In short
This Bloomberg Tech episode covers AI infrastructure deals, AI-driven labor shifts, major legal fallout in the OpenAI/Musk dispute, and data-center expansion.
Guests and backgrounds
Mandeep Singh (Bloomberg Intelligence tech research) analyzes the Google–Blackstone “NeoCloud” plan. Marta Norton (Empowered Chief Investment Strategist) discusses AI-related workforce changes and adoption. Dorothy Lund (Columbia Law School professor) explains the statute-of-limitations ruling in Elon Musk’s OpenAI lawsuit. Parag Agrawal (Parallel founder/CEO) launches “Index,” a content marketplace for agentic AI. Riley Griffin (Bloomberg) reports on Meta’s Louisiana data-center impacts. Laurie Beer (JPMorgan CIO) discusses AI transformation and cybersecurity. Dan Wright (Armada CEO) explains modular AI data centers.
Key claims/examples
Google and Blackstone will create a NeoCloud using Google TPUs, with Blackstone funding ~$5B (leveragable to ~$25B) and targeting 500MW by 2027, aiming to speed TPU capacity without Google building all data centers. Meta plans to reassign 7,000 workers to AI product teams while cutting 10% staff, and is building a ~$200B Louisiana campus (5GW compute). Musk’s OpenAI case was dismissed because he sued after the statute of limitations; appeal will focus on procedural issues. Parallel’s Index pays publishers/data providers based on marginal task value using Shapley values; launch partners include The Atlantic, Fortune, PitchBook, and ZoomInfo.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Chapters
Tap a time to open that second in VOGoogle and Blackstone's AI Cloud Partnership
2:25 to 3:47
Exploring the implications of Google's partnership with Blackstone to create an AI cloud business.
“This is at plans AI-related layoffs this week.”
Impact of Google’s AI Cloud Plans
3:47 to 5:50
Discussion on how Google’s new cloud business with Blackstone might influence the tech industry and its competitors.
“Bloomberg Intelligence says this pact could be a major boost for the tech giant's chips business.”
AI-Related Workforce Changes
5:50 to 7:40
Examining the trend of AI-related job cuts and the broader implications for the workforce.
“Mandeep Singh, who leads technology research at Bloomberg Intelligence, thank you very much.”
Elon Musk vs. OpenAI Co-Founders Verdict
7:40 to 11:00
Detailing the jury's ruling on Elon Musk's lawsuit against his OpenAI co-founders and its implications.
“But yes, there are these idiosyncratic reports that suggest we are seeing a little bit of movement on that front.”
Legal Technicalities of Musk's Case
11:00 to 14:02
A discussion on the legal nuances of Musk's case regarding the statute of limitations and its consequences.
“A jury ruled that Musk waited too long to sue over OpenAI's transition to a for-profit business.”
Musk's Legal Battle with OpenAI
14:02 to 16:42
Explore the complexities behind Musk's lawsuit against OpenAI and the statute of limitations.
“And so then OpenAI said, well, you waited too long to sue.”
Appeals Court Strategies
16:42 to 17:43
Understanding Musk's options for appealing the verdict in his case against OpenAI.
“So I'll give the Internet gremlins a second to reset.”
Introducing Index by Parallel
20:46 to 24:45
Parallel's CEO discusses the launch of Index and its implications for content creators.
“is launching Index today, a new marketplace that pays publishers and data providers based on how much their content actually helps an AI agent complete a task.”
Shapley Values and Content Compensation
24:45 to 27:00
Learn about the Shapley Value concept and its role in compensating content creators.
“Number two, content that gets used in high value work gets paid more.”
Meta's Data Center Plans
27:00 to 27:32
A look ahead at Meta's ambitious data center plans and potential challenges.
“Now, coming up, we dive into Meta's mega data center plans in Louisiana and some of the hiccups along the way.”
Show all 20 chapters
Semiconductor Market Insights
27:32 to 28:01
Insights into the recent performance and outlook of the semiconductor market.
“There's still a semiconductor story in financial markets.”
NVIDIA and Dell CEOs on China Market Dynamics
28:01 to 29:50
Learn about the perspectives of NVIDIA and Dell CEOs regarding the potential for the Chinese market to open up for American technology companies.
“what's going on, but we get NVIDIA earnings on Wednesday.”
Meta's AI Workforce Changes
29:51 to 31:30
Discover how Meta is restructuring its workforce and focusing on AI-related products amid economic challenges.
“Your net conclusion on whether or not China will become open to American technology companies to do business there?”
Impact of Meta's AI Facility in Louisiana
31:31 to 33:54
Explore the economic implications of Meta's AI facility in Louisiana and the local community's response.
“And there's big readership on this, a lot of interest.”
J.P. Morgan Tech Conference Insights
33:55 to 36:12
Hear insights from Laurie Beer on the accelerating pace of change in technology and its implications for financial services.
“So will 500 jobs be enough to provide that economic lifeline long term?”
AI Tools and Job Market Dynamics
37:26 to 42:00
Delve into the discussions around AI tools, job market changes, and the balance of innovation and risk in technology.
“Morgan Tech Media and Communications Conference in Boston, where Bloomberg Surveillance co-host Lisa Abramovitz is standing by.”
Balancing Innovation and Trust in Banking
42:00 to 43:56
Learn how banks are navigating innovation while maintaining customer trust.
“going at the appropriate speed of innovation.”
Transforming Banking Over the Next Decade
43:56 to 44:10
Explore the transformative changes expected in banking over the next ten years.
“At the end of the day, we have to maintain that trust with customers and clients, and we have to make sure that their unique needs are served across the over 100 countries we operate in.”
Breaking News: Apple's Hardware Restructuring
44:10 to 45:26
Gain insights into Apple's hardware development reorganization to enhance product speed.
