Nvidia's AI Push, Paramount's Merger Pause & China's AI Race

21 Jul 2026 · 39 min · 21 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

The episode covers three main tech/business stories: Nvidia’s next-gen AI chips and the compute rebound; a legal pause on the Paramount–Warner Bros. Discovery merger; and China’s accelerating AI race amid U.S. scrutiny of Chinese models.

Guests and backgrounds

Samir Dallakia, partner at Bessemer Venture Partners (Growth Investment Practice), discusses frontier vs open models and VC opportunities. Stephanie Aliaga, global market strategist at J.P. Morgan Asset Management, focuses on China’s AI model economics and market policy risk. Joanne Feeney, advisor at Advisors Capital Management, argues AI efficiency boosts overall demand and highlights durable moats in data-center suppliers. Jessica Malujin, Competitive Enterprise Institute (Center for Technology and Innovation), comments on the merger as a political/antitrust dispute. Lucas Shaw, Bloomberg Screen Time, explains the merger timeline and antitrust theories.

Key claims/examples

Nvidia’s “Verirubin” rollout is in full production; servers are already running (MV72 rack) and installation can take “tens of minutes.” TSMC may raise chip prices 5–10% starting 2027, affecting advanced and mature chips. U.S. plans to scrutinize Chinese models for IP theft/distillation; examples include Moonshot and DeepSeek accusations. Paramount–Warner pause: judge issues a temporary restraining order; concerns cite movie and cable-network market shares (25–35% for films). Aliaga notes “API cheap but compute/electricity clusters still costly,” and cites Jevons paradox (efficiency expands demand). Feeney warns memory pricing (e.g., Micron) may be cyclical as supply ramps. Dallakia highlights Fireworks’ ARR growth (100M to 1B in <1.5 years) and Moonshot’s Kimi K3 fundraising/listing plans (up to ~$50B valuation).

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

NVIDIA's AI Chips and Market Dynamics

0:00 to 0:22

Discussion on NVIDIA's latest chip designs and the semiconductor market trends.

“With the highest number of young STEM graduates per capita in the EU, Ireland has the people and skills your company needs to succeed here.”

NVIDIA's AI Chips and Market Dynamics

1:00 to 1:30

Discussion on NVIDIA's latest chip designs and the semiconductor market trends.

“For midsize and large companies, risk can affect multiple parts of the organization at once, from property and liability to cyber and regulatory challenges.”

NVIDIA's AI Chips and Market Dynamics

2:15 to 5:56

Discussion on NVIDIA's latest chip designs and the semiconductor market trends.

“And the AI race is entering a new phase, with Washington promising to scrutinize Chinese models as investors hunt for the next big winners.”

TSMC Pricing Strategy and AI Implications

5:56 to 9:19

Analyzing TSMC's pricing strategies and their effect on the AI chip market.

“The increases would affect both advanced AI chips, but also more mature semiconductors, potentially impacting customers from NVIDIA and Apple to Amazon and Alphabet.”

China's AI Landscape and Market Confidence

9:19 to 14:00

Exploring China's AI development and its impact on global markets.

“Joining us now is Stephanie Aliaga, J.P.”

AI Demand Remains Strong Despite Slowdown

14:00 to 14:31

Investors remain optimistic about AI companies' ability to monetize demand amid anticipated CapEx growth slowdowns.

“And we do think that we're going to see a deceleration in this pace of CapEx growth into next year, and particularly 2028.”

Judge Halts Major Merger

14:42 to 15:27

A judge's decision to pause the Paramount-Warner Brothers merger raises questions about antitrust concerns.

“Joining us next, Jessica Malujin, director of the Center for Technology and Innovation with the Competitive Enterprise Institute.”

Impact of the Paramount Merger Delay

16:26 to 17:11

Discussion on the implications of the merger delay for Paramount and Warner Brothers, including financial stakes and market reactions.

“Clown a balanced account required to be eligible for cashback points.”

Antitrust Concerns and Political Dynamics

17:11 to 20:03

Analysis of the political context surrounding the merger and the role of state attorneys general in the decision.

“a federal judge ordered a pause on the merger over potential antitrust violations.”

President Trump's Remarks on Iran

20:03 to 22:13

Summary of President Trump's statements regarding tensions with Iran and military strategies during a press conference.

“Let's stick with the Paramount-Warner Brothers Discovery merger.”
Show all 21 chapters

Market Reactions to Political News

22:13 to 24:27

Discussion on technology stock performance in response to President Trump's remarks and broader market conditions.

“Bloomberg's Washington correspondent, Tyler Kendall, out in D.C.”

Chip Suppliers and AI Demand

24:27 to 27:33

Exploration of the importance of chip suppliers in the AI landscape and their competitive advantages in the market.

“They've already actually shipped some Vera Rubin-based platforms to some of the biggest AI companies where workloads are expected to run imminently.”

Challenges in the Memory Market

27:33 to 28:00

Analysis of the memory market's cyclical nature and the implications of increased supply on pricing.

