In short
Bloomberg Tech episode covering major AI/tech business moves and markets: NVIDIA in talks to acquire Hugging Face for about $14B; Dell’s AI-server surge and raised outlook; Uber cutting 10% of staff; Cognition raising ~$1B at ~$47B valuation; plus segments on Napster’s AI-agent pivot, humanoid robotics scale-up forecasts, and earnings previews (Snowflake, HPE, Broadcom).
Guests (and backgrounds)
- Ian King and Rachel Metz (Bloomberg reporters who broke the Hugging Face story).
- Michelle Canale (President/Owner, Porsche Capital Management; covers servers/semis).
- Natalie Lung (Bloomberg reporter on Uber restructuring).
- Rebecca Torrance (Bloomberg reporter on Cognition funding).
- Maggie Eastland (Bloomberg reporter at G20 tech/data-center discussions).
- John Acunto (Napster CEO; agentic AI “streaming intelligence”).
- Zanitsa Todorov (Head of Thematic Equity Research, Barclays; humanoid robotics).
- Brody Ford (Bloomberg Intelligence analyst previewing Snowflake).
- Wu Jin-ho (Bloomberg Intelligence analyst previewing HPE).
- Kunjun Sabani (Bloomberg Intelligence analyst previewing Broadcom).
Key claims + notable examples
- Hugging Face deal: $14B includes ~$1B retention to keep talent; Hugging Face described as a platform for publishing/sharing AI models and open research; NVIDIA wants broader AI adoption beyond hyperscalers.
- Dell: AI server revenue expected to triple to $74B; $95B AI server backlog; margins rising (operating margin >11%).
- Uber: CEO memo cites micro-teams; cut ~3,300 roles (~10%) to reduce coordination; reallocating capital to autonomy and core delivery/ride-hailing.
- Cognition: annualized revenue passed $900M (from $492M in May); valuation jumps from $26B (May) to ~$47B; SpaceX previously approached but was rejected.
- Napster: streaming “music” shifted to agentic AI on Microsoft Azure; claims cost ~penny/min vs others; agent enables one-on-one conversations online/in-store.
- Humanoid robotics: Barclays forecasts $200B market by 2035; 60,000 deployments in 2026; biggest bottleneck is lack of real-world physical training data (synthetic/simulation may hit limits).
- Earnings: Snowflake seen as “data layer” benefiting from AI processing; HPE read-through from Dell’s AI-server momentum; Broadcom expected to benefit from custom silicon/ASIC volume and has ~40% revenue from software.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Chapters
Tap a time to open that second in VONvidia's Potential $14B Acquisition of Hugging Face
1:42 to 2:04
Discussion on Nvidia's acquisition talks for AI startup Hugging Face.
“Plus, Dell shares jump after the company boosted its annual sales forecast by$25 billion due to surging demand for servers to run AI tasks.”
Understanding Hugging Face's Role in AI
2:04 to 2:46
Exploring what Hugging Face does and its significance in the AI ecosystem.
“NVIDIA could be closing in on its biggest AI acquisition yet.”
Nvidia's Motivation Behind the Acquisition
2:46 to 3:41
Analyzing why Nvidia is interested in acquiring Hugging Face.
“Rachel, it seems a strange place to go next, but what is Hugging Face?”
M&A History of Nvidia and Its Implications
3:41 to 4:47
Discussion on Nvidia's acquisition history and what it means for Hugging Face.
“At the moment, the industry is very concentrated on the hyperscalers and what they produce and what open AI does.”
Dell's Earnings and Market Performance Analysis
4:47 to 5:59
Insight into Dell's financial performance and its implications for the market.
“So it could be a really great matchup for both sides.”
Market Trends in Data Center Spending
5:59 to 8:07
Discussion on projected data center spending and its implications for Dell.
“Joining us now is Porsche Capital Management President and Owner Michelle Canale.”
The Semiconductor Industry's Future Landscape
8:07 to 10:31
Exploring the future of the semiconductor industry and potential market shifts.
“I want to zoom out here for some perspective and look at today's big number, which is$32 trillion.”
Alphabet's Advertising Exchange Ruling
10:31 to 12:29
Overview of a federal judge's ruling on Google's advertising practices.
“I'm trying to think of one etched as a good example.”
Uber's Global Workforce Restructuring
12:29 to 13:20
Details on Uber's plans to cut 10% of its global workforce.
“This is Shares, very muted move after a federal judge ruled that Google doesn't have to sell off its advertising exchange and instead needs to make its ad tech tools work with those operated by rivals.”
Uber's Restructuring Efforts
15:05 to 15:27
Discussion on Uber's decision to cut jobs and restructure management.
“At LPL Financial, we offer more ways for advisors and their clients to thrive.”
Show all 25 chapters
Uber's Focus and Future Plans
15:27 to 17:44
Analysis of Uber's business strategy and future investments.
“That's according to an email from Uber's CEO viewed by Bloomberg News.”
Cognition's Funding Round
17:44 to 20:05
Insights into Cognition's massive funding round and valuation increase.
“AI coding startup Cognition is set to close a new funding round of around$1 billion, according to sources.”
Investor Interest in AI Startups
20:05 to 20:22
Discussion on the surge of interest in AI coding startups like Cognition.
“Okay, to recap, Cognition is set to raise a billion dollars at about a$47 billion valuation, according to Rebecca's reporting.”
