OpenAI-Linked Stocks Slip; Investors Focus on AI Ahead of Big Tech Earnings

0:00Is your multi-entity management creating more confusion than clarity? You need the Intuit ERP. Intuit Enterprise Suite. It's the AI-native ERP solution that's powerful, painless, and proven. Learn more at intuit.com slash ERP. If you follow markets, you know the value of long-term thinking. You plan, you diversify, you prepare for volatility. But in life, even the best strategies can't prevent every bad day. A fire, a loss, a disruption that demands immediate attention. When that happens, what matters isn't just what you planned, it's who shows up. That's where Cincinnati Insurance comes in. For more than 75 years, they've helped individuals and businesses navigate life's toughest moments with care, expertise, and personal attention.

0:49Together with independent agents, Cincinnati Insurance focuses on relationships, not transactions. Their approach is grounded in experience, follow-through, and trust built over time. Bad days happen, and when they do, you deserve an insurance partner who understands risk, respects what you've built, and is ready to help you move forward. The Cincinnati Insurance Companies. Let them make your bad day better. Find an independent agent at c-i-n-f-i-n.com. So there's a lot of noise about AI, but time's too tight for more promises. So let's talk about results. At IBM, we work with our employees to integrate technology right into the systems they need.

1:29Now, a global workforce of 300 ,000 can use AI to fill their HR questions, resolving 94 % of common questions. Not noise. Proof of how we can help companies get smarter by putting AI where it actually pays off. Deep in the work that moves the business. Let's create smarter business. IBM. Bloomberg Audio Studios. Podcasts. Radio. News.

2:18on Bloomberg Radio. So a very big week when it comes to tech earnings. And it comes because we are, I think, continuing to question the AI spend, the return on investment when it comes to artificial intelligence, that build out. And some of it had to do with the conversations we've had around OpenAI today. Yeah. The backdrop of questions over the growth of that company, the Wall Street Journal reporting late yesterday that OpenAI had fallen short of several internal targets as rivals gained ground. OpenAI described the report as, quote, prime clickbait, though we did see that report resonate far and wide in today's trade.

2:57Oracle shares took a hit. Core Weave was down. ADRs of SoftBank, an investor in OpenAI, took the hit. The journal also reported that OpenAI CEO has expressed concerns, or CFO rather, has expressed concerns that the company may not be able to afford its reporting, its future computing needs if sales don't grow fast enough. Yeah, we've seen OpenAI push back about some of those concerns. Also today, Amazon says it will make OpenAI's AI models available to its customers after Microsoft dropped its exclusive rights to sell OpenAI products. So take that. That's it? That happened today? That was like a full page of things that happened.

3:29It's kind of why we wanted to pull in three key members of the Bloomberg News team and the Bloomberg Intelligence team and really just kind of do a deep dive into what's going on with OpenAI. But AI generally, especially as we get ready for these big tech names to report, and they're going to obviously be weighing in on the AI spend. With us right now is Mandeep Singh, Bloomberg Intelligence Global Head of Technology Research in Boston. We've got also with us Sarah Fryer, Bloomberg News Big Tech Team Leader. She's out there, I believe, on the West Coast. And Rachel Metz, Bloomberg News AI reporter.

4:00Sarah and Rachel joining us from specifically the Bloomberg San Francisco Bureau. Guys, thank you, thank you, thank you so much. Mente, if I want to start with you, that open AI, and they pushed back, and now they've got a new partner in Amazon Web Services. What's your read on that news? I mean, look, we had the signs since December when they announced the Code Red. I think I was surprised by, you know, the$120 billion funding round that they had at$850 billion valuation. And now it seems like, given they miss the target, it has to do with the distribution they had through Microsoft and what they're looking to do with Amazon.

4:44So it sounds like they feel they can get more distribution with Amazon to the extent they are probably giving away some of the revenue they were getting from Microsoft. So a little bit of a shift in terms of ecosystem, but I don't think we are at a point where you know, Entropic is the clear leader and OpenAI is falling behind. It's still too early to call that out. Rachel Metz, come on in here. You're Bloomberg News AI reporter. You cover these companies and more very closely. How is this report and sort of the indication or the idea that there might be some targets that are not being met, some spend might be being pulled back?

