ready to move beyond demos and actually put AI to work, take a look at Rackspace AI Launchpad and see how a managed path to production can accelerate results. Visit rackspace.comailaunchpad to learn more. If you're using AI to code, ask yourself, are you building software or are you just playing prompt roulette? We know that unstructured prompting works at first, but eventually it leads to AI slop and technical debt. Enter zenflow. Zenflow takes you from vibe coding to AI first Engineering. It's the first AI orchestration layer that brings discipline to the chaos. It transforms freeform prompting into spec driven workflows and multi agent verification where agents actually cross check each other to prevent drift. You can even command a fleet of parallel agents to implement features and fix bugs simultaneously. We've seen teams accelerate delivery 2x to 10x. Stop gambling with prompts. Start orchestrating your AI. Turn raw speed into reliable production grade output at Zenflow Free. Most companies don't struggle with ideas, they struggle with turning them into real AI systems that deliver value. Robots and Pencils is a company built to close that gap. They design and deliver intelligent cloud native systems powered by generative and agentic AI with focus, speed and clear outcomes. Robots and Pencils works in small, high impact pods. Engineers, strategists, designers and applied AI specialists working together to move from idea to production without unnecessary friction. Powered by RoboWorks, their agentic acceleration platform teams deliver meaningful results including initial launches in as little as 45 days, depending on scope. If your organization is ready to move faster, reduce complexity and turn AI ambition into real results, Robots and Pencils is built for that moment. Start the conversation@rootsandpencils.com aidaily brief that's robotsandpencils.com aidDaily Brief Robots and Pencils Impact at Velocity Today's episode is brought to you by superintelligent. Superintelligent is a platform that, very simply put, is all about helping your company figure out how to use AI better. We deploy voice agents to interview people across your company, combine that with proprietary intelligence about what's working for other companies and give you a set of recommendations around use cases, change management initiatives that add up to an AI roadmap that can help you get value out of AI for your company. But now we want to empower the folks inside your team who are responsible for that transformation with an even more direct platform. Our forthcoming AI Strategy Compass tool is ready to start to be tested. This is a power tool for anyone who is responsible for AI adoption or AI transformation inside their companies. It's going to allow you to do a lot of the things that we do at superintelligent, but in a much more automated, self managed way and with a totally different cost structure. If you are interested in checking it out, go to aidaily Brief AI Compass, fill out the form and we will be in touch soon. Welcome back to the AI Daily Brief. For basically the last quarter of last year, one of the big topics of conversation around AI was whether we were in a bubble. Now we've talked extensively on this show about why it's important to separate the market conversation about whether we are in a bubble in terms of things like valuation and capitalization for AI companies from whether it's actually significant and meaningful technology that you need to be paying attention to. And yet still these things are not disconnected. There will, for example, be impacts to the speed at which new, better AI comes to market if additional capital and financing for companies is no longer available. Now, so far in 2026, markets frankly just haven't had all that much time to worry about AI. Between Venezuela and Greenland and domestic unrest and future government shutdowns, let's just say that the risk off dance card has been a little full yet. Of course, this question about an AI bubble has never really gone away. In fact, part of what makes it such a potent conversation is that it is ultimately a debate. There is no way, at least in the short term, of knowing who is right and who is wrong. There is only the loudness with which either side can present evidence. Now, with all that as background, we got some of our first chance to see where markets are in a renewed version of that conversation as we got Meta and Microsoft earnings, Bloomberg host Caroline Hyde summed it up this way Meta wins, Microsoft loses when it comes to after hours market reaction to earnings. Both beat on earnings and revenue both yanked up capex. Both talked of the positive impact of AI spending, but investors found it hard to see the CapEx rewards for Microsoft. Given the slight slowdown in Azure growth rate versus the previous quarter. Zuckerberg and team did a better job at highlighting the strength of its AI. Investments are already