“Thank you, Bloomberg Surveillance co-host Lisa Abramowitz there.”
Armada CEO Discusses Modular Data Centers
45:38 to 49:55
Discover the revolutionary concept of modular data centers and their impact on deployment.
“and be focused on helping train new AI models.”
Transcript
Automatic transcript. May contain errors.0:00The right technology can strengthen human judgment. That's why Deloitte brings together AI and data analytics with multidisciplinary teams who can help you connect the dots across your enterprise. From risk to operations to customer needs. So opportunities don't slip by and surprises don't spread. Because the smarter your systems, the sharper your instincts. That's how technology makes people better at what they do best. Deloitte. Together makes progress. Learn more at Deloitte.com slash Together Makes Progress.
0:58mobile app is available for select mobile devices. Message and data rates may apply. JPMorgan Chase Bank N.A. Member FDIC. Copyright 2026. JPMorgan Chase and Company. Sonesta Travel Pass makes traveling more rewarding. Designed to help you get more out of every stay. Sign up at Sonesta.com to enjoy instant savings, bonus points, and valuable perks like early check-in, late checkout, room upgrades, and free stays over time. With Sonesta Travel Pass, every stay brings you closer to your next reward. Choose from more than 1 ,100 hotels across 13 distinctive brands and unlock the best available rates when you book direct with Sonesta Travel Pass.
1:38Here today, roam tomorrow. Join now at Sonesta.com. Terms and conditions apply. Bloomberg Audio Studios. Podcasts. Radio. News.
1:56Bloomberg Tech is live from coast to coast with Caroline Hyde in New York and Ed Ludlow in San Francisco.
2:07Caroline Hyde:This is Bloomberg Tech. Coming up, Google agrees to create an AI cloud business with Blackstone, which will run on Google's homegrown AI chips. Plus, the legal battle between Elon Musk and his open AI co-founders ends with a jury ruling Musk waited too long to sue. We'll dive into the details of the verdict. And Meta bets big on Louisiana for the world's biggest AI facility. This is at plans AI-related layoffs this week. AI-related layoffs, a theme for the market. But AI, lack of exuberance, pounding the market once again with actually yields pushing higher. We're still focused on geopolitics, Ed.
2:43But we're looking at a third straight day of losses on the Nasdaq 100. We're off by about a percentage point. We're off by about 3 % in those three trading days. As we just get a little bit of worry that we went too far, too fast on the hardware side of things. I shine a light on the semiconductor index. The socks are currently off by 2%. In fact, we're down by 8%, 9%. Look, we're on the cusp of some sort of recalibration of where we've been in this run-up. Yes, it was about 70 % from the end of March. But now we're on the verge of a correction. We're down almost 8 % or 9%, as I say. But a rotation.
3:12Some money moving into the software stocks. We've got key earnings coming from software and hardware over the next few days.
3:19Caroline Hyde:Neoclouds down. The news, Blackstone and Google forming a new company to be a Neocloud. Blackstone brings$5 billion. There's debt financing. 500 megawatts is the target. Google's TPUs, Tensor Processing Units, are the chips. And look at those names that are down. CoreWeave, Iron, Nebius in particular, all very NVIDIA-focused architectures. But Neoclouds all the same. competition seems to be coming. It does. Let's discuss the impact of Google. Blackstone, the partnership ed. Bloomberg Intelligence says this pact could be a major boost for the tech giant's chips business. Manip Singh, who wrote that react, joins us now.
3:55And what's so interesting is, yes, the equity comes from Blackstone,$5 billion. It could be leveraged up to$25 billion. But this is just another way for Alphabet to get its hardware out there, but not actually have to build the data centers, it feels like. Yeah. And look, when it comes to NVIDIA, their ecosystem of NeoCloud players is pretty big. I mean, you had a few of those on your screen. And when it comes to Google, you have got terrible cipher mining, but there aren't a lot of TPU providers. And that's where when Anthropix says they want to use almost$200 billion of capacity from Google, Google has to find a way to wrap that up outside of their own cloud.
4:38And that's where I think this sort of transaction is quite interesting, where Blackstone is putting the money and also has got live capacity that they can bring online by 2027, which is what makes this very attractive from a Google standpoint.
4:55Caroline Hyde:Hey, Mandy, what's the Bloomberg Intelligence thesis on the business model of setting up a NeoCloud for the TPU? Right. Because Google Cloud, you know, they have capacity that they could give to third party customers. Why set up a new standalone Nuco? Yeah, so in the case of Google Cloud, they are serving that traffic from Google's data centers. I mean, in this case, having a new cloud provider use Google TPUs and then everything else is taken care of by a blackstone in this case, whether it's the power or the cooling or even getting the land for that data center. So from that perspective, having that live capacity is what makes this different from Google having to do it on their own, which will be a lot more staggered in terms of how much gigawatt they can bring online in 2026, 2027 and beyond, whereas this would come live much faster given the involvement of Blackstone.
5:56Caroline Hyde:Mandeep Singh, who leads technology research at Bloomberg Intelligence, thank you very much. Let's take a look at today's big number. 7 ,000, that's how many workers MEDA's resigning to new jobs related to AI. That's according to an internal memo. This comes on the eve of a 10 % staff cut tomorrow to improve efficiency and, quote, offset MEDA's investments in AI. AI-related job cuts are the talk of Wall Street right now, with Standard Chartered CEO Bill Winters also saying his company plans to cut more than 15 % of its support staff by 2030, aiming to replace, quote, lower value human capital with AI.