5:22How is that resonating in your space today? I mean, I think you're clearly seeing people wondering what is going on, if something's going on. I mean, the company really pushed back on this. And it's also a little bit unclear to me, like how any moves they've been making with Codex, they've really been pushing Codex, which is their AI coding product lately, how that might be counterbalancing what they're saying here. So it's going to be interesting, like a few months from now to sort of see what happens. But, you know, there's not going to be one winner here. And these companies always seem to be shifting places.

5:58So I think that is something that we should expect is that things are going to be changing all the time in this like really nascent market. Yeah, I get a little nervous. I mean, we talk about the circular financing, that it seems like the money going back and forth. Sarah, come on in on this, because I do wonder, do we feel like we have enough transparency on a company like OpenAI? It's private, a lot of money going in. We talk about these valuations, but do we really kind of understand the business? You know, we're reporters. We always want more transparency, right? I think once we see if they do have an initial public offering at some point later this year, that'll be fascinating.

6:35I think the SpaceX filing will maybe give us insight into the space through XAI. We'll see a little bit of the financials of an AI company. XAI, I don't think, is at the level of OpenAI. But, you know, through some of these edging closer to public markets that these companies are doing, we're going to see and hear a lot more about their finances. And we're going to hear a lot more about what they're doing to shore up their finances to try to, to get to the point where they have a, you know, a line of a few revenue lines that are really in infrequent use by actual customers that are paying the bills for all of that compute power that is so necessary for continuing to grow.

7:17And listen, they have to continue to grow. That is the only way that they are going to continue to get investment in the private markets and in the public markets eventually is showing that this is just the beginning, as Mandeep said, that this is just the foundation of what will eventually be this much more transformative, much more lucrative change to the economy and how we all do business and how we all live. That's the promise. No, right, totally, right? And we see this kind of day in and day out. Mandib, I feel like you've been pretty optimistic about what's going on in terms of AI, but do we feel like we have enough transparency on the balance sheet of open AI and really understanding their business?

8:01And I do wonder, we had a guest on earlier, Rudina Cessari, she's a venture capitalist. She's really focusing on agentic AI and thinking about vertical forms of data and saying that at some point, we won't need all these large language models that not necessarily all of these will continue to exist in the forms that they are today. And I'm just curious where you are on that. So you're right. Look, these models have different areas where they do well. And what Enthropic has shown is you can have, you know, expertise in one area. And that will give you a very long runway, as they have shown with coding agents.

8:39And now everyone is chasing coding agents. So to my mind, you know, these model providers will try to tap in what the best use case is. Everyone was going after customer service for a while until they realized, you know, coding agent is a more lucrative market. And look, that's where it will be use case specific. It will be vertical specific. But there is no doubt that, you know, the numbers are getting very big in terms of both the CapEx investments as well as, you know, the funding rounds. Like the$120 billion funding round,$50 billion was committed by Amazon. And the$5 billion that they gave to OpenAI is going to flow through in the form of AWS revenue over the next 12 months.

9:25So that circular financing aspect is huge as well. But they have to find a lot more use cases like they have found this big coding agent use case. You need probably 10 of those to justify the level of funding that is going on and the CapEx investments that are being made. So, Rachel Metz, what are the different use cases that not just these companies are talking about, but other AI companies in your universe are talking about? I mean, Jensen Wang has been talking about physical AI for a few months and longer than a few months. But, you know, and Elon Musk with XAI and with Tesla, the optimists, robots, humanoid robots as well.

10:05What are these companies thinking as these 10 other use cases? Yeah, that's a great question. And I guess I should say that while there are a bunch of other use cases, and I'll detail some of them, they're all really, really early stages still, as far as I can tell. Obviously, coding is one that has taken off and people are using it and they say that they're seeing a return on investment. And it is one that's great because it's very easy to generate the code and then you can check it. Does the website work? or can you use the app that you make? That's like a great use case in that it's pretty easy to see if it's actually working and working well.