bestowing on its recommendation system, its video impressions and therefore its ad success and 25% revenue growth. Zuckerberg also spoke excitedly about the coming year in terms of LLM improvements, meaning current systems are primitive compared to what will be possible soon. That's his quote. So that kicks us off with a high level, but let's dive in and go a little bit deeper. The theme of Meta's earnings call was full steam ahead on AI regardless of costs Mark Zuckerberg painted a bold vision of what the next few years of Meta's AI product would look like now. Seen as one of the early, if not only winners in AI wearables, Zuckerberg is doubling down on making Meta Ray Bans the default options for portable AI. He told investors, billions of people wear glasses or contacts for vision correction, and I think that we're at a moment similar to when smartphones arrived and it was clearly only a matter of time until all those flip phones became smartphones. It's hard to imagine a world in several years where most glasses that people wear aren't AI glasses. And while one could be forgiven for being a little bit concerned about hyperbole there, frankly Meta's numbers sort of validate that vision. Sales have tripled over the past year, which Zuckerberg characterized as some of the fastest growing consumer electronics in history. Now, when it comes to AI models, Meta has so far failed to make contact with the consumer, but it's clear they hope to turn that around with their next generation. Zuckerberg basically characterized 2025 as a rebuilding year, and although he didn't make this comparison, Google's turnaround on AI of a couple years ago to now has to be on the forefront of investors minds, said Zuckerberg. In 2025 we rebuilt the foundations of our AI program. Over the coming months we're going to start shipping our new models and products and I expect us to steadily push the frontier over the course of the new year. Unlike some of the other AI companies, for Zuckerberg, commerce is the clear focus. He talked, for example, about new agentic shopping tools that will allow people to find just the right set of products from the businesses in our catalog. Meta also aims to catch up on agents, which are of course dominating the enfranchised insider conversation as new products like cloudbot demonstrate just how far the technology can go, Zuckerberg said, we're starting to see the promise of AI that understands our personal context, including our history, our interests, our content and our relationships. A lot of what makes agents valuable is the unique context that they can see, and we believe that Meta will be able to provide a uniquely personal experience. Now, on top of all that chatter, the actual earnings were incredibly strong, giving Meta a solid base to pursue these bets. Meta delivered a year over year growth rate of 24%, which blew analysts estimates out of the water. And yet at the same time, the number that got the most attention was of course, a massive increase in Capex guidance. Meta said that they plan to increase spending on data centers to as much as $135 billion this year, around 20% higher than the median analyst forecast and and almost double what they spent last year. They also announced a 40% rise in expenses, reflecting their huge salary commitments to their new AI research team. Now Meta is generating around 60 billion in quarterly revenue at the moment, so these spending goals suggest reliance on debt funding. Still, Zuckerberg has said that the aim is to front load capex and build as fast as possible to get back in the AI race. Rounding out his statement, Zuckerberg told investors, this is going to be a big year for delivering personal superintelligence, accelerating our business, building infrastructure for the future, and shaping how our company will work going forward. Forward now. Some had big critiques of Meta's apparent lack of a plan. One analyst asked Zuckerberg to sketch out his three, five or ten year plan for delivering ROI on all of the AI spending and Zuckerberg kind of sidestepped the question. He responded, I have to say up front that I think my answers to a lot of your questions on this particular call may be somewhat unfulfilling because we're in this interesting period where we've been rebuilding our AI effort. We're six months into that and I'm happy with how it's going. But we are going to be rolling out our initial set of models and products and businesses around that and I'll have a lot more to share on all those fronts at that point. He added, not much of this is going to be particularly detailed, but it will be exciting as we roll it out. Others honed in on Meta's new focus on debt financing, Kakashi on X wrote, zuck has finished almost all of Meta's cash. Without the issuance of 30 billion in debt, Meta's cash cash equivalents and restricted cash equivalents would be less than 10 billion by the end of 2025. Prominent finance commenter Michael