“Is it healthy to sort of go from a situation where you enter a bear market to and there have been big swings in the socks?”

Market Dynamics and Future Earnings

28:00 to 31:10

Discussing the impact of supply and demand on market pricing and stock volatility.

“And the reason is that those profits, those sales numbers have in large part been driven higher by massive price increases, hundreds of percent.”

Copyright Settlement in AI

31:10 to 32:44

Overview of Anthropic's historic copyright settlement and its implications for the AI industry.

“Anthropic has reached a landmark$1.5 billion settlement in a copyright lawsuit over the use of millions of pirated books to train its AI models.”

Venture Capital Opportunities in AI

33:28 to 37:51

Exploring venture capital trends and the emergence of significant AI opportunities.

“China's Moonshot AI is preparing to begin discussions in August on a final round of fundraising before listing in Hong Kong at a valuation of as much as$50 billion.”

AI Growth and Market Potential

37:51 to 41:37

Discussion on the rapid growth of AI companies and market potential across sectors.

“And so the frontier model, the penetration in enterprises is so tiny relative to what it will be that the frontier models will continue to excel.”

Tesla's AI Spending Concerns

41:37 to 42:00

Examining Tesla's AI spending compared to competitors and investor expectations.

“As big tech rushes to pour trillions into AI investments, Tesla is lagging behind.”

Tesla's Earnings and Investor Expectations

42:00 to 44:05

Explore what investors are looking for in Tesla's upcoming earnings report, with a focus on capital expenditures and AI advancements.

“And investors are growing a little bit concerned.”

Tesla's Earnings and Investor Expectations

44:37 to 45:03

Explore what investors are looking for in Tesla's upcoming earnings report, with a focus on capital expenditures and AI advancements.

“If you've ever waited on a refill or couldn't schedule an appointment, you get it.”

Tesla's Earnings and Investor Expectations

45:07 to 45:34

Explore what investors are looking for in Tesla's upcoming earnings report, with a focus on capital expenditures and AI advancements.

“in Lacqua on Bloomberg Television or wherever you get your podcasts.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00With the highest number of young STEM graduates per capita in the EU, Ireland has the people and skills your company needs to succeed here. IDA Ireland, the National Investment Development Agency, can help you find and nurture the people you need to internationalise and thrive. Our talent is just one of the extraordinary benefits Ireland has to offer. Learn more at idaireland.com. Invest in extraordinary.

0:30Ed Ludlow: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. Now, 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. When you're running a business, the best days are the ones where priorities stay on track. For midsize and large companies, risk can affect multiple parts of the organization at once, from property and liability to cyber and regulatory challenges.

1:13At that level, managing risk becomes an ongoing discipline. At the Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. And when losses do happen, that work is paired with insurance coverage shaped by years of underwriting, risk engineering, and claims experience. Learn more at thehartford.com slash risk mitigation. Policies provided by Hartford Fire Insurance Company and its property and casualty affiliates, Hartford, Connecticut. Bloomberg Audio Studios. Podcasts. Radio. News.

1:53Bloomberg Tech is live from the heart of Silicon Valley with Ed Ludlow in San Francisco.

2:02Ed Ludlow:This is Bloomberg Tech. Coming up, we check in on NVIDIA and its next generation AI chips as the battle for compute gets into high gear. Plus, a judge hits pause on Paramount's merger with Warner Brothers Discovery throwing the blockbuster media deal into limbo. And the AI race is entering a new phase, with Washington promising to scrutinize Chinese models as investors hunt for the next big winners. We're going to discuss with Samir Dallakia of Bessemer Venture Partners. Chipmakers, semiconductor stocks, the rebound continues and is gathering some pace. We went from closing in a bear market Friday, which was catalyzed by the Kimi K3 release, to now being a little more confident after that pullback.

2:44Ed Ludlow:The Philadelphia Semiconductor Index, or SOX, is up more than 4%, a big point being that even in yesterday's session, we saw some pretty wide swings and choppy changes between the level of gains that that index had. We're also checking in on shares of NVIDIA. Actually, just in the last 30 seconds, taking a little bit of a tick higher, up 1.4 % have been higher. NVIDIA says it's full steam ahead on its latest chip designs. Accelerated computing systems based on the new Verirubin platform have been delivered to major AI companies and are about to start running workloads, Bloomberg's reporting. The Verirubin rollout has been under scrutiny by both investors and analysts.

3:24Ed Ludlow:For more, Bloomberg's Ian King joins us fresh off a trip to Inverio HQ in Santa Clara. And you've been speaking to one of the key leaders there about the actual state of play. Yeah. What is it? Ian Book, who's basically the head of their data center business, dragged us all into Jensen's office and we all got a walkthrough of where they are, basically. And really what they were keen to do was to kind of slay the dragon, this kind of lingering myth that there is somehow a problem with the rollout. They said, you know, we're going to go faster. Everything's working. They even took us to a secret location and showed us a rack of servers that's already actually in use, up and running.