G20 Meeting and AI Discussions
20:22 to 22:01
Overview of tech leaders' discussions on AI at the G20 meeting.
“Let's get out to New York, where Bloomberg's Yaira Hanand is standing by.”
AI's Role in Economic Growth
22:01 to 25:17
Exploration of AI's potential impact on global economies and growth.
“NVIDIA CEO Jensen Wang is urging the world's largest economies to accelerate their adoption of AI as a way to enhance growth.”
Napster's Transition to AI
25:17 to 28:00
Interview with Napster CEO about its pivot from music to AI services.
“You probably remember Napster as that company that upended the music industry, right?”
AI Disruption and Napster's Transformation
28:00 to 32:34
Explore Napster's evolution from music streaming to AI and the company's future.
“Well, I mean, it's obviously real enough to be native in Microsoft.”
AI Disruption and Napster's Transformation
34:02 to 34:38
Explore Napster's evolution from music streaming to AI and the company's future.
“Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done.”
AI Disruption and Napster's Transformation
34:45 to 35:13
Explore Napster's evolution from music streaming to AI and the company's future.
“At LPL Financial, we like the sound of that.”
The Future of Humanoid Robotics
35:13 to 42:00
Understand the projected growth and challenges in the humanoid robotics market.
“Humanoid robots are moving out of the lab and into the real world, with deployments expected to surge in the years ahead.”
Snowflake's AI Positioning and Market Performance
42:00 to 42:50
Explore Snowflake's growth and its role in the AI landscape.
“That was the most eloquent, poetic earnings preview that this show has ever seen or heard.”
HPE's Earnings Insights and Market Dynamics
42:50 to 44:38
Insights into HPE's upcoming earnings and the influence of AI.
“The stock up roughly about 112 % in 2026.”
Broadcom's Earnings Expectations and Market Trends
44:38 to 45:51
An analysis of Broadcom's earnings and the semiconductor market outlook.
“Bloomberg Intelligence's Wu Jin-ho, HPE on deck after the market closed today.”
Broadcom's Software Revenue and Its Impact
45:51 to 47:19
Understanding the significance of Broadcom's software revenue in their growth.
“This week, custom silicon has been a big talking point.”
Market Recap and Closing Thoughts
47:19 to 48:25
A quick summary of market performance and final thoughts from the host.
“We again expect some healthy numbers from that business tonight.”
Transcript
Automatic transcript. May contain errors.0:00The most effective people at work aren't working harder than everyone else. They're working smarter, inside better systems. Superhuman Go, from the makers of Grammarly, is the AI assistant that works inside every tool you already use, always ready, already aware of what you're working on. It's a teammate whose only job is to help you be better at yours. With Go working with you, you can show off what you do best. See what Superhuman Go can do at superhuman.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.
0:40Now, 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. What if you could have even more and more and more help to pursue your goals? At LPL Financial, we offer more ways for advisors and their clients to thrive. So, what if you could? Paid advertisement investing involves risk including potential loss of principal. LPL Financial LLC member FINRA SIPC. Bloomberg Audio Studios.
1:19Podcasts. Radio. News.
1:27Bloomberg Tech is live from the heart of Silicon Valley with Ed Ludlow in San Francisco. This is Bloomberg Tech coming up in videos in advance talks to acquire AI startup Hugging Face in a transaction totaling$14 billion. Plus, Dell shares jump after the company boosted its annual sales forecast by$25 billion due to surging demand for servers to run AI tasks. And Uber cuts 10 % of its staff globally to streamline its core engineering, science, and delivery groups. Okay, so let's get right to our top story. NVIDIA could be closing in on its biggest AI acquisition yet. According to sources, the chip makers in advanced talks to buy Hugging Face in a deal that could total$14 billion.
2:15For more, Bloomberg's Ian King is here with us, along with Rachel Metz, both of whom broke the story last night. We've known about, Ian, the number of$12.9 billion. That's the kind of acquisition, and we've known that it's close. This$14 billion figure, what does that include? What's the new bit here? Yeah, I mean, according to our reporting, that includes an extra sort of billion dollars of retention. Obviously, NVIDIA doesn't want the talent, the core of that company to just disappear overnight when it acquires it. That's about keeping the staff. Rachel, it seems a strange place to go next, but what is Hugging Face?
2:50Like, we've been talking about them so often of late because they were subject to a mistaken hack by open AI models as one example. But what is it that they actually do? Yeah, so Hugging Face does a couple different things. It offers a platform that a lot of companies use to showcase AI models that they're releasing. It's also a place where anybody could post AI models or similar kinds of work that they're working on that they want to share with other people. And the company also, for a long time, has really focused on doing its own research. It's very focused on open research, open models, and has been fostering that sort of ecosystem for a long time.
3:32Ian, why would NVIDIA want to buy Hugging Face, as Rachel's just described it? What she said was key there, which is open models, right? At the moment, the industry is very concentrated on the hyperscalers and what they produce and what open AI does. And what NVIDIA really wants is a broad adoption of AI across the economy. It basically wants more customers. And Hugging Face is an area where it's sort of a bazaar where things can be propagated, where people can come and get what they need. and in theory, according to probably NVIDIA, a better thing for the overall industry. Rachel, is it a surprise that this is happening?