Green said the risk embedded here is astonishing. And yet ultimately it seemed that investors liked what they heard. With meta shares rising 8% in after hours trading overnight, the story from Microsoft earnings Call was something of the inverse, where the perception was that too much caution had led them to miss out on AI opportunities. The headline figure, as we heard from Caroline at the beginning, was a continued drop in cloud growth. One could be forgiven for thinking that this slowdown doesn't seem to actually be that large of a problem. Azure revenues are still growing at a 38% pace, and the slowdown was only a single percentage point compared to the prior quarter. Azure is also a much larger division than it was a year ago, making hypergrowth more difficult to maintain. Still, we are in the middle of an AI infrastructure boom, and to some it seems like Microsoft is failing to take part to the fullest extent. Caution has come to define Microsoft's strategy over the past year. Last February, for example, there was a massive news cycle around Microsoft canceling numerous data center projects. CFO Amy Hood insisted that this was largely about forecasted softness and demand, and Microsoft didn't want to be left holding useless data centers if demand dropped off, given the long lead times to complete construction and power up GPUs. Compare that sort of framing to Zuckerberg, who has basically relentlessly said that the risk of overspending by hundreds of billions on infrastructure is dramatically less than the risk of underspending. In any case, when it comes to Microsoft, they're now faced with a different problem. They disclose that their cloud sales backlog has more than doubled from a year ago to reach 625 billion. Much of that growth came from a new $250 billion commitment from OpenAI. In total, OpenAI now accounts for 45% of Microsoft's backlog. And so whereas Meta's stock jumped 8% overnight, Microsoft stock fell by about 5% in after hours trading. And indeed, going back to the theme of this episode, Bloomberg suggested this was fears of the AI bubble popping manifesting in the market. Yet to me, that doesn't really carry water. It seems more like Microsoft is being punished for failing to grow as fast as possible during this boom period. They have over a trillion dollars worth of sales, waiting in their backlog and slowing cloud growth rather than powering up GPUs. Then Microsoft chose to miss out on sales. To the extent that there are bubble fears, they're almost entirely about OpenAI's ability to make good on their commitments, Jeffries analyst Brent Thill said. The backlog is really good, but the disclosure that OpenAI is 45% of their backlog. It goes back to the situation where can OpenAI achieve these financial goals to pay Oracle, Microsoft and many of the providers Now Microsoft does appear to be correcting their underbuilding issue, if ever so slightly. Capex for the quarter came in at 37.5 billion a, up 66% from a year ago and slightly above analyst expectations. They didn't adjust CapEx guidance, so it seems they won't be attempting to catch up in a major way. Of course, for investors, it's not as simple as just spending more on capex. Microsoft also disclosed an operating margin of 45.1%, falling short of the 45.5% consensus forecast. Not a huge miss, but spending more money on capex now will make operating margins worse in the short term due to the construction lag overall. CEO Satya Nadella doesn't seem too bothered by the state of play and is happy to play the long game. He told investors, we are only at the beginning phases of AI diffusion and already Microsoft has built an AI business that is larger than some of our biggest franchises. We are pushing the frontier across our entire AI stack to drive new value for our customers and partners. He reiterated those themes in a thread on X actually spending a bunch of time on agents, nadella wrote. Like in every platform shift, all software is being rewritten. A new app platform is born. You can think of agents as the new apps. We're providing everything customers need to build their own agents, including model catalogs, harnesses for orchestration, context, engineering plus AI safety and observability. The bottom line, he says our TAM will grow substantially across every layer of the tech stack as this diffusion accelerates and spreads. And we feel very good about how we're delivering for our customers today and innovating to capture the opportunity ahead. My honest read here is that as much as some commentators are trying to make this just about capex spending, it's more about a conversion on AI and b perception of strength in the AI race. Zuckerberg advised a way bigger capex jump than Microsoft did and got rewarded by traders. That means that this can't be about CapEx alone, but as we heard the story with Meta was a big increase in the ad business. That is a conversion on AI spend and tailwinds in the AI race