4:04Ed Ludlow:Ah, a rack of servers. This is a representation of MV72, 72 Rubin GPUs. Obviously, this is not life-sized. It wouldn't fit on set with you and I. There was the key takeaway from the real version of this, that it's in full production. You know, that, I think Jensen's been quizzed about that quite a lot recently, but that is pretty definitive answer to what the market's been questioning. Right. There was a report out there saying that they had some actual materials issues, that it wouldn't be ready. And what they did was say, look, here it is. Guess what? What used to take hours for a human to put together, a human who would make mistakes, now we've designed it so it can be put together by a robot.

4:51Ed Ludlow:Great pictures, by the way, that you took on your trip down to NVIDIA HQ. Yeah, with the iPhone. and they said, look, when we drop one of these things off at the dock at a data center in the past, you know, it would take a long time to get it up and running. Now we're talking tens of minutes to have it from the loading dock to being up and running. So really what they're pushing the idea is we're going faster, try and catch us. A big part in the moment of the rebound in chip stocks is Intel's having a pretty good day. And as an aside to that, we got some, a statement from the company about them right sizing, streamlining one of its divisions.

5:24Ed Ludlow:What do we need to know? Yeah, I mean, they've already got rid of multiple tens of thousands of employees and reduced their cost count appreciably. And that looks like it's something which is going to continue. They've said that their data center group is going to be dealing with some more redundancies to get those costs down and get that company into the kind of lean position it feels it needs to be in. Bloomberg's Ian King, who leads our coverage of semiconductors. Thank you very much. The cost of building the AI future may be about to get even more expensive. According to a Nikkei report, TSMC has finalized plans to raise chip making prices by as much as 10 % starting in 2027.

6:04Ed Ludlow:The increases would affect both advanced AI chips, but also more mature semiconductors, potentially impacting customers from NVIDIA and Apple to Amazon and Alphabet. For more, Bloomberg's Peter Elstrom, executive editor who leads our coverage of Asia Tech, is with us. This, again, is a Nikkei report, but what's in it? What do we need to know about TSMC's actions? Yeah, hi, Ed. Well, this is one of the questions that the tech industry has been asking for a long time. TSMC has this unique position where they have a lock on the most advanced manufacturing of chips. So we talk about NVIDIA as a chip maker.

6:41They don't do their own manufacturing. They have to farm that out, and they farm it out to TSMC in particular. So when you look at NVIDIA's income statements, you see they have gross margins of something like 75 % these days. And one of the big questions is why doesn't TSMC, which is essentially the only company that can make their chips, why don't they raise their prices? Why don't they take some of that profit from NVIDIA? And one of the reasons they've been cautious about raising their prices is they don't want to give companies like NVIDIA, Amazon, Apple, a reason to go to the other foundries out there.

7:12There are only a couple of foundries that can do this. They don't have very much market share. They're not considered as good as TSMC, but Intel and Samsung in particular are trying to get that business, and TSMC wants to keep it. So they've been raising prices, but very, very cautiously. And despite those rich profit margins and some of their customers like NVIDIA, they've been cautious about this. So now what we're hearing is that they are going to raise prices, but very incrementally, 5 % to 10 % to be able to capture some more of that profit margin. That's partly key for them because they are changing their business model in a sense.

7:43They're manufacturing now chips not just in Taiwan, but they're trying to build these facilities in other countries, particularly the United States. So part of this is just to try to capture some of that profit that we've seen in the rest of the AI industry.

7:56Ed Ludlow:There are lots of reports on the Bloomberg this morning relating to Chinese AI. And one of them is a report from Reuters that ahead of President Xi's September trip to the United States of America, there is a plan for the U.S. and China to hold talks. What do we know? Well, Scott Benison actually gave an interview today where he talked about some of these issues. There are many different issues that they want to discuss around AI, including AI security, proper uses of AI models. But there's also this question about IP and AI. As we've heard, Anthropic and OpenAI have accused some Chinese companies of distilling their AI technology, trying to essentially take a shortcut to be able to develop their own models.

8:41We talked about Moonshot a lot over the past few days. Moonshot is one of those companies that have been others also, DeepSeek, that's been accused of distillation. So Besson, and I think the U.S. side, wants to raise this question and try to set some boundaries about what's appropriate and what's not. The AI industry is evolving so quickly at this point, it's not clear how much distillation you can do. Many companies, American AI companies, also do distillation themselves to check their models and verify. One of the big questions is how much can you do without crossing that line into what's considered either illegal or untoward in the industry?

9:18Ed Ludlow:Bloomberg Tech Executive Editor Peter Elstrom, thank you very much. Let's stay with China and AI. Joining us now is Stephanie Aliaga, J.P. Morgan Asset Management, global market strategist, who writes, China's AI models remain a wild card. And I was so interested to read your research and how you model this for the markets. It's one of the factors of recent days has been us re-evaluating the economics of particularly open source or open weight models. Outline your point of view on that. Yeah, absolutely. I mean, these AI model labs, you know, these are native AI services. They're quite new in nature.