4:14Like in Hugging Face's history of either raising money or what we know about their co-founders selling to NVIDIA, did we have that on the bingo card? I think it sounds a little bit surprising on its face, but then if you think about it, it actually could make a lot of sense. The company has been around for about 10 years. They have a lot of investors, including NVIDIA. My understanding from people that I've spoken to that are close to the company say that NVIDIA has long been a dedicated investor of the company. And as Ian mentioned, they're aligned, especially on the open model. So it could be a really great matchup for both sides.
4:56A bit of M &A history for NVIDIA, Ian. We say$14 billion would be their biggest acquisition in AI. Where has NVIDIA done actual acquisitions as opposed to these investments, which it makes many of? Yeah, I mean, it doesn't have a huge record of acquiring. It tends to like to do things in-house. You could probably, the most recent one of note was what it did in terms of getting hold of the assets of Grok. That's right. Grok with a Q. That's right. Grok with a Q, the AI chip investor. But really the question here is what happens to Hugging Face if NVIDIA owns it? NVIDIA is a very powerful force in AI.
5:35Arguably, Hugging Face's greatest value is that everybody uses it. Rachel, real quick, is this imminent? Do we know if this will happen, when it will happen? Yeah, according to our reporting, it sounds like it is pretty imminent. And we are expecting that it will happen this week. okay bloomberg's ian king and retro metz with the story one of the most read stories this wednesday morning thank you very much let's get to the earnings story shares of dell higher off session highs but the company boosting its annual sales outlook by a massive 25 billion dollars with ai server revenue now expected to triple to 74 billion dollars this fiscal year Dell also reported a$95 billion AI server backlog and earnings in the quarter well ahead of Wall Street's expectations.
6:26Joining us now is Porsche Capital Management President and Owner Michelle Canale. Dell continues with momentum. This is the fifth straight quarter where they basically raised their outlook. What do you see as the story behind that? Well, first, thanks for having me, Ed. It's just a powerhouse in the server area. And a few years back, Dell was considered dead or almost what I would call a dinosaur. But they've been great in terms on the server side. And it's not just the AI servers I'm going to point to. It's also their servers where the revenue was up over 110 % year over year. Because if you're going to have AI servers, the corporations also need their traditional server infrastructure upgraded.
7:16I'm looking at Dell shares on a sort of longer time horizon. So even year to date, actually, 252 % gain. But it took some time for them to get credit, or at least for the market to understand, you need to have a Dell to put the server design together. Do you expect this to be a name that continues to have momentum on the equity side? I do, because they're the leader in this space. And they are very good in terms of their competitive advantages. They have a great breadth of products. Their supply chain is phenomenal. Their client relationships are great. And they've also been in the process of going through an optimization program in terms of cleaning out the product lines that aren't going to work and maximizing their margins.
8:02And that's another thing I think they don't get enough credit for. This year, their operating margins were over 11 percent. Last year, they were six. I mean, that's a huge trend. I want to zoom out here for some perspective and look at today's big number, which is$32 trillion. Yes, trillion. And that's how much data center spending is expected to reach by the year 2050, according to a report by PwC. So, Michelle, I don't know how you feel about such reports from the likes of PwC. But on your point there on the margin expansion, right, the story on the technology side has been that NVIDIA dominates the content ownership of the server.
8:43You know, Dell doesn't really contribute that much to it. But if we get to that, that$32 trillion number, surely Dell is going to take a big slice of that market. It's definitely the case because they've got the entire company lined up to continue to take share. And they're by far away the number one player in the server and AI server space. So, yes, if that survey is correct, which is very interesting because the last survey I saw was through 2030 going to$8 trillion. But now we're looking at$32 trillion. As you said, that's a huge number. And Dell will definitely benefit. And I think a lot of analysts on the sell side are seeing that.
9:25And you're seeing a lot of the estimates today push up by quite an amount. More broadly, across all the layers of the AI stack, where are you most focused right now? What are you looking at most closely? I started my career as a semiconductor analyst. And so I still am attracted to what's going on there on the public side, as well as the private side. And you're going to see more and more focus on the fact that it's not just customization by Broadcom or others, but you had NVIDIA buying MediaTek. And there are other private semiconductor firms that we're going to start hearing more about because copper is not going to make do with these chips in terms of transmitting the data.
10:15We need optics. We need other materials involved. And as this happens, the semiconductor story is going to broaden out and there are going to be a lot more winners. And there's going to be some market share losses along the way. In the private markets, we've seen that manifest in particular in specialized inference platforms. I'm trying to think of one etched as a good example. When you say there will be market share changes, who loses out in that context? I don't have exact names on that, but things are moving so quickly. You have NVIDIA that is smart enough to make these partnerships and make these acquisitions, but if you do not have the war chest to do that, maybe something like a Broadcom may have more problems in front of it.
11:07That's why the earnings announcement today is going to be very interesting and see if they're going to be able to prove that their AI revenue can stay within the$120 billion to$150 billion. How high or low has the bar been set for Broadcom? I'm really interested in what you just said after the shocker of NVIDIA's earnings last week, where they gave us a rare fiscal year outlook for growth of 70%, which was modeled on supply. It wasn't a growth outlook based on demand, but supply. Can you extrapolate through to Broadcom? I think they will benefit in the short term from the same issue. Last night when I was listening to, I think it's SemiCon over in Taiwan, the main theme was supply.