because while their models haven't kept up with others, they do have a thing that they are the undisputed leader of in this AI race, which is of course, the Ray Bans and AI wearables. Compare that with a theme explored by Bloomberg columnist Dave Lee in his piece published on Thursday morning called Microsoft Has Lost its AI Sparkle. Dave writes, who could forget those heady days when Microsoft Chief Executive Officer Satya Nadella looked like the smartest man in tech? When ChatGPT emerged in late 2022, his decision to back OpenAI put Nadella's company at the forefront of the AI boom. More recently, the relationship has cooled and with it the perceived value of Nadella's foresight. Microsoft has now invested in Anthropic and look to incorporate its capabilities into its products, and OpenAI, hungry for computing power, has turned to Oracle, Google and Amazon. All of this AI polyamory has put Microsoft's eggs in a few more baskets, but it has also highlighted that Microsoft's early mover advantage has run its course. The AI sparkle that illuminated the market value that more than doubled has diminished. The thorny issue of ROI is stalking all the top AI players, but the pressure has mounted on Adela because of the shifting narrative of the AI industry. The AI story of 2026 so far is one that says Google, with its Gemini model, is crushing it with consumer use cases and benefiting from the avail and lower cost of its own chips. At the same time, Anthropic's coding capabilities are at the leading edge of redefining software engineering. Whatever the reason, with its huge base of enterprise users and ownership of the coding mega platform GitHub, Microsoft stands accused of letting its advantages dwindle. Now, Dave does end on a positive note, reminding us all that Microsoft's critics today were Google's critics yesterday, but pointing out that Microsoft has to do something to shift the narrative back in its favor. Now, one more market story to really round things out and let us know where we stand. We also got earnings reports from South Korean memory giants Samsung and SK Hyeing. If you've been keeping an eye on RAM prices recently, you probably already know how this went. Samsung reported that profits had doubled from this time last year, reaching 13.7 billion. This was a 61% increase from the previous quarter, vastly outpaced already lofty analyst estimates. Stripping out just the memory division, profits were up fivefold year over year. Samsung said that they would, quote, proactively address market demand by focusing on high value added products. In other words, they will prioritize high bandwidth memory for AI chips and make as much money as they possibly can. The story was similar for SK Hynix who doubled operating profit to 13.5 billion, again a massive beat on estimates. Numerous comments from analysts noted that the current memory shortage isn't likely to see relief anytime soon. Richard Claude of Janus Henderson said there are no easy wins and levers from a supply perspective to meet this new demand driver compared to consumer electronics companies. He added, hyperscalers and AI customers are looking at what they are willing to pay for memory through a very different lens. Now both memory companies are keeping capex relatively restrained. Samsung reported that their capex had actually declined slightly in 2025, commenting that they are maintaining a conservative investment approach. They did state they expect to ramp up investments this year, but declined to give full guidance. SK held their guidance steady at around 30% of revenue. However, increase in revenue means they'll be spending quite a bit more, but they don't seem to be looking to stand up extra capacity in any great hurry and risk overbuilding. Citigroup analysts expect the cost of DRAM to rise by 120% this year and NAND chips to increase by 90%, so no one is expecting the shortage to be rectified in the short term. Remarking on the nature of tech spending in this cycle, Sanjeev Rana, the head of research at CLSA Securities Korea, commented, the companies are spending real money on real stuff. We're in uncharted territory in terms of valuations, share prices, the demand cycle. Everything is unprecedented. Overall, from this first glance at what the bubble narrative might be in 2026, it feels like the big takeaway from this early round of tech earnings is that there is still pretty much a full on gold rush in AI. Yes, bubble fears are present and yes, we're seeing how markets are going to want to see translation of AI spending into something resembling increased growth either in cloud business or in ad sales when it comes to meta. But the market seems to be rewarding firms not only for taking part in the boom, but for where their place in the narrative is. So that's the story from here and that's going to do it for today's AI Daily Brief. Appreciate you listening or watching as always. And until next time, peace.