9:54You know, our clarity around what the margins of using these models is still quite new. And I think there is a little bit of a misunderstanding around how cheap these AI models are. They may be cheap to call, so the API price is quite low, but to use them, especially if you're using something that is very compute intensive, right, that's very capable, and in order to achieve that, it still requires a lot of compute, you still need a lot of those resources. You'll need clusters of GPUs. You'll need significant electricity. So this isn't something that you can just run on your laptop. You still need that infrastructure around it.

10:33And we think the U.S. model labs are still well positioned to provide that. But there's just more that needs to be done around model routing, right, for the more efficient queries and getting them to where they should be.

10:46Ed Ludlow:Friday, chip stocks close in a bear market, in part because of the concern that more efficient AI means less need for hardware. By Monday morning, we'd reverse back to the assumption that there's demand there. So there will be greater hardware demand on the inference side. Where have we netted out, do you think? I think recent years have given us so much evidence to Jevin's paradox that as we get more efficient at using AI, at calling on AI, at routing our queries to more efficient models, the overall pie of AI demand continues to expand. And where we are today is that overall pie of demand is growing way faster than we can reliably service.

11:28So I think this emphasis on efficiency right now is very needed, but it's a bullish indicator, I think, on overall AI infrastructure, not a bearish one, because it's going to help the economics of all of this really work for more parties involved.

11:43Ed Ludlow:Stephanie, there's a critical report on the Bloomberg about China making intervention in the stock market following heavy selling in Chinese technology stocks. It makes the point in this report that this is probably the broadest effort in quite some time that China, its regulators, state-backed investors, etc., have moved to shore up confidence. Again, global market strategists' view, how do you interpret that? What does it mean for markets going forward? Investors should not forget the lessons learned in 2022, that Chinese equities are still very, very susceptible to sudden changes on the policy front from their government and so forth.

12:26It's the reason why we've been more cautious on Chinese equities, particularly in this AI wave. But we continue to still have a lot of confidence in the broader EM Asia story, but really around whether companies or compute providers are helping service some of that AI demand that is growing in China.

12:44Ed Ludlow:We are right on the verge of a massive earnings season. every earnings season is big, but it starts in earnest with Alphabet from the hyperscaler perspective, Tesla. What are the learnings going to be in the next few weeks for you? What will you pay most closely to on the corporate side? Absolutely. It's been a really volatile last three to four weeks. But we think that volatility has actually been healthy for the market, the digestion of what had been some meteoric gains in parts of the AI ecosystem. And the pressure on valuations has also left things in healthier levels. I mean, the valuations for the hyperscalers right now, that have been singled out, all of this caution and concern around ROI, well, their valuations are trading at some of the lowest levels that we've seen since early 2023, before the AI boom was really even priced into markets.

13:39So I think positioning is a lot healthier today. And I think that we may be near the inflection point when it comes to upward guidance on free cash flow margins. You know, yes, CapEx has grown to a level that is likely consuming 100 % of these companies operating cash flows, but their operating cash flows are nearing$1 trillion. And we do think that we're going to see a deceleration in this pace of CapEx growth into next year, and particularly 2028. and what our investment teams have really underscored to us in their conversations with many of these companies in Silicon Valley is that across so many different cylinders, AI demand is still very red hot.

14:20So we feel comfortable that these companies will be able to monetize AI and after this period of digestion in the markets, maybe investors will feel more ready to reward them for it.

14:31Ed Ludlow:Stephanie Aliaga of JPMorgan Asset Management, thank you very much. Now, coming up, a judge just halted the highly awaited merger between Paramount and Warner Brothers Discovery. Joining us next, Jessica Malujin, director of the Center for Technology and Innovation with the Competitive Enterprise Institute. Everyone thought this deal might close Wednesday after European regulators make a decision on it. Then a U.S. judge came in. We need to go over it. This is Bloomberg Tech.

15:05With the highest number of young STEM graduates per capita in the EU, Ireland has the people and skills your company needs to succeed here. IDA Ireland, the National Investment Development Agency, can help you find and nurture the people you need to internationalise and thrive. Our talent is just one of the extraordinary benefits Ireland has to offer. Learn more at idaireland.com. Invest in extraordinary. This is Ashley Akinetti from the Ben and Ashley I Almost Famous podcast. Ben, I've been reflecting a lot lately on how many big life moments our listeners have experienced with us. Yeah, it's pretty wild.

15:47They've watched us go from dating shows to engagements, weddings, babies, buying homes.

15:52Ed Ludlow:Every stage comes with new decisions. The milestones are the fun part, but the planning is where things get real. That's why Klarna stands out. Klarna gives people smarter tools to help manage everyday money with flexible ways to pay, cash back that builds automatically, everything in one place, and support that makes everyday decisions feel less overwhelming. Since life moves so fast, the right tools can help you feel more confident. Download the Klarna app today or visit Klarna.com to learn more. California resident loans made or arranged pursuant to a California finance law license. NMLS number 1353190.