6:33Caroline Hyde:Marta Norton is Empowered Chief Investment Strategist, and I'm going to ask you straight up. It's an interesting balancing act, cutting roles and saying that this is not necessarily a trim, a downsize. It's a reallocation of priority to AI. As an investor, what signal does that give you? Well, it's not the only piece of data that we have as companies do that. We hear a lot of different announcements of companies making reductions to their workforce related to AI. I think a lot of these are idiosyncratic when we're looking at the broad economy at the moment. We know that the broad economy added labor upon labor in the wake of the pandemic, and now there's a bit of right-sizing.
7:18And so in some cases, I do think companies are adding the AI moniker to make it feel very forward-looking, to maybe AI whitewash it, but not necessarily a sign of the times in terms of very broad layoffs across the economy. In fact, one of the areas that I'm looking more closely for signs of an AI impact is more on the hiring side, where people are maybe slower to hire than they otherwise would because of AI. But yes, there are these idiosyncratic reports that suggest we are seeing a little bit of movement on that front. I'm sorry, language matters, lower value human capital. It's an extraordinary choice of words.
7:55But Malta, putting that to one side, there is anxiety out there of labor and people who are affected. But when you're thinking about how we see actual businesses lean into this moment, are we still seeing adoption taking root in the way you anticipated? Are we seeing the people, the labor that is able to use AI being more productive with it? You know, I think there's early signs that are pretty positive in terms of adoption by current employees. So I think everyone at this point is using the chatbots on a regular basis to do, you know, low value research. But I think the bigger question is this AI agent deployment, what that looks like.
8:38And you are seeing companies talking about that in real time and beginning to structure workflows around it. I think we're very early, obviously, on the adoption of that. But that is where I think a lot of this AI productivity can stem from.
8:53Caroline Hyde:Marta, the market's trading in part in anticipation of NVIDIA after the closing bell Wednesday. Yesterday, I spoke to NVIDIA CEO Jensen Wang. This is how he frames the state of play. We have the largest supply chain in the world. Our partners have done a great job securing supply for us. And so all of the pieces go together. The silicon photonics is lined up. Everything is all lined up. It's just that the demand is much greater than the overall capacity of the world. The demand is far outpacing this industry's ability to supply. When you go into an earnings, how do you decide, therefore, the bar in terms of what is positive and what is negative if you're in a supply constraint world?
9:39Well, I think you do want to see signs that they are monetizing it, right? If you're in a supply-constrained world, that demand should be moving earnings, and you want to see some evidence of that. But, of course, when we're talking about the supply constraint, we're also talking about supply constraint in areas that would infect the likes of NVIDIA. So memory chips, for example, and what does that mean for gross margins, and how are they navigating that? I think those are important questions at the moment. But I think the bigger picture, and I caught that interview yesterday, I think the bigger picture that Jensen Wong is pointing to is this massive supply and demand mismatch that will carry through not just this earnings report, but for several quarters, years to come as this all plays out.
10:21So I really agree with his comment that this is inning one, inning two on the AI build. Luckily, we all caught that interview. It was a great one with Ed. And Martin Norton, we so appreciate having you on the show today of Empower. Thank you for your time. Coming up, a bitter feud. Personal grievances, billions of dollars. The decision comes down to timing. We'll discuss a jury's rejection of Elon Musk's claims against his OpenAI co-founders. And what comes next? This is Bloomberg Tech.
10:58The decision in that closely watched legal battle between Elon Musk and his OpenAI co-founders came down to timing. A jury ruled that Musk waited too long to sue over OpenAI's transition to a for-profit business. Bloomberg's Madeleine Mecklenburg has been following the case every step of the way. You join us now. Madeleine, what did the jury decide? What didn't it decide on? Right, it was a bit of an anticlimactic verdict, if I can say that, because the jury didn't actually make a decision on any of the claims that Elon Musk presented over the course of this lawsuit. What they said instead was that he had not proven his case within the statute of limitations set by law for the specific claims that he brought.
11:39So essentially, they found that he had concerns about this conduct and should have known about this conduct years before he actually filed his lawsuit. And so therefore, the case that the jury was hearing should be tossed out, which is what happened.
11:54Caroline Hyde:We're going to get an appeal, according to Mr. Musk, who posted on X that he would appeal. I believe his legal team outside the courtroom said that we would appeal. If that's the case, what do we know about what that appeal would be based on from a legal perspective? Right. It should come as no surprise to anybody who follows Elon Musk as a litigant that he plans to appeal this and fight this vigorously. We saw him celebrating on X afterwards saying that, you know, just because they decided on the statute of limitations, they never actually touched the merits of the case. I think an appeal on this grounds is a lot more narrow than if the jury had actually ruled on the merits of his claims.
12:35So I think a lot still kind of remains to be seen of how this is going to play out. But it's going to be an issue that the Ninth Circuit Court of Appeals is going to have to sort through when he eventually files that.
12:46Caroline Hyde:Bloomberg's Madeleine Mecklenburg, thank you very much. Let's dive more into the legal considerations. Dorothy Lund is with us, a professor at Columbia Law School focused on corporate law, governance and contracts returning to the show and has been covering this trial with us. In our Bloomberg News story, we write that Mr. Musk was too late in his suit. And we later down in the story explain the idea that he had knowledge in 2024 and so basically should have filed his suit sooner. But I've got a lot of sympathy with people around the world that don't understand the legal technicality of that. The statute of limitations seems to play a role here.
13:23Caroline Hyde:Is this codified or is this the jury's discretion in deciding that Mr. Musk ran out of time? Yeah, so the jury had to basically answer this question of whether or not the statute of limitations applied to the two claims that were brought here. So the two claims, just to remind everybody, were that there was a violation, that OpenAI, Brockman and Altman, violated a promise that was made to Musk that it would have a permanent charitable mission to develop safe open source AI technology and that they received undeserved benefits because of those broken promises. And so then OpenAI said, well, you waited too long to sue.