11:59There is not enough supply. So in the short run, the main players will benefit from that. But that may change with time. And as the means of building these small chips as they get smaller and as the amount of data is being communicated increases, the shares are going to change in terms of who's winning and who's losing share. Michelle Conor of Porsche Capital Management for the first time on Bloomberg Tech. Thank you very much. A story that broke this morning, Alphabet. This is Shares, very muted move after a federal judge ruled that Google doesn't have to sell off its advertising exchange and instead needs to make its ad tech tools work with those operated by rivals.
12:44The decision allows Google to avoid a forced sale. You'll remember at April 2025 ruling that the company had illegally monopolized two advertising technology markets. I would say I think that this was a probable or one scenario outcome and the stock move very muted after those headlines this morning. Now, coming up in the show, Uber is planning on cutting 10 percent of its global stock. That's in a massive restructuring effort. We're going to have the details next. This is Bloomberg Tech.
13:15The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. They've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files. Those aren't small inefficiencies. They're hours wasted every week. Superhuman Go gives you those hours back. From the makers of Grammarly, Go is an AI assistant that sits inside every tab and tool you already use, always available and already aware of what you're working on. Ask it to draft something, summarize a long thread, pull up a file, or prep you for a meeting.
13:56Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work, instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com. Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done. ChatGPT Work is a new way of working in ChatGPT that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work.
14:39It's designed to help you move from a chaotic starting point to a reviewable first version. So all the source materials, briefs, and scattered information that you have to grind through to turn into something useful can just become something useful. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com by selecting Work Mode, available on Plus and Pro plans. What if you could have even more and more and more help to pursue your goals? At LPL Financial, we offer more ways for advisors and their clients to thrive. So what if you could? Paid advertisement investing involves risk, including potential loss of principal.
15:18LPL Financial LLC member FINRA SIPC.
15:27Uber is looking to cut about 3 ,300 roles in its latest restructuring efforts to reduce its management layers. That's according to an email from Uber's CEO viewed by Bloomberg News. It's about 10 % of its global staff, Bloomberg's Natalie Lung, broke the story. Interesting. So there's a real emphasis here on cutting out management layers. Why is Uber so focused on that now? They have been brewing so much in the past five years. And there are actually a lot of teams apparently called micro teams that have one to two reports. And part of the effort here is to cut those micro teams in half. And as a result, there will be 20 % fewer managers.
16:04And in the CEO's note today, he talked about internal surveys of people complaining that they're doing a lot more coordination and alignment versus just building things. So this is really an effort to make the organization simpler and faster. Natalie, we say, as it's outlined in Dara Khoshwashoggi's memo obtained by Bloomberg News, that it's a restructuring. But there are areas that Uber's really focused on. It wants to grow in. What did Dara say about that in his correspondence? Yes. Obviously, the core ride hailing and delivery business are very important. They're investing more into drivers and merchants and trying to get ahead in global markets and expansion markets where DoorDash, its rival, is also competing in, especially in Europe.
16:50They also talked about, you know, building an autonomous future, which is very important because they committed, you know, more than$10 billion in investment in the next few years. And they're going to have to reallocate capital to be able to do that. I think most people are familiar with Uber, right? You know, as a ride-hailing platform, largely with human-driven cars, but summarize what the robo-taxi strategy is. For Uber, it's kind of multifaceted. Yes. So Uber doesn't want to make their own robo-taxis. They're not going to build their own Waymo car or Tesla, but they are partnering with a bunch of different providers such as Neuro, Lucid, Rivian and Averight to sort of be able to aggregate all these cars and in the future become the go to platform where people can hail multiple kinds of autonomous vehicles.
17:38Blueberg's Natalie Leung with the reporting breaking the story on Uber, cutting 10 percent of staff. Thank you very much. another big story. AI coding startup Cognition is set to close a new funding round of around$1 billion, according to sources. The latest round will place the startup at a$47 billion valuation. Bloomberg's Rebecca Torrance covers all things startup, venture capital, and broke the story. This is interesting because, very simply, it is a marked and very quick jump from Cognition's last valuation. That's right, Ed. Cognition just raised at a$26 billion valuation in May. This is a massive jump from that, obviously, and there has been a ton of investor interest in this round.
18:18We reported that the round is looking to be around a billion, but it could exceed that amount because it's received, the company has received nearly$10 billion in total investor interest. That's due in part to the growth of the company and in the investor interest more broadly in AI coding startups. I should say that in our reporting, and I teamed up with you on this one, we don't have a sense of all the details yet, right? We don't necessarily have who's participating in this round or leading it, but the sense is that it is well underway for Cognition to raise this money and that this valuation is getting settled.
18:53That's right. The deal is pretty set to close and certainly moving forward. It's also notable that Cognition's annualized revenue has passed$900 million at this point. That's up from$492 million in May. So that's one of the things that's really driving a lot of interest here. Russ, right, because I think Scott Wu, the CEO, had come on the show and shared that AR figure. Just very quickly, you and I had reported that SpaceX approached cognition with the idea of acquiring them. Cognition said no. We published the story. Scott Wu, the CEO of cognition, and Elon Musk weighed in. But just summarize that historic reporting for us, because I think it's relevant here, given the timing.