16:31Clown a balanced account required to be eligible for cashback points. Limitations, terms, and conditions apply. Get the news you need in just 15 minutes. Start your day with Bloomberg Daybreak, the podcast with a global view on the stories that matter. I'm Nathan Hager. And I'm Karen Moscow. Join us each morning for curated stories on current events, politics, business, and foreign relations. Plus one conversation on the day's biggest developments, all in just 15 minutes. Subscribe to Bloomberg Daybreak for a precise, thoughtful take on the stories that matter. Listen to Bloomberg Daybreak each morning on Apple, Spotify, or anywhere you listen.

17:07Ed Ludlow:Just a day before the Paramount Warner Brothers Discovery deal was set to close, a federal judge ordered a pause on the merger over potential antitrust violations. Now the companies need to wait two more weeks as a judge deliberates if the deal is legal. Joining us now, Bloomberg's Lucas Shaw, who leads the team at Screen Time. And this was the biggest deal for the industry that we cover at Screen Time, right? What do we know the basics at this point about the timeline and procedure of what will happen? But also, you must be getting a sense of how terminal or serious this is being treated as. Yeah, well, what we know is that we have a couple weeks before the judge is going to make a ruling as to whether we will delay the deal even more, right?

17:56This is a temporary restraining order. They could decide in a couple weeks, you know, go ahead, go for it. Or based on a lot of people are reading into the language of the ruling this week that they may say, actually, we're going to have to pause on this because this case from the states has a reasonable chance of winning. We don't know which way they're going to go. obviously Paramount is hoping that they can kind of move ahead. They, I think, were hopeful that there is going to be an update in two weeks. But also you see in their language, you know, they have gotten less confrontational and dismissive because now they realize they're in a real legal fight and they need to win.

18:33Right.

18:34Ed Ludlow:The judges scheduled a hearing for August 3rd, but that hearing is to decide whether to extend the hold for longer. What are the concerns here, Lucas, that, you know, we're quite far down the road at this point on this deal. If there are some antitrust concerns, were they known prior to today or this is something that has sort of been ruminated on for a while? There were definitely antitrust concerns all along, you know, as much as, you know, there were concerns about Netflix back when they were going to buy Warner Brothers Discovery because of how much bigger it is and them becoming even more dominant in streaming.

19:07By some measurements, Paramount and Warner Brothers is sort of the more logical antitrust case because you're combining two companies that are very similar, right? So the primary cases that the states have outlined are around movies and cable networks, that if you combine two big movie studios, they have too much clout in that business. You know, they would be 25 to 35 % of the film business every year. And that cable networks, they'd have an even larger share. Now, obviously, Paramount feels that these are too narrow of definitions. Cable networks in particular doesn't account for streaming, but Paramount was also the one that was previously defining the market as just streaming.

19:42And there are real stakes in this for Paramount, not just because they want to get the deal through, but if they are delayed longer, given the nature of what they agreed to, they owe a ticking fee. And so they will owe millions of dollars a day if this deal passes beyond the end of September.

19:58Ed Ludlow:Bloomberg's Lucas Shaw, who leads our coverage of the media and entertainment industry. Thank you very much indeed. Let's stick with the Paramount-Warner Brothers Discovery merger. That is now being paused. Joining us now is Jessica Lemelugin, director of the Center for Technology and Innovation with the Competitive Enterprise Institute. And again, to start your reaction to a federal judge putting a hold in place. I think it's sort of the latest chapter in all of this being, unfortunately, a political football a little bit. Going back to the Netflix attempt to purchase the studio, we saw that there was a lot of reporting about the White House being involved.

20:36and there was things beyond just what we should be focused on, which is what does this mean for consumers and competition? That's the job of these agencies to review these mergers. But there's always been some chatter in the background that there was more going on around it than just that. And we sort of see that it's the next chapter of that, right? It's not coincidentally, I don't think, 12 Democrat AGs, most of them running for office in the fall who are challenging this. That's an unusual move, usually. I mean, California sometimes can surprise you. But usually once the feds have cleared something, the states go along with it.

21:12That's not what we see here. There's also been reporting that the DOJ cleared this without the final recommendation of the staff investigating it. That's a little unusual. And they also issued a letter that in some ways some have noticed anticipates the AG arguments.

Read the full transcript

21:29Ed Ludlow:I'm sorry to interrupt you. The president of the United States is speaking at the White House. that was president trump speaking in a bilap with the president of lebanon stocks technology stocks in particular at session highs a lot of discussion on the situation of the war in iran trump saying that there is no interest in meeting with iran and holding talks for peace or a ceasefire if they are not going to take it seriously or until they're ready he said that iran wants to meet, but that this time the U.S. has no interest and that they will continue deploying military might to pare back Iran. A lot of questions focused on also what's happening with the Strait of Hormuz.

22:20Ed Ludlow:And let's get a recap. Bloomberg's Washington correspondent, Tyler Kendall, out in D.C. for us. wide ranging, lots of questions, but Iran are focused. What do we need to know? Well, Ed, you really hit one of the main headlines there. President Trump saying that Iran is desperate to meet in his words, but that the U.S. doesn't have any interest until Iran is willing to meet in what he called a meaningful way. This is Bloomberg News confirmed earlier today that Iranian officials are in Pakistan for talks with mediators. Now, he was also asked about a headline that we have been tracking very closely here, which is that threat from the Iranian-backed Houthi rebels to institute immediately a maritime blockade against Saudi Arabia.