14:07you know you brought your case on august 2024 and the statute of limitations was only three was three years which meant you the things that you're complaining about that took place in 2018 2019 you you you you're too late whereas surely did elon musk and his his team of very expensive
14:29Caroline Hyde:lawyers not understand the law or something when they filed the suit no musk you know so musk of of course, had these fabulous lawyers. And, you know, the statute of limitations is told if you've concealed, if harm's been concealed, or you haven't discovered the harm. And so his argument was, well, you know, Sam Altman and OpenAI have been lying to me. They concealed that they were doing all this stuff. It wasn't until 2023 that I really understood what was going on. And so this, and this gets back to your original question, which is, why did this go to a jury, right? The jury had to sift through testimony from both parties, documents, to try to understand this timing question.
15:12So when was it that Musk knew or reasonably should have known about the harm that he was claiming happened? And instead, we all get consumed with billionaire versus billionaire and how much equity they own in the underlying company and what they're like in terms of managers or how trustworthy they are. At the end of the day, were we misguided to think that? When they go back to an appeals court, will it be argued in a different way? Yeah, so it's a little disappointing that we didn't get an answer on the big question here, right? The claims that I was telling you about just now are unaddressed.
15:53And in this appeal, we won't get any resolution on this either. So this is going to be a narrow appeal on this decision of the statute of limitations. And because that statute of limitations question turned on a factual determination, that is going to be really hard to be overturned. Appellate courts rarely overturn a jury's fact-intensive findings, especially in this case where the judge also reviewed the evidence and said, I came to the same conclusion. And so the only path forward for Musk is going to be arguing a different set of things about, you know, that argument would have to be some error that come here, improper instructions or improper evidentiary rulings.
16:41Caroline Hyde:We're having some technical issues, Dorothy, with your Zoom. So I'll give the Internet gremlins a second to reset. But my question to you is exactly that. I don't want you to speculate, but what options does Mr. Musk have if he needs to go to the appeals court but argue something completely different to the original case? Yeah, his argument is going to have to be that there were significant legal or procedural errors that prejudiced the outcome. You know, that the instructions to the jury were improper, their evidentiary rulings were wrongly decided. There won't be any discussion on the substance of this larger question of whether or not OpenAI violated its charitable purpose when it restructured to create a for-profit affiliate.
17:30Called a public opinion, therefore, in many ways. Dorothy Lund, professor at Columbia Law School. Fantastic to have you. Thank you. Coming up, Parallel is moving past the era of flat-fee AI licensing deals. The CEO, Parag Agrawal, is going to be joining us next. This is Bloomberg Tech.
17:54The right technology can strengthen human judgment. That's why Deloitte brings together AI and data analytics with multidisciplinary teams. People with deep industry experience who can challenge assumptions and help you connect the dots across your enterprise. From risk signals to operational pressure points to shifting customer needs, Deloitte helps you see what's coming sooner. So opportunities don't slip by and surprises don't spread. It's not just dashboards. It's real clarity in the moments your decisions are made. When models reveal patterns, people can ask better questions. When data and people are connected, leaders can move faster with confidence.
18:34And when your teams are aligned, smart choices can scale from the front line to the C-suite. Because the smarter your systems, the sharper your instincts. That's how technology makes people better at what they do best. Deloitte. Together makes progress. Learn more at Deloitte.com slash together makes progress. 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.
19:13So 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. chances 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.
19:54SEC Registered Advisor. Crypto services by zero hash. Sample prompts are for illustrative purposes only, not investment advice. All investing involves risk of loss. See complete disclosures at public.com slash disclosures. You need to make a huge presentation in an hour. Adobe Acrobat uses AI to take all your documents and generate a presentation with a single click. Build slides quickly and streamline the process. Need a last-minute pitch deck? Do that with Acrobat. Need to level up your presentation design? Do that with Acrobat. You have 30 plus documents that need to be simplified into a proposal.
20:31Do that. Do that. Do that with Acrobat. Learn more at adobe.com slash do that with Acrobat.
20:45Caroline Hyde:Parallel, a web API purpose-built for AI and led by the former CEO of Twitter, is launching Index today, a new marketplace that pays publishers and data providers based on how much their content actually helps an AI agent complete a task. Joining us is Parallel founder and CEO, Parag Agrawal. So in answer to the question, what did you get done this week? You launched Index. In this era of agentic, the agents need the web data and the proprietary data sets, you, I guess, saw a gap where the marketplace for that data doesn't exist, then now you think it does. Yeah, we've been building towards this content marketplace now for the last two and a half years.
21:31We started parallel with the notion that agents will use all of the content on the web, thousand X more than humans ever have. And when that happens, the technology that we've built over the last couple of decades isn't enough. But the business models also break down, ads, subscriptions. These business models no longer work. And that, I think, is the opportunity for us to figure out how to align the interests of AI agents doing real work, along with the content owners who are creating valuable content that feeds these agents. And that's what Index is all about.
22:07Caroline Hyde:To help the Bloomberg Tech audience out, the question, what did you get done this week, was posed to you by Mr. Musk just before you left Twitter. how do you assign a value then to the data? You've called this index. Very interesting, right? On the one side, you have the agent or the entity the agent is acting on behalf of. And then on the other side, you have the source of the data who sets a price or a value. Yeah. So if you think about other deals that are being done, these deals are typically between a large lab or a hyperscaler or Google alongside a large publishing house. And these are flat fee deals where over time, I don't believe content owners get to participate in the growing economy of agents.