19:34Sure. So we had reported that SpaceX had approached cognition in recent months about a potential acquisition. Cognition had rebuffed them. Cognition has always been very vocal about wanting to remain an independent company. That's why they said they did this last fundraise. So it's unsurprising to see another fundraise well underway here. And SpaceX obviously just closed its$60 billion acquisition of Cursor in August. And we had also reported previously that that had driven investor interest into Cognition and certainly seems to continue to be doing that now. Okay, to recap, Cognition is set to raise a billion dollars at about a$47 billion valuation, according to Rebecca's reporting.
20:13Bloomberg's Rebecca Torrance, thank you very much.
20:22Lots going on around the world. Let's get out to New York, where Bloomberg's Yaira Hanand is standing by. Hey, Yaira. Hi, Ed. It's time now for Talking Tech. First up, Anthropic has released a cheaper version of its powerful Fable AI model. The company says Fable 5.1 is better at coding and science tasks. Enterprise users will pay the same price per token as Fable 5, but less when the model resurfaces previously processed information. Plus, Apple's new CEO, John Chernis, will get about$58 million in compensation next year. That's according to an SEC filing. That includes$3 million in base salary and a$55 million stock grant.
21:04Meanwhile, Tim Cook, who has moved into the executive chairman role, will be paid around$47 million. Not bad. And TSMC's need for chip-making tools has nearly doubled since the end of last year. That's according to a senior executive. The comment was made at Semicon Taiwan, where others in the industry also touted increased demand. Here's Berthold Schmidt, the CTO of German laser maker Trump, a key supplier to ASML. Right now on EUV level, we are trying to triple our output within the next two years. And in areas like the plasma generators, with the demand we are having, the order books that are filling up, we see solid growth and we probably also double, maybe even triple on the generator side within the next two years.
21:53Schmidt added that despite the demand, he did not say the industry was in a super cycle. Ed. Thank you, Hyra. NVIDIA CEO Jensen Wang is urging the world's largest economies to accelerate their adoption of AI as a way to enhance growth. Speaking at a G20 meeting in North Carolina, Wang said, quote, Every single country needs to build infrastructure to support their local economies and called AI the great equalizer. Bloomberg's Maggie Eastland was in the room and she's with us now. This is a sort of side meeting to the main ministerial G20. But the biggest names in technology are there and they are having a pretty clear message.
22:35Go through what Jensen Wong said and who else was in the room. Absolutely. So we've heard already this morning from Jensen Wong, as well as Tom Brown, the chief compute officer and co-founder of Anthropic. And they both emphasize this message to the foreign nations that have joined the U.S. here to say we need to build data centers not only in the U.S., but also abroad. And this is, of course, coming at a time when within the U.S. there's growing backlash in communities towards having data centers built there. So it's a very interesting dynamic. But the one message that's been consistent across all the CEOs and tech executives that we've heard from is that we need to build data centers.
23:19We had heard from Elon Musk yesterday who framed this as AI will equate to economic growth. Just give me the bigger picture why they've got all these tech CEOs into the room. Absolutely. So from the U.S. perspective, they're working on ironing out and getting all of the nations to agree to what White House official Michael Kratios is calling the Carolina principles. So that's essentially a framework for regulating or in this case sort of not regulating technology. So it outlines, you know, having a very tailored approach and considering the opportunity cost of not innovating fast enough. So they're trying to get other nations to agree to this.
23:59And in doing that, they're presenting all of these tech executives and having them weigh in on the importance of innovation and the importance of keeping regulation moderate so that all of these tech founders can continue doing what they're doing. OK, and elsewhere at the actual ministerial G20, AI has also been a topic. Here's Treasury Secretary Bezum. I think that the AI companies, whether it is the builders of the data centers, whether it is the labs themselves, have done a horrendous job, horrendous job of explaining themselves to the American people. And I think we need a big reset on this.
24:38They're going to have to take some of the blame and they are going to have to convince the American people that all the benefits will not accrue to a small group. Maggie, real quick, that's a slightly different AI perspective. Absolutely. And one thing it makes me think of that I heard from Tom Brown just this morning is, you know, he actually said that he loved President Trump's post about data centers. So Anthropic, that's a company that's historically been a louder voice in terms of needing to pay attention to the safety risks of AI, is now coming out at this public summit and saying, hey, we really need to build data centers.
25:15Boomberg's Maggie Eastland, terrific reporting on the ground. Thank you very much.
25:24Welcome back to Bloomberg Tech. You probably remember Napster as that company that upended the music industry, right? Yes, that Napster, 1999 Napster. Now it's trying to do something similar with AI. Napster's reinvented itself as an agentic AI company with Microsoft selling its platform through Azure. Joining us now is Napster CEO, John Acunto. Okay, that was sort of an abridged history. You joined Napster in March of 2025 when your other company had acquired it. But let's just start with the very basics here. What are we talking about, John? Well, we're going from streaming music to streaming intelligence, and we're just unbelievably excited at making that transformation.
Read the full transcript
26:08We know that Napster really changed the world of music, and we're really looking to change the world of agentic AI, specifically streaming intelligence. The Microsoft part of this, why is it that Napster needs to go through Microsoft and Azure to do this? Explain the deal, essentially. Well, I mean, in essence, there are many companies that are partners, obviously, with Microsoft. And we saw an opportunity to sort of continue to sit and work with all the large language models with our agent. And quite frankly, Microsoft was the best partner for us at this point to really make sure that our customers own the data, own their own customers, own their experience, and really own their models.
26:59And I think that's the path that we saw Microsoft on. Okay. So, yes, dear Bloomberg Tech audience, there'll be loads of you watching. Napster, the music streaming service now offers an AI agent, essentially. What is the agent called? What does it do? Why is it better than other AI agents? Well, first off, the agent's on the edge. And so the cost difference is the first thing you feel, right? Is, you know, our costs are about a penny a minute versus others that can be 25 cents to a dollar a minute. So that's the first experience that you have is the cost savings. But most importantly is having a one-on-one conversation, whether it's online or in-store with an agent, really we think is the very next step of AI.
27:48John, we have seen shoe companies pivot to AI. We have seen crypto companies pivot to AI. There will, with respect, be loads of people watching this interview thinking, okay, he's going through the full lexicon and glossary of AI words. What are we actually talking about? Like, how real is this? Well, I mean, it's obviously real enough to be native in Microsoft. And listen, our hat's off to everybody that wants to try to reinvent a brand. We feel like we're really just transitioning disruption. We feel Napster did its part and its job and history of disrupting music, and we're going to do our part in disrupting AI.
28:28And this is really just the first step. And in terms of Allbirds and some of these others, we truly wish them luck. I think the first step is finding a great product, which we already have. That's kind of demonstrated here with this relationship. I think what was interesting about what you said there is that you are basically leveraging your existing expertise and core competencies, right? Like Allbirds was a sneaker company and it pivoted to compute. What is it that Napster, its people is good at that makes you a prime candidate to be an AI company? Well, I mean, ultimately, we're a collection of actually a lot of companies that we've acquired over the years to really make it so that our customers can have a one-on-one conversation online with their customer.
29:16And when you think about it, how many websites have you went today where you had a one-on-one conversation with your customer? So we really spent the past five, six years building and developing those expertise through acquisition and through working with customers globally. John, going back to some of that history, so Infinite Reality acquired Napster. The overall entity is a private company. just from sort of a financing perspective and Napster's place in the world? How are you thinking about its future as a private company, looking at public markets? And I guess in that context, what your capital needs are.
29:55You talked about how your economics are pretty good. Yeah, the economics are great. You know, we really see the fourth quarter of this year and most of 27 being a true launching pad. we're in the adoption phase of AI where our customers are really looking to adopt the product and being able to access it directly in Microsoft and have that build and that relationship we see is a big lift for us in late 26 into 27. I went back through the archives, actually it wasn't that long ago so like January of this year the music streaming part of Napster is gone, it's shut down Is that it? That's the end of that story?
30:40No, not yet. I mean, you know, we really want to have fun with the brand. We really want to bring back sort of music through the eyes of AI and the ability for creators to still create but use AI tools to do that creative presentation. It's a very delicate thing right now, obviously, in the creative space, AI. So we're going to kind of see how that settles out, see how the major labels want to approach that in a way which doesn't disenfranchise creators. Just really quick, I want to go to your infrastructure or lack of infrastructure, right? So it seems like the pitch on your economics is in part that Azure allows you to do it with a sort of asset-like model, right, through a significant cloud computing partner.
31:24So you don't have any serious outlay on your own compute capacity, essentially. Well, exactly. Our agents sit on top of all models. So we're really, quite frankly, the interface, what our customers would use to have an interface with a customer. And so it's super important that we keep our focus on how our customers can have the best interaction with their customers and really pivot the discussion away from cutting costs and how AI displaces folks and really sort of pivoting the conversation to how we can learn through these online interactions, for example, why 90 percent of people aren't buying from your website.
32:14and start to figure out how AI can be a tool to the blue sky side of the equation. Launched as a music streaming platform in 1999, Napster. Today, AI agent through Microsoft Azure, Napster. Jonathan Tau of Napster, thank you very much for joining us on Blue Boat. Thank you so much. Coming up on the program, Zanitsa Tarova, who's the head of thematic thick research at Barclays, is going to join us to talk about the next phase for the humanoid robotics industry. That is her expertise and deep research area. And it's been a big topic on the show for quite some time now. Humanoids, this is Bloomberg Tech.
32:55The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. They've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files. Those aren't small inefficiencies. their hours wasted every week superhuman go gives you those hours back from the makers of grammarly go is an ai assistant that sits inside every tab and tool you already use always available and already aware of what you're working on ask it to draft something summarize a long thread pull up a file or prep you for a meeting go handles it without you ever leaving the page you're on This is what it looks like when AI actually fits into your work, instead of adding to it.
33:45It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com. Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done. ChatGPT Work is a new way of working in ChatGPT that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work. It's designed to help you move from a chaotic starting point to a reviewable first version.
34:24So all the source materials, briefs, and scattered information that you have to grind through to turn into something useful can just become something useful. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChachiBT.com by selecting Work Mode, available on Plus and Pro plans. What if you could have more wins? More support? More sound effects? At LPL Financial, we like the sound of that. Because LPL offers more. Advisors, what if you could have more ways to help your clients? Ready to invest? What if you could find an advisor that really understands you? When it comes to your finances, your business, your future, At LPL, we ask, what if you could?
35:09Paid advertisement. Investing involves risk, including potential loss of principal. LPL Financial LLC, member FINRA SIPC.
35:21Humanoid robots are moving out of the lab and into the real world, with deployments expected to surge in the years ahead. Our next guest thinks the industry is entering a major scale-up phase, but says a shortage of real-world training data remains a key hurdle. Zunita Todorov is the head of thematic thick research at Barclays. Back on the show. I think let's start by, Zunita, you reminding us what your kind of longer term forecast is for the addressable market or size of this humanoid robotics market globally. Hi, it's great to be on the show again. Right, so humanoid robotics is scaling really fast.
35:58We estimate that by 2035, it could become a$200 billion market and it could be part of a much bigger physical AI market worth a trillion dollars. Now, if we want to put some numbers in terms of deployment units behind these forecasts, our forecasts look like this. So in 2026, we expect that we're going to see about 60 ,000 new humanoid robots deployed worldwide. That's a massive scale up compared to what we saw last year when we just had 15 ,000 robots. But I think more importantly, we're just scratching the surface of what's possible. And we're just we're going to see much more. And according to our latest research, by 2035, we could see about 13 million new humanoids deployed each year across two major markets, the U.S.
36:43and China. And I think that if we continue to see the same technological progress as we have seen so far in 2026, and if investor appetite for the theme continues to hold strong, I think we could easily see about 30 million humanoids in operation by 2035, which is quite a sizable fleet if you think about it. OK, right now we're looking at video of Tesla's Optimus. And what I find so interesting about this is the number that you shared of 60 ,000 units of humanoid robotics globally in this year. Don't most of those come out of China? Well, about 85 to 90 percent of us will likely come out of China.
37:25That's for sure. China seems to have the edge right now. But I think there are small ecosystems popping up worldwide. And I think the U.S. is very quickly narrowing the gap in terms of maybe not so much in terms of the number of units that are deployed out there, but certainly in terms of the expertise of the robots and what they can do in the real world. The limiting factor that I'm being told on a near daily basis by industry is real world data. So what we mean by that is the ability of the robot to interact with the humanoid robot, to interact with its immediate physical environment, be that a factory, other industrial setting, whatever it is.
38:05One solution seems to be synthetic data. Could you just quantify the challenge and how you see the market trying to address it? Right. I absolutely agree. I think that the number one challenge right now is not really demand for humanoids. I think that there is strong demand for humanoids in all kinds of applications. The biggest challenge is the lack of physical AI data. And I think that this is something that people find very surprising because the reality is we have tons of digital data, but it's just not useful for training humanoid robots. And I do think that the data necessary to train and to scale, more importantly, a truly general purpose humanoid robot does not exist on the Internet.
38:43It needs to be retrieved in the real world, in the physical world. And that could mean something like recording humans doing everyday tasks from doing the laundry to household work and even deliveries. And this way, generating data that is required to train this very complicated general purpose models. Now, you mentioned simulation, and I think this is definitely one solution, one bet to this challenge. However, I do feel that as humans, we live in a very fuzzy, very unstructured world. And simulation has its advantages. It's definitely a cost-effective option. But I think ultimately simulation will hit a wall in terms of how much of that fuzziness, how much of that messiness it can capture.
39:26So ultimately, I think we'll have to lean on the real world and the physical data that's out there. And we definitely see more and more companies popping up, which reflects a general industry-wide push towards assembling this data that is going to feed models for the next five to ten years. You know, the most bullish on the humanoid industry put a lot of faith in the digital twin, the idea that a digital twin can ground the humanoid in the real world physics around it. That's on the software side. On the hardware side, is there even a supply chain for the humanoid robot anywhere in this planet, on this planet?
40:03it's i it's it's beginning to it's beginning to be put together it does not exist uh at the scale that we would think when it comes to other technological products such as evs for example other types of industrial robots i think that the reason is that this is going to come all the time we just need to find the skill application for humanoid robots they are most efficient functional form and i think i'm confident that the supply chain is going to emerge around it i think that if you think about it at a high level we don't need to reinvent everything from scratch. I think there are lots of complementarities between I would say electric vehicles and humanoids, especially when it comes to actuators and the motion system so I think we can borrow a lot from that.
40:48It's definitely not going to be plug and play but I think that the basics of that exist and I think as the technology improves, the supply chain is definitely going to scale accordingly. Zonica Tororova, who is the head of thematic thick research at Barclays, but is just all in on the research behind humanoid robotics. Thank you very much indeed.
41:18Many tech earnings coming after the bell today, and we start with Snowflake, the company revenue expected to rise 30%, which would mark a third straight quarter of growth at or above 30%. for the most Brody Ford joins us now. What's the Snowflake story right now? The data layer is one that's been spared from the kind of scourges of the SaaSpocalypse. Things have been going pretty well for Snowflake. I mean, they're down a bit today, but overall their stock has been at its highest point in years because you have all your data already in this platform and Snowflake's been able to say, well, now you need to process it with AI and we're going to sell you additional products to do so.
41:55So their revenue growth has been pretty healthy and Wall Street's been pretty happy. So they just want to see that adoption continuing. That was the most eloquent, poetic earnings preview that this show has ever seen or heard. Snowflake is successful, right? Like they have top line growth that seems strong, profit. Why is it that they're not sort of like the high flying name in the world of AI? Well, the question, you know, I mean, they're not making any of the real core interesting AI products everyone's excited about. what they are doing is a platform where you can use all of them on your data.