23:03President Trump saying that if it happens, we will take care of it. Also, for context, we had confirmed earlier today that this potential threat is extending to any vessels going to and from Saudi ports, threatening potentially the 4.5 million barrels per day that Saudi Arabia exports via the Red Sea. President Trump also went on to confirm that Israel is currently in the process of withdrawing troops from southern Lebanon in a test of the truce between Israel and Lebanon in Israel's fight against the Iranian-backed proxy group Hezbollah. This move comes as you're seeing President Trump is meeting with the Lebanon president, Ed, marking the first time that we have seen a meeting between the two leaders of this country in nearly 20 years.

23:48Ed Ludlow:That's Tyler Kendall in Washington, D.C., summarizing the president's comments. Thank you very much. This is what markets look like. Again, technology stocks pushing to session highs as the president was speaking. The Nasdaq 100, a gain of almost 2 % on track for its biggest jump in about three weeks. The SOX similarly just coming off its session high of 5%, but on track for its biggest jump in about a month. But, you know, it's been a lot of swings and moves. NVIDIA is up 1.6%. At the top of the show, we brought you some news based on a Bloomberg News interview with NVIDIA personnel that they are very much on track with Vera Rubin.

24:23Ed Ludlow:Their latest generation system, full production is underway. They've already actually shipped some Vera Rubin-based platforms to some of the biggest AI companies where workloads are expected to run imminently. We'll keep tracking it. The AI trade is evolving. Investors are looking beyond the biggest winners and asking where the next wave of profits will come from. Joanne Feeney, who advises capital management, says the recent fallback on AI-related stocks is an opportunity that investors should focus on the companies with durable competitive advantages. Joanne, it's great to have you back on the program.

24:56Ed Ludlow:You know, the situation's moving fast and changing quickly, right? Chip stocks closed in a bear market Friday night on the idea that better, more efficient AI, less demand for hardware. By the time I came into work Monday morning, we were back to where we were prior to that. There's lots of demand for hardware. Try and explain it to me. Yeah, Ed, you know, you and I have been talking about this for a long time. And right now, I think investors are starting maybe to parse through where the strength in competitive advantages actually lie. You know, first, obviously, it's the suppliers to the data center build out to the enabling of these AI models like NVIDIA, like Broadcom, both of which we've owned for a long time for clients.

25:38But now I think people are trying to think about, okay, what about those model companies? Because although most of them are not public, we're starting to see more and more of them. We saw a couple of announcements in the last week of new, very high-powered models out of China. We're seeing the price of the implementation of AI in terms of the token costs falling. And it just seems like there's a concern that the model makers are going to struggle to be able to continue to differentiate what they're offering. And that suggests profit margins there are going to be challenged and that you'd better look elsewhere.

26:12And that's why we like the chip suppliers still and the data centers, because they have other things that they're offering besides just running these models. The security, the stability, the reliability, the data handling, all of that gives them, we think, a more durable moat. And that would sustain our profit margins.

26:32Ed Ludlow:what one interpretation post Kimi K3 release was it's a 2.8 trillion parameter model that's priced three dollars per million tokens on the input side 15 dollars per million on the output side and that would basically be an indication that the Chinese AI companies feel good about where they're pricing it they see demand and when we enter the inference phase there'll be even more demand for nvidia-based or gpu-based platforms is that an argument that that tracks for you Yeah, so I think this competition story among the models is going to take some time to play out. I think you'll see that gap shrinking over time as more and more models come out with better and better capabilities.

27:14But, you know, the demand will rise, yes, because of lowering costs on the inference side. And that only then spurs more demand for more data centers, for more NVIDIA chips, more Broadcom design chips, et cetera. So, you know, you just have to think about where the profit margins are going to be maintained and where they're going to start to shrink over time.

27:33Ed Ludlow:Is it healthy to sort of go from a situation where you enter a bear market to and there have been big swings in the socks? And I'll remind the audience of your many years covering the semiconductor industry. But, you know, it's amazing how quickly things return to normal, so to speak. You know, I think one of the areas to really be a little bit worried about is the memory space. and I think they're so-called priced cheaply, like Micron at whatever it is, six times future earnings. Bring it up. There's a reason for that. And the reason is that those profits, those sales numbers have in large part been driven higher by massive price increases, hundreds of percent.

28:14That's not going to be sustainable for forever. And at some point, we're going to get more supply. And even with some of the innovations we're seeing in high bandwidth memory, We're going to see those prices come down as more supply comes on. Why? Because you have three players at the leading edge. You have some at the trailing edge. They're all building more capacity. And this is how the memory world works. Some people are saying, oh, this is different. We're no longer in a cyclical industry. I just disagree. The game theory behind when these guys add capacity and how much they add has always pointed to them collectively adding too much capacity.