22:52Now, what's really happening with agents is they are starting to do real work, real knowledge work in the enterprise. So if you think about our customers, our customers span people building AI scientists, our customers are building AI lawyers, customers are building for finance using agents. And all of these agents that our customers build are doing real work. When they produce value, content owners don't get to participate in that economy of the value being created. With Index, what we're trying to do is build a system where if very valuable work is done using somebody's data, they get more money.
Read the full transcript
23:34If someone has very high quality data, they get paid more. And we've built a model to figure out what this value is, what this marginal contribution is. And it draws apart from this game-theoretic concept called Shapley Values. Let's talk about the Shapley value. It's contribution to the work performed by an AI agent. So basically, instead of content access crawls or citations, you consider the value. I mean, but where's the monetary number coming from? What are you pegging it against? And how? How do you work out what the NADI agent has eventually learned itself off of? Well, Shapley Values has been my obsession for the last three years as we've been working on this.
24:15So just in a sentence, what Shapley Values is, is about like if a few people decide to collaborate on building something in a positive some way where the whole is bigger than the sum of parts. The question is, how do you divide up this value? And Shapley value is the concept that tells us the way to divide this value so that everyone is incentivized to participate. The net outcome of using this concept to reward content owners, it creates three really nice properties. One, high quality content gets paid more. Number two, content that gets used in high value work gets paid more. And most importantly, as agents do more work and create more value, content owners grow alongside them, which we believe allows content creation on the web to be sustainable.
25:06There are a lot of content creators out there right now. So how do they sign up to Index? How do you make sure that it's not just quite large either individuals who are writing, I think of Alex Heath, or it's actually those that have got a minor substack, but with really valuable content that somehow the agent has managed to access? How do they access that? Yeah, with Index today, we're announcing a collection of launch partners. These span premium news publishers like The Atlantic and Fortune to really draw factual amazing content from the PR newswire, business intelligence and data providers like PitchBook and Traction and ZoomInfo, as well as independent creators like Alex Heath, Azimazar, and several others.
25:51The really interesting part about this is that we want everyone, all content owners, big and small, to be able to participate in this economy. And we want all agents, not just ones built by large providers, to be able to get access to all of this amazing content.
26:08Caroline Hyde:We just have 60 seconds, but can I ask you about X, the platform for me known as Twitter? Of course. You're still using it, your view of it, and as part of XAI, as part of SpaceX. I'm still using it. I'm still calling it Twitter. I think we built a product over a decade that I was there, which endures in value forever. And I'm really proud of the platform that we created. It allows the theme of content access and democratization out in the open versus in permission groups. And that theme is what carries forward into my current company parallel, where the goal is to have as much content as possible out in the open for everyone to have access to.
26:53everyone's agents to have access to instead of just a few. And be compensated. Parallel CEO, Raghagawal. It's great to have some time with you. Thank you for coming on. Now, coming up, we dive into Meta's mega data center plans in Louisiana and some of the hiccups along the way. From New York, from San Francisco, this is Bloomberg Tech.
27:31Caroline Hyde:Welcome back to Bloomberg Tech. There's still a semiconductor story in financial markets. And looking at the stocks of the Philadelphia Semiconductor Index, we're down for a third straight day. A part of this is kind of a breather after what we were calling the melt-up. The performance in chip stocks since the end of March was astonishing. But there's also this element that we are really close to correction territory. In three days, at one point in the session, we're down almost 10%, as Cara pointed out earlier. NVIDIA is a part of it, right? And the idea that post-President Trump's visit to China, we don't really have a net conclusion of what's going on, but we get NVIDIA earnings on Wednesday.
28:09Caroline Hyde:So fresh off his trip to Beijing, NVIDIA CEO Jensen Wang says he's confident the Chinese market will eventually open back up to his company. Dell CEO Michael Dell shared the optimism for economic collaboration as well. Here's what both leaders told me at Dell World in Las Vegas yesterday. The president wants America to win everywhere. Right. The president wants America to lead the AI revolution. And so H200s are licensed to sell to China. The Chinese government has to decide how much of their local market do they want to protect and how much of their local market do they want to expand with more AI capacity?
28:49My sense is that the demand in China is so incredible. Just like it is here, agentic AI is also making enormous progress there. My sense is that over time, the market will open. President Xi was very clear that he wants China to be an even wider open market. Premier Li Chang was very straightforward and to explain very eloquently that China will be in an open market. So I'm looking forward to China being a more open market.
29:20Caroline Hyde:To clarify, Jesse, you were able to meet with those officials directly to discuss whether or not you can sell to those Chinese tech companies. I didn't discuss directly with them about H200. I was there to represent the United States, and I was honored to do so. I was there to support President Trump, and really glad to do so. But that was really the focus of my trip. President Trump had some conversations with the leaders. And I'm looking forward to what they decide. Michael, you did not go to China. But I think what's interesting is you are a member of the President's Council of Advisors for Science and Technology, as is Jensen.
29:57Caroline Hyde:Your net conclusion on whether or not China will become open to American technology companies to do business there? You know, we have a business in China. Obviously, we comply with all the restrictions and various controls that are in place. But I hope that there's more economic collaboration between the United States and China. That ultimately is what will lead to greater outcomes of prosperity for everyone and a greater likelihood of a successful relationship between the countries and around the world. Dell CEO, Michael Dell, of course, NVIDIA CEO, Jensen Wang, speaking there with Ed yesterday.
30:41And let's just revisit today's big number. With Meta reassigning 7 ,000 workers to new jobs. Of course, they're related to AI. An internal memo from Chief People Officer Janelle Gale, viewed by Bloomberg, reveals employees will move into one of several new groups focused on AI-related products, including agents and apps. The new corporate structure will, quote, be flatter and have smaller teams. Now, this comes as the company is getting ready, of course, to cut 10 % of its staff this week as part of an effort to, as they say, improve efficiency.