42:32And they've been successful at doing so. I mean, one of the debates is whether their new AI products cannibalizes their kind of core computing revenue. This doesn't appear to have played out yet, but it's certainly something we'll be watching. It can be a watch. Exactly. Bloombo's Brady Ford on Soflake. Thank you very much. Next up, HPE set to report earnings after experiencing its best year in the stock market driven by AI fever. The stock up roughly about 112 % in 2026. Bloomberg Intelligence Analyst Wu Jin-ho joins us. What is the BI preview on HPE and what do you think is on deck for earnings?
43:08Yeah, so Ed, quite straightforward. Look, AI is going to matter. We do expect a little bit of upside on AI, primarily from memory inflation. However, the biggest story is traditional servers and the uptake by customers on traditional servers. And there was outperformance on Dell last night. And I expect similar outperformance on AI servers as well as networking for HPE tonight. So that's the read through that I'm looking for. Dell basically is seeing momentum in AI servers. Is it therefore just expected HPE sees the same? And just the basics of the role of those two companies. Yeah, look, from a scale standpoint, HPE doesn't even match Dell's scale, right?
43:54But at the end of the day, there is a fair amount of demand that HPE is eventually going to benefit. But from the enterprise server side, corporate server, corporate IT, that's where a fair amount of meaningful upside that Dell saw, HPE is going to see the same. Woojin, big picture. What's your takeaway been from this earnings season in the parts of the infrastructure that you cover? What is the big story that's happening in that market? Yeah, look, at the end of the day, everyone is spending. Despite memory inflation, customers are willing to absorb it, whether it's going to be the cloud or for enterprise IT.
44:33The only people that's not willing to spend it are the PC customers. Bloomberg Intelligence's Wu Jin-ho, HPE on deck after the market closed today. Last but certainly not least, Broadcom also reporting earnings after the closing fell. Investors hoping for a reprise after a difficult summer in the stock market where shares went on a$520 billion tailspin since they hit a high, which was back in early June. Bloomberg Intelligence analyst Kunjun Sabani is here for more. Actually, the way I look at it is how high is the bar or low is the bar for Broadcom after the shock that NVIDIA gave us on that outlook for growth last week?
45:12The bar is very high, specifically for their fiscal 2027 outlook. Remember, last quarter, they sort of missed on the 3Q guide and they reiterated the$100 billion plus revenue for fiscal 27, which investors did not like. Since then, a lot of tailwinds have gotten in their favor. The new OpenAI chip is running ahead of schedule. Programs at Meta and Anthropic have even gotten stronger. So there's a lot of expectation. As well, you said, Ed, with the big guide that NVIDIA provided, there's a lot of expectation to give a much higher number, almost 20 to 30 % higher number to that$100 billion number.
45:51This week, custom silicon has been a big talking point. I had an interview with Jensen Wang where he basically said, I'm not worried about custom silicon, but I'm going to make it easy. for custom chips to work with our tech. You put research out this morning, Broadcom cements number two AI chip position as XPU scales. Explain that to us. Yeah, look, so when you look at the other world, which is the custom chips, right? Broadcom has almost close to over 80 % share when it comes to revenue, right? Broadcom has almost now seven to eight hyperscale wins. So when you think of ASIC or custom silicon volume increasing, the number one beneficiary remains Broadcom.
46:29And the step that NVIDIA is doing, like we talked about, is hedging its risk. So even if you go with a TPU or with a Tranium-based data center, NVIDIA can still play a role in there with its NMILink portfolio. But Broadcom definitely benefits across the board here as ASIC volume rises. Kunjan, you're a chip guy. You're a semiconductor analyst. And Broadcom is kind of like an interesting business that has a software component. Do you pay attention to that software component? We definitely do. I mean, almost depending on what they report tomorrow, but now close to like 40 % or so the revenue is from software.
47:05And one thing people always forget, this software business is what really helps them generate significant free cash flow, which then they go and in turn deploy in the semiconductor business to drive the growth which we are seeing in the AI. So that business is very important. We again expect some healthy numbers from that business tonight. Yes, it's something that, again, doesn't surface on the top, but it's really the core driver in terms of free cash flow generation. It's very interesting. The other part of the BI thesis is that whatever they might say tempers cyclality. We just have 20 seconds, but what is it that they're trying to put to rest with that?
47:44There's not a lot of cyclality. It's just like, again, we're expecting a big beat to the fiscal 27 number. in a lack of that which could be a difficult time for sentiment. Bloomberg Intelligence says Kunjan Sabani Broadcom after the market close. That does it for this edition of Bloomberg Tech. There's been a lot of news so just a real quick check on what markets are doing generally in the technology space. NASDAQ 100 is flat. The Mag 7, some outperformance, chips outperformance and NVIDIA continues to have momentum and it has all week in part that MediaTek deal which I encourage you to go and check out.
48:18Also check out the podcast And also find me on Twitter or X. Find me on Instagram. Let's talk. We love your feedback. This is Bloomberg Tech.
From the publisher
Bloomberg’s Ed Ludlow breaks down Nvidia's advanced talks to acquire AI startup Hugging Face in a $14 billion deal. Plus, Dell shares jump after the company boosted its annual sales forecast by $25 billion due to surging demand for servers to run AI tasks, and Uber cuts 10% of its staff globally to streamline its core engineering, science and delivery groups.
See omnystudio.com/listener for privacy information.