28:50It's in their interest to build more capacity than their current market share justifies because they want to gain market share. And so we're going to see those prices come down at some point. And we always see the prices of the stocks come down well ahead of that. So I think that's a particularly dangerous area. I don't know for how long this current positive momentum is going to last. But I also don't think it's a problem, Ed, that we saw a little bit of a pullback. Investors probably had become overly concentrated. We are pretty active in making sure our investors, our clients are not overweight, any one name.

29:23And so you see some diversification, which is simply smart. And it does lead to these kinds of periods of pullbacks of volatility.

29:30Ed Ludlow:Every day you want to be able to answer the question, why? Why is something happening? We're showing Micron's up almost 11%, which is only its biggest gain in about three weeks. But you're right, trades at 6.5 times forward 12-month earnings. Why? What is going on in the here and now? Well, it's hard to know. I mean, look, stock prices are noisy, right? I spent a lot of time studying asset pricing back when I was an economist. And we learned looking at the data back then that you don't look at any one day. You don't even look at a week. You don't look at a month. You've got to look at the long term.

30:06And what we've been focused on is how are earnings driving the fundamental valuation of a company? What kind of cash flow can they generate over time? And so these periods where you have these big run-ups and run-back downs are just noise. And so we really focus on long-term investing for our clients and looking at the fundamentals. Ultimately, that is what supports a company's valuation. And so for Micron, there is a period ahead of them where they're going to continue to have really strong pricing, where they're going to continue to have strong profit growth. It's just that you don't know when we're going to get the news that, oh, this is how much new capacity is coming online from all three of these guys.

30:47And so, therefore, this is how much prices are going to fall. It's not that their earnings stop growing. It's that their earnings growth actually at some point probably turns negative because if you get a collapse in prices, then, you know, you have a period of negative growth. And that's happened in Micron's history in the past. And so that's why it is trading where it is. And it also is why I think it's a little bit of a risky place to go.

31:08Ed Ludlow:Joanne Feeney of Advisors Capital, back on Bloomberg Tech. Thank you very much. Some news. Anthropic has reached a landmark$1.5 billion settlement in a copyright lawsuit over the use of millions of pirated books to train its AI models. The deal, which received final court approval yesterday, is the largest copyright class action settlement in U.S. history and could set the stage for similar cases against OpenAI, Meta and Google. and we will track what happens in that space. Now, coming up, we're going to take a look at where venture capital is finding the biggest opportunities as Moonshot's Kimmy rattles the AI industry.

31:47Ed Ludlow:Joining us next, Samir Dallacchio of Bessma Venture Partners back on Bloomberg Tech. Stay with us. This is Bloomberg Tech.

31:58This is Ashley Akinetti from the Ben and Ashley I Almost Famous podcast. Ben, I've been reflecting a lot lately on how many big life moments our listeners have experienced with us. Yeah, it's pretty wild. They've watched us go from dating shows to engagements,

32:13Ed Ludlow:weddings, babies, buying homes. Every stage comes with new decisions. The milestones are the fun part, but the planning is where things get real. That's why Klarna stands out. Klarna gives people smarter tools to help manage everyday money with flexible ways to pay, cash back that builds automatically, everything in one place and support that makes everyday decisions feel less overwhelming. Since life moves so fast, the right tools can help you feel more confident. Download the Klarna app today or visit Klarna.com to learn more. California resident loans made or arranged pursuant to a California finance law license and MLS number 1353190.

32:55Klarna balance account required to be eligible for cashback points. Limitations, terms and conditions apply. The Bloomberg Sustainable Business Summit returns to Singapore on July 22nd. Our fifth annual Asia-Pacific Summit will explore how business and finance leaders are shaping the next phase of globalization by strengthening resilience and driving a multi-speed energy transition across Asia's diverse markets. Join us for solutions-driven discussions and networking opportunities. Thank you to our summit advisor, Bangkok Bank. Learn more at BloombergLive.com slash SBS dash Singapore.

33:30Ed Ludlow:China's Moonshot AI is preparing to begin discussions in August on a final round of fundraising before listing in Hong Kong at a valuation of as much as$50 billion. Sources say the Chinese AI lab should close a financing round that began this summer at a valuation of around$31.5 billion and then intends to open discussions with potential backers about a very quick follow-on capital raise. Of course, it's been in the news cycle constantly for four days. The AI conversation has shifted yet again after the release of Moonshot's Kimmy K3. Treasury Secretary Scott Besson says the U.S. will scrutinize Chinese AI models for evidence of intellectual property theft.

34:10Ed Ludlow:Against that backdrop, where is venture capital seeing things? Where are the biggest opportunities? Are we worried? Joining us now is Samir Dallaki, a partner on the Growth Investment Practice team at Bessama Venture Partners. There's so much in your portfolio that's interesting and highly relevant to this. But could I just start by asking what your net takeaway from everything Kimmy K3's been? All-consuming, Kimmy K3, in the news. And I think it's a great thing. I really do. I think we have a moment in time where I'm a free trade, competition is good for us kind of a guy. I believe that the right tool for the job is the right answer for our enterprises, for customers.