10:42Some of the other areas that people are looking at include materials discovery, drug discovery, and a lot of things related to molecules, science, scientific work in a bunch of areas. Biology-related things are pretty hot. Google's done a lot of work in that area in particular and OpenAI and Anthropic are also working there. You see various things related to math. You also see like a really big push lately, especially from OpenAI and Anthropic to build out things that people can use for office work. Things that we have to do all the time in offices, but we probably don't enjoy very much like making a PowerPoint presentation, working on spreadsheets, things that are necessary a lot of times when you work in an office environment, but they're not like the fun parts of your job.

11:32Um, some of those, I think we're seeing more and more traction. We're also seeing, um, increasing, uh, uses of like design work. We're seeing like legal work, those kinds of things, but they're all really early stages. Um, for instance, there are no drugs on the market yet that have been like largely discovered or come about because of AI. So Mandeep, come on back in here. I mean, how do we know, like, what do you look for, for signs of froth or, you know, this is getting a little silly. I know a lot of people make comparisons to the tech craze, 1990, late 90s into 2000. And of course, the tech bubble bursting.

12:09It doesn't feel like the same thing to me. But I'm just curious, how do you look for where maybe the silliness is or the euphoria in terms of the AI build or the AI trade or the AI spend? Yeah, I think when one of these companies these files their prospectus, you know, for their IPO, will know exactly what is the margin profile of these businesses, how much money they are losing. And clearly, right now, they talk about, you know, how training is separate from inferencing compute. But when you look at the overall, you know, income statement, we'll get a very good sense of, you know, what the long-term gross margins are going to be in this business.

12:50And they clearly will be far below where the current software company's gross margins are, around 70 to 80%. It won't be that because we know generative AI is more compute intensive, but it's just the level of cash burn and what the long-term margin profile, that will determine how much investment investors will continue to make in these businesses. And look, Microsoft just talked about adding Accenture, you know, 700 ,000 plus employees for Copilot. That's a great anecdote in terms of if they're able to retain, you know, such a customer for a long time, because that adds real revenue, recurring revenue that they can show over time.

13:31But the question is, is Microsoft Copilot the right kind of product or is it more agent, a coding agent, that kind of a product? And I don't think the jury is out yet in terms of what is more sticky when it comes to generative AI. Sarah Fryer, I want to bring you in here and just talking about the comparisons, the historical comparisons here, because we've spoken about this in the past. But I think many would argue, especially with the way that these employees are being compensated, the venture capital that's going into some of these firms, the rounds and how valuable and how much money these companies are raising before they even IPO.

14:10that's unparalleled historically. So if there were some sort of bubble to burst, that would have pretty wide ranging consequences. It's all so interconnected, right? Google just said it's going to invest up to 40 billion in Anthropic. It's also competing with Anthropic on agentic coding. I mean, there are so many examples like that every day in the news. We have a lot of dominoes that could fall if something goes wrong. But listen, it's also built upon the foundation of what big tech has already become so good at and sort of the foundation of the Internet itself. Right. Like it's built on on Google search business, Meta's social media business.

14:59Like all of this is the money that is going into AI and feeding back into the growth of this new technology. So it is very interconnected, but it's also all in service of, you know, making that growth possible via these legacy businesses. They're all trying to find what the next big business is. They've all decided that it's AI. And so now they're all competing with one another, but also investing in one another. So it's quite a financial party. Yeah, really close. I will say, I'm just going to point out, if you have a Bloomberg terminal, check out John Arthur's latest column. It's called, Hate to Suggest Partying Like It's 1999, But...

15:41Dot, dot, dot. Dot, dot, dot. Yeah, check it out. I don't know. Well, Rachel, you've got the big picture, obviously looking at ongoing developments in AI. I mean, it feels like every day there is new things. Or do you feel like, as you continue to cover it, that you are seeing some leaders and some laggards, if you will, kind of form in this industry? I mean, I think at the moment you see a ton of attention heaped upon Anthropic and OpenAI, and you see a lot of that back and forth, back and forth. But I don't think that this is going to be ultimately a situation where you have one or two companies.

16:15I mean, it's clear at this point that we're only going to have a handful of companies building extremely computationally intensive and large AI models. Like, that's clear. It's so, so expensive. So, you know, there's only a limited amount of money that companies can gather for that. But ultimately, I think you're going to have a lot of different companies that are doing lots of different aspects of AI, building different types of models. And I think we're just going to have to see in a few years how things shake out. And I wouldn't forget about legacy companies, big tech companies like Meta and Google.

16:47I mean, I think it's really important to remember that those companies have been working on AI themselves for a very long time, created some really important foundational research, some of the key research that companies like OpenAI and Anthropic are now drawing on for their own models. I want to bring back Mandy, who covers the public companies, many of the public companies in this space. But to ask you a question that we kind of spoke about a little earlier with Eddie Gabor over at Key Advisors, and that's about what the actual effect on companies that are not native to this technology, but are deploying the technology.

17:23So the ones that are right now using AI to actually decrease headcount, to increase productivity, are you seeing yet the proof in the pudding that this technology actually works, Mandeep? Yes. And I think the most visible point is tied to developer productivity. I mean, all these large tech companies last year were touting the fact that 30 % to 35 % of new code was written by coding agents. This year, the number seems to have gone up to almost 75 % based on the couple of anecdotes that we have heard so far. So clearly, you will see the impact in terms of headcount additions around that aspect in particular.

18:08But look, all of these companies are looking for more productivity across the board. And I I think that's where the deployment of agents can have a multiplier effect. One employee using multiple agents to do tasks in parallel. And I think for use cases where the agents can do longer duration tasks, you will see that productivity benefit, and then it will show up in terms of how these companies are talking about headcount. I mean, Microsoft did that voluntary, you know, kind of headcount reduction that they, the offer that they gave to their employees. That's huge in the context of how they are thinking about the future.

18:57And I'm sure, you know, a number of IT services companies are probably thinking along the same lines as well. All right. So I want to move it forward to what we're going to be talking about in less than 24 hours time and all of those earnings that we're going to get after the closing bell. And for so many of them, you know, Meta, Microsoft, Amazon, Alphabet, obviously we're going to be looking for confirmation of this AI spend and build and what they have to say. Sarah, let me start with you. The overview, the big tech team will be all over it. I mean, it's important what we get from them. I mean, this is what we're going to be looking for, again, when they report.

19:33There's already so much enthusiasm for spending in AI. These companies that are reporting tomorrow have planned$650 billion in AI infrastructure, AI spending for this year. Just a completely mind-boggling, unprecedented number. And the investors that are listening on their earnings calls tomorrow are going to be wondering, what's your conviction on that money? Are you even able to deploy it at the rate that you want to deploy it? And are you seeing efficiency gains? Are you seeing revenue that is directly related to progress in AI? At Meta, for instance, you see it already in their advertising business.

20:16It has been dramatically accelerated by improvements they've made to their ad targeting algorithm to the way that they show users content with AI. So that's a company where you can see the benefits in the business already from AI, and they're just doubling down. At a company like Google, where they're thinking about how does search approach the era of AI? How fast can we tweak it while also keeping in mind that it is our cash cow? It's what's funding our growth. Those are harder questions, right? Those are more difficult. But Google has really shown leadership in a lot of what it's done with Gemini, the way it's integrated it in and built it for enterprise customers.

20:59So we want to see, investors want to see whether their cloud business is keeping up with that progress and whether they're seeing more demand. The cloud business really across the companies, across Amazon and Microsoft and Google is going to tell us so much about AI demand. Because if you are a company that's going all in on transforming yourself with AI, you're going to need one of those cloud providers to be the provider of the compute to run those models. And so it's a really significant indicator for us of how much the enthusiasm is progressing. And I think investors are probably going to want to see themselves be blown away or else we might get some negative reactions.

21:44Even if the numbers look good, if they're not great, we may see some disappointment even on good numbers. So, Rachel, I want to bring you in here on what Sarah is saying and the idea that what investors are looking for and, you know, they want to be blown away. Look, they want to see margin improvement. They want to see efficiency. but meta platforms being better at targeting ads through AI technology isn't really like, you know, creating new drugs to help cure rare diseases. Like not yet, but I mean, is this, this is what, this is what we're getting right now, Rachel. Yeah. I mean, I think that's, uh, like a decent benchmark for the reality of the situation is that there are some things that you can do with this technology right now.

22:35There are a lot of things you can't. And, you know, and Meta is really good at doing these sorts of things. I mean, Meta has spent years working on targeting advertising, has been using AI to target ads for quite some time. And it's the kind of company that I would hope would be good at it, right? They've spent all this money and time and energy on building that up. I mean, we're going to I think we will see some applications start to emerge, especially over this year. I mean, it seems to me over the past couple of years that some of the larger AI companies have sort of picked a couple things that they're targeting.

23:08You can see that with their messaging, like coding, for instance. You know, people started to realize, hey, we can do this quite well. And then you see more messaging about it and you see more products related to it. We'll probably be seeing that more with a couple other applications this year, such as some scientific applications, perhaps. and then more applications will start to come out. Some of them might be accidental things that people just realize, hey, wait, we've been using this in the office for this and it's great and others won't. But I think we're not going to see, we're going to see a lot of things that aren't the most exciting and we're just going to have to be okay with that, I think.

23:46Patience, patience. Hey, Mandib, final thought for you. You know, Meta, Alphabet, Amazon, Microsoft, How important is what we get from them tomorrow for showing whether or not, you know, this AI narrative that we've been talking about for, what is it now, three years, like, makes sense? What are you watching out for? Sure. So all of these companies have guided for over$100 billion in CapEx for 2026. And so when you roughly equate two gigawatt capacity, it is around two gigawatt capacity based on our calculation. So you want to see what is the type of revenue per gigawatt that they can show as ROI or monetization around the AI data center capacity that each one is investing in.

24:32And at what margins are they able to do that? And the other aspect I would call out is the impact of higher memory prices, because clearly that has been a big trend over the past few months. And so you want to see how that trickles through when it comes to this AI data center build out. Thank you so much, you guys. Well set up for what happens tomorrow after the closing bell. Or Mandeep Singh, Sarah Fryer, and Rachel Metz, of course, all of our Bloomberg team here at New York headquarters reporting this out. Stay with us. More from Bloomberg Businessweek Daily coming up after this. If your finance team spends more time finding data than using it, if there's one entity here and one here and one here and one here, if scaling your business feels like starting over, you need the Intuit ERP.

25:24Intuit Enterprise Suite is the AI-native ERP solution that's powerful, painless, and proven. Learn more at intuit.com slash ERP. Everyone has been there. Your team's feedback is scattered across emails, chats, and sticky notes. It's a mess. But PDF Spaces and Adobe Acrobat gives you one collaborative workspace to streamline every file and comment. So, if you need six departments to finally agree on a proposal, do that with Acrobat. Need to turn a mountain of feedback into one plan of action? Do that with Acrobat. Want to stop searching for files and finally get everyone on the same page? Do that.

26:04Do that. Do That With Acrobat. Learn more at adobe.com slash do that with Acrobat. Support for the show comes from Public. Public is an investing platform that offers access to stocks, options, bonds, and crypto. And they've also integrated AI with tools that can assist investors in building customized portfolios. One of these tools is called Generated Assets. It allows you to turn your ideas into investable indexes. So let's say you're interested in something specific like biotech companies with high R &D spend, small cap stocks with improving operating margins, or the S &P 500 minus high debt companies.

26:41Chances are there isn't an ETF that fits your exact criteria. But on public, you just type in a prompt and their AI screens thousands of stocks and build a one-of-a-kind index. You can even backtest it against the S &P 500. Then you can invest in a few clicks. Go to public.com slash market and earn an uncapped 1 % bonus when you transfer your portfolio. That's public.com slash market. Ad paid for by Public Holdings. Brokered services by Public Investing, member FINRA SIPC. Advisory services by Public Advisors, SEC Registered Advisor. Crypto services by ZeroHash. Sample prompts are for illustrative purposes only, not investment advice.