31:11Caroline Hyde:Yeah, sticking with Meta's big AI plans, Mark Zuckerberg's plan to build the world's biggest AI facility has enmeshed his company deeply in Louisiana's politics and economy. The project's providing an economic boost to one of the nation's poorest regions, but also bringing a host of growing pains. Thanks. Bloomberg's Riley Griffin joins us more with the big take. And there's big readership on this, a lot of interest. You did what every good beat reporter does. You go there. You see what's happening on the ground. The headline is$200 billion, right? That's the commitment. What is happening with Meta in Louisiana?
31:49Yeah, a really important point here is that$200 billion spend that Meta is thinking about with Richland Parish and its data center there is largely about the chips. This is going to be a five gigawatt of compute capacity data center, another 2.5 gigawatts to support the broader campus, which is spanning nearly 4 ,000 acres. President Donald Trump has said that this is going to be a Manhattan-sized project after speaking with Mark Zuckerberg. But a lot of that spend isn't going directly into the community. It's going into the chips inside the facilities. And we still have to ask the question, what does this mean for the people of Richland Parish in the Louisiana Delta, a region that has struggled for decades and where farmers are really losing$400 on every acre of cotton they sow right now.
32:34Well, let's go to Dustin Morris, because you paint a picture of how his economic reality is currently. He's a soybean, a corn farmer, and he's flying around his own land trying to work out whether it's the right time to be selling out, right? Yeah, it's a story of expected land rush, right? As with all of the data center stories of this size around the United States, many of them shifting into rural regions, particularly in the South. And people are looking around and wondering, you know, do I trade this history? Do I trade this land? Do I trade this crop for a data center? And it was, I mean, it was really an honor to be able to go multiple times to Richland Parish and hear their concerns and think about the way they're looking at Meta as a lifeline.
33:20And I, over years... Let me jump in.
33:22Caroline Hyde:What are Meta saying about that? Like, what's Meta's response to it? And what do people genuinely feel about the company? Are you even aware it's Meta building there? Well, it took some time. Meta hashed out this deal over the course of a year in private. Meta is certainly saying they want to be there in the thick of it with the community. They're investing in the schools. There's actually a school at the corner of the data center. There are discussions about whether it should be moved. I have heard that Meta is reviewing plans. So they're in the trenches. But notably, Meta is only bringing 500 jobs.
33:54This is a community that's looked for 5 ,000 over the years from an auto manufacturer and has been unable to court that. So will 500 jobs be enough to provide that economic lifeline long term? I'm not sure.
34:06Caroline Hyde:That's Riley Griffin with the big take, top of the website and the Bloomberg Terminal. Caro, there's plenty more news out there today. There is. It's time now for Talking Tech Ed. And first up, Chinese AI pioneer Moonshot is restructuring to secure its path to public markets. Now, the developer of the popular Kimi chatbot has informed investors it will revamp its corporate structure to comply with Beijing's rules and pave the way for a highly anticipated Hong Kong IPO. Speaking of which, Chinese robotics unicorn Linkobot is exploring a Hong Kong IPO that could launch as early as this year. and the maker of highly dexterous robotic hands is targeting a$6 billion valuation and counts Samsung, Tencent, among its clients.
34:44And China's demographic time clock is triggering an unprecedented automation boom. A new report from Barclays shows humanoid robots could offset as much as 60 % of that labor decline by 2035, turning what is a structural crisis into a hardware gold rush. Ed.
35:03Caroline Hyde:Coming up, we're going to go live from the JP Morgan Tech Media and Communications Conference, and we'll hear from Laurie Beer, JP Morgan, Global, CIO. That conversation next. This is Bloomberg Tech.
35:31One 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. Chances 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.
36:10That's public.com slash market.
36:31Sending a file is easy. Making sure your clients understand the file is the hard part. But with PDF spaces in Adobe Acrobat, you can give your clients the full picture with custom intros, audio summaries, and a helpful AI assistant to your docs. So if you want to stop the endless follow-ups, do that with Acrobat. Need to make your docs crystal clear? Do that with Acrobat. Want to make sure your clients get everything they need to hear? Do that with Acrobat. Learn more at adobe.com slash do that with Acrobat. Game night rush or any night of the week, really. Genius keeps every order moving. From online ordering to your kitchen to the front counter.
37:13Big league reliability for any business. That's genius.
37:25Caroline Hyde:Let's get out to the J.P. Morgan Tech Media and Communications Conference in Boston, where Bloomberg Surveillance co-host Lisa Abramovitz is standing by. Lisa. Thank you so much, Ed. I am here at the J.P. Morgan Conference for Technology, Media and Communications. And I'm here with Laurie Beer, the Chief Information Officer for J.P. Morgan. And Lori, wonderful to speak with you. In many ways, you are the woman of the moment because you have to understand how to oversee change that seems to be happening at an accelerating pace. Can you give a sense of just how quickly it has accelerated from, say, six months ago, a year ago to right now?
38:02Yeah, I think it's an amazing moment, first of all, to be in technology and to be in financial services. And we definitely saw the Gen 1 tools come out a couple of years ago. the last six months have rapidly accelerated. Not only new models getting created, but how we're able to apply them. And so it really ties back to how do we think about driving and adapting change so quickly? Yeah. One thing, it's created a lot of nervousness, a lot of anxiety. There was a story that was out this morning about Standard Charter CEO coming out and saying he was going to cut 50 % of staff, support staff, and removing lower value human capital and replacing it with AI tools and investment in AI.
38:46What's real, what's not when it comes to sort of the mass job cuts? What's been your experience? Yeah. So overall, we've seen definitely productivity opportunities. So even when we think about technology, we've seen 10 to 20 to 30 % improvement just from the initial generation one tools. As you look to a Gentic, you're going to see even more, but there's a huge demand. There's a huge demand to create new products and services and experiences for our customers. We're using AI to figure out new products to offer. And frankly, when you think about the other side of it, the AI is also creating increased cybersecurity risks.
39:21So how do we think about also accelerating the investments to protect our customers and clients too? So the demand remains significant and we're able to get a lot more done than we have historically. How scary is mythos. So with mythos, AI in general, but it's really good on both sides of the equation. One, to understand vulnerabilities. And so the key with that and any other AI is how do we capture those? How do we clearinghouse those? How do we update those in a safe and secure way? But on the other side of this, as you apply these models into the software development process, we're going to have safer and more secure software.
40:06On the flip side, our cyber teams can use the power of these models to be able to protect, defend, identify fraud, identify issues more quickly. So there's really two sides of this equation to think about. Based on how quickly you see technology shifting and what tools are really coming to the fore, how does the mindset have to shift of the people who work for you? Yeah, I think it's a great question. The technology is amazing. I think this is truly a shift, a transformational shift in technology where this is getting embedded in horizontally across all layers of the stack. The harder part is the change management and the leadership.
40:44And I think really thinking about where are things going, we don't know all the answers. So how do you navigate change in an uncertainty? It's still very early innings. So it's really had to shift from how do we help guide our teams, How do we make sure our software engineers understand what's coming? My management team spends a lot of time on the weekends building their own apps. Pick your favorite coding, authentic coding tool. But it helps them understand what's coming. And I think this is a moment where leaders, in addition to software engineers, truly have to understand the change. And we have to understand, you know, how do we actually create an environment that allows us to pivot and navigate when times are uncertain and really make sure we don't lose track of what's really important.
41:31How do we deliver products and services and experience for our customers and clients? How do you balance innovation with risk? And the idea of the tried and true, you need to do this project versus moonshot projects. Yeah, I think this is one of the most interesting times when you have to think about the balance of innovation and risk-taking. We wouldn't be in business as long as we have been if we haven't driven a lot of innovation historically. but we're also in the business of managing risk. And so I think in this moment, you have to really think about going at the appropriate speed of innovation.
42:04Some of these moonshots are becoming real. And so you have to move aggressively at the innovation front, but we obviously are a business of trust and we have to make sure our customers, our clients, the bank is protected every day. And so finding this balance and actually, again, using these capabilities to also protect ourselves. And so it's this really interesting time. While along with that, you don't necessarily know who's going to win in the end. And so making sure we're not locking ourself in to one provider or one scenario. And we're constantly testing, sandboxing what's on the horizon in the future.
42:44Does it change your decision whether to build or buy in terms of different software solutions? You know, it's been interesting because over the last several years, we've made sure that we're really focusing our engineering community. And we have a large engineering community on building the things that are competitively differentiating and create value for our customers and clients. And over time, you have to balance where that still makes a lot of sense and where we focus our energy on those unique products and services for our customers and clients. And so when you think about the balance of trade and the fact that software engineering overall is increasing in speed and it is more cost effective when you look at the end to end product development lifecycle, you know, there are different tradeoffs you would consider in making those build versus buy decisions.
43:33but you're always going to have some of those enterprise software solutions, frankly, that aren't competitively differentiated, that sort of help us run the bank without delivering those unique products and services. Just 20 seconds. Do you think that we're going to be able to recognize what banking looks like, say, in 10 years? I think we're watching it transform as we speak right now. At the end of the day, we have to maintain that trust with customers and clients, and we have to make sure that their unique needs are served across the over 100 countries we operate in. But how we do it and the products and services will absolutely evolve over the next several years.
44:09Caroline, that is Laurie Beer, the Chief Information Officer at GPMorgan. I'll send it back to you. What a great day. What a great set of interviews. Thank you, Bloomberg Surveillance co-host Lisa Abramowitz there. And we've got some breaking news in the meantime.
44:23Caroline Hyde:Yeah, some breaking news on the terminal about Apple. Apple Chief Hardware Officer Johnny Shrugy is reorganizing hardware development and shifting oversight of key functions such as product design, part of an effort to speed up work on future devices. Sources say the hardware shakeup is also meant to better integrate teams working on in-house silicon with those that are creating products. And, of course, Bloomberg's Mark Gurman, the one carrier that broke the story. Some other details, Shelley Goldberg, Dave Pakula are now going to oversee product design. maybe not names as familiar to the Bloomberg Tech audience, but in this new era for Apple, a lot of focus on who's right at the top and running the show.
45:03They'll become known. Ed, thanks. Meanwhile, coming up, look, we're going to be speaking with Armada CEO Dan Wright as the company raises$230 million in a Series B funding, boosting its valuation to$2 billion and some new strategic investors and big projects for the hyperscaler for the edge. This is Bloomberg Tech.
45:26Caroline Hyde:Some more breaking news crossing the Bloomberg terminal. OpenAI founding member Andrei Karpathy says he's joined rival Anthropic. In a post on X, Karpathy says he will be working on R &D with Anthropic and be focused on helping train new AI models. Karpathy previously helped lead the team behind Tesla's autopilot system before then returning to OpenAI. He left the ChatGP maker for a second time last year and had launched a startup focused on AI and education. In that post, Carrey, he says that he's still passionate about education, plans to resume his work on that. But in the market of AI talent, this is a bit of a wow.
46:02Caroline Hyde:This is a big story that I didn't see coming breaking this morning. It certainly is. The wows keep coming. Meanwhile, let's talk about a fundraise now. Armada has just raised$230 million in an oversubscribed Series B round, boosting the company's valuation to a cool$2 billion. The company is also announcing a partnership with Johnson Controls to produce modular data centers. And they're at scale. Joining us now, Armada CEO Dan Wright. What's a modular data center? How are we going to be making it? Yeah, thanks so much for having me on. A modular data center is a full stack AI factory that you can deploy anywhere.
46:35And we can do it with the three S's. We call it speed, scale, and sovereignty. You can deploy these within weeks. You can scale up with demand versus traditional data centers. Sometimes you end up overbuilding or building the wrong thing. And then modular data centers also have the benefit that you can get sovereignty down to the site level. People used to talk about data sovereignty in terms of countries, but with increasingly sophisticated cyber attacks, now you want data sovereignty at the site level.
46:59Caroline Hyde:This is all about speed, movement, building. Funding, this helps you move even quicker, or what's the plan here? Yeah, so this round was co-led by Overmatch, 8090 Industries, as well as BlackRock, and really what it allows us to do is scale up to accelerate deployment of the USAI stack. As part of this, we're also announcing a global agreement with Johnson Controls, where starting with Galleon Forge 1 in Arizona, we're going to do continuous manufacturing of these modular data centers. And if you look at them, they kind of look like shipping containers. So they're easily transportable anywhere in the world.
47:36You can deploy them wherever you want. And the nice thing is you can use any source of power. And that's obviously important. Power is a big concern when it comes to data centers. Energy is distributed. We're saying compute should also be distributed. Where's the demand coming from exactly? Who wants these? Yeah, so energy companies, we're working with many of the largest energy companies in the world. All of them have a hard segregation between IT and OT, meaning that the only way to do anything with AI is to deploy the infrastructure on the rigs, on the refineries, and run those models behind the firewall at the edge.
48:09We're enabling that. They're all trying to be fully autonomous over the next few years. You need compute locally in order to enable that. And then also defense is a huge driver for us. For example, with the conflict in the Middle East, we got a call from an ally. They said, hey, we need one of these. Like immediately we were able to deploy within days. They called back, said we need more. We were able to ship more out. And who are your partners within this data center? I can understand that you're helping with the construction of it all, but still you need the chips from someone. You need the servers from someone.
48:36How are you bringing this all together? And how open are you to different types of partners? Well, I think that's actually part of the magic of Armada is that we built this ecosystem of amazing partners. And it's kind of like actually what the hyperscalers did 20 years ago, is you're bringing all of these different services and you're making them accessible to more people. That's what Armada is doing for the Edge is the hyperscaler for the Edge. And a good example of this, we just announced a partnership recently with Microsoft to bring Foundry, bring Azure Local, allow them to run, these customers to run services that they love from Microsoft, models that they run from Microsoft, at the Edge AirGap.
49:13Same thing, we announced a partnership with OpenAI. We announced one with NVIDIA, Dell, as well as Palantir,
49:19Caroline Hyde:bringing all those capabilities to the edge. So, Dan, real quick, I was with Michael Dell and Jensen Wong yesterday. The bigger theme is on-prem. People want, at the edge, their facility close to them. Is that an addressable market for you? We just have 30 seconds. Yeah, absolutely. I mean, that's what I believe is going to happen, is you're going to have sovereign AI factories. And really what that means is that every country and every company is going to have its own AI factory. and I know that Michael and Jensen would agree with me on that because they've been saying that publicly. And in order to do that, you need both the chips, you need the servers, and then you also need the infrastructure and that's what Armada's enabling as the hyperscaler for the edge.
49:57Caroline Hyde:Dan Wright, Armada CEO, back on Bloomberg Tech after a big fundraise. Great to have you back. Thank you very much. That does it for this edition of Bloomberg Tech, Caro. Yeah, you don't want to forget to check out this podcast. What a rich podcast we're going to be having. all eyes as we build up towards NVIDIA's earnings after the bell tomorrow. You can find it on the terminal as well as online on Apple, Spotify and iHeart. Ed, still a phenomenal interview from yesterday as well. Yeah. And, you know, the market is waiting. Anticipation for the NVIDIA print Wednesday night, post melt up in chip stocks.
50:30Caroline Hyde:Wow. This is Bloomberg Tech.
50:37If you follow markets, you know the value of long-term thinking. You plan, you diversify, you prepare for volatility. But even the best strategies can't prevent every bad day. For more than 75 years, Cincinnati Insurance has helped individuals and businesses navigate tough moments with expertise, personal attention, and independent agents who focus on relationships. Not transactions. The Cincinnati Insurance Companies. Let them make your bad day better. Find an agent at c-i-n-f-i-n dot com. So as a pizza genius, I know pizza shop orders come from, well, everywhere. With Genius by Global Payments, online orders actually sink straight into your kitchen.
51:23It's as simple as pie. And with digital menu boards, your specials, your prices, your brand, always front and center. It's one system, ready for game night crowds. Any night of the week, really. Big league reliability for any business. That's genius. A business gift should do more than check a box. It should reflect your brand and show someone they're appreciated, recognized, and truly seen. 4imprint offers thousands of high-quality, customizable products, like premium apparel, drinkware, tech, and more, making it easy to create a gift that feels meaningful and on-brand. And with 4imprint's expert support and their 360-degree guarantee, you can be 4imprint certain your order will arrive exactly as intended.
52:09Explore gifting with confidence at 4imprint.com. 4imprint. 4certain.
From the publisher
Bloomberg’s Caroline Hyde and Ed Ludlow discuss Google's decision to create an AI cloud business with Blackstone which will run on Google's homegrown AI chips. Plus, the legal battle between Elon Musk and his OpenAI co-founders ends with a jury ruling that Musk waited too long to sue. And, Meta bets big on Louisiana for the world's biggest AI facility, as it plans AI-related layoffs this week.
See omnystudio.com/listener for privacy information.