34:53and they will use Frontier Lab models like Anthropic for the stuff that they need to use that for and they will use open source models from China and other places in the world over time. We need our open source US and Western-based open source models, open weight models to get better and bigger and that's great and that will happen if we have free competition with the Chinese models. But that's my general take.

35:17Ed Ludlow:I'm very positive on the development. Let me ask you one sort of final specific thing on Kimi K3. Bessemer is an investor in Anthropic. One of the questions posed was, okay, if a startup of Moonshot's size and scale and resources can put out a 2.8 trillion parameter model, price it such that it did, what's the moat of an Anthropic or an open AI? Why do we value them such as we do? Your industry colleague, Son Monamedi, from Greylock was on the show on Friday, and he basically said, well, Anthropic's got a real business with multiple streams of revenue. It's completely different. But where would you put your response to that question?

35:57Well, because we're big investors in Anthropic and they're in the quiet period, I can't comment specifically on them. But I will say about the frontier models generally, what I would say about their modes, they're still at the frontier. They're still extending the lead. We're in a stage of recursive learning improvements to the models. The models are being released faster and faster as we get ever better capabilities. I believe the frontier will continue to stay at the frontier. And you'll have others that will catch up. Now, there's also a lot of talk on X right now around whether there should be, Secretary Besson mentioned, should we be going after IP theft and distillation?

36:38And how do we feel about that? I'm a free markets guy, but I don't believe in IP theft and distillation. So I do hope that we will figure out ways to better protect the IP of the closed models on the frontier. But that would be my take on that.

36:53Ed Ludlow:Samir, it's been way too long since you've been on Bloomberg Tech. So take us inside the portfolio company's perspective on that. You know, you have concern about IP theft on behalf. How would that impact if there were to be regulation or there were to be an action protecting those companies that you're backing and investing in? Yeah, I think at the frontier, you know, it's important that they put up defenses. I think it's important that there are protections for IP. I think that we'll see those model providers are going to continue to explode in growth. I keep trying to explain to my colleagues in the market just to get a sense of the size of this market, Ed, is astonishing.

37:37This is like having the opportunity to invest in Rockefeller's Standard Oil in 1870. The market for intelligence is the largest market. I used to say in technology, it's the largest market in human history. And so the frontier model, the penetration in enterprises is so tiny relative to what it will be that the frontier models will continue to excel. But the open source ecosystem is going to continue to flourish. it should, companies like Fireworks, we'll talk about, are going to benefit from that.

38:10Ed Ludlow:The big news of late, so Fireworks took ARR from 100 million to 1 billion, but they did it in less than a year and a half. And you're, with Reset, somebody with experience in software, you're basically saying we've not seen that growth trajectory before. It's extraordinary. And I used to be the CEO of a SaaS company. I took it public in 2017. To build our SaaS company to IPO scale took us eight years. These guys are adding 100 million of ARR per month. That took us years and years and years. Do you know what's interesting about that? Sorry to interrupt you. Those guys at startups like Fireworks, they're good at software, right?

38:47Ed Ludlow:They're good at AI. Are they good at go-to-market? Do you know what I mean? So they're able to do that 100 million of ARR per month. Have they been able to keep up with themselves on the sales channels and having good operational leaders in that business? Yeah. Well, I'll tell you, it's one of the biggest reasons we invested behind Fireworks, and we're so excited to back them. The CEO, Lynn, is an incredible strategist, a deep technologist, came from Meta, was leading the team from Meta that built PyTorch, which is sort of the leading open source framework for machine learning. But she's recently brought in George Hu as president.

39:24And George and I got to work together at Twilio after they acquired my last company, SendGrid. and I will tell you without hesitation, George is one of the absolute best go-to-market leaders, software executives in the industry. He was the COO at Salesforce, then the COO at Twilio and so the companies that are scaling at this pace and have the wisdom to bring in some folks that have seen scale and growth and can put in process and operational rigor, we think are going to be real winners in this market and that's why we think Fireworks

39:55Ed Ludlow:will be the winner in their market. to lots of people, AI is still quite abstract. It can mean lots of different things. So, you know, at Bessma, you guys basically say you invest across the stack, the model layer, the infrastructure layer, the vertical application layer, you know, but those are all quite different things. Like Lease AI is a high-flying portfolio company of yours. But they all have slightly different factors behind them. They do, and you have to look at each of the markets, and every business is different. But what's extraordinary is just the demand and the pull from the market for all of them.

40:29You're just seeing exponential growth, growth curves like we have never seen. That's what they have in common. That's what they have in common. I've been in software now for 31 years. I've never seen a lease AI selling AI to property management companies. They've posted publicly. They crossed 200 million of ARR and are still doubling. The companies in software history that reached what I would call triple-digit growth at triple-digit scale, those triple-triples, that's extraordinary. Any company that did that in the prior era went on to become household names. And we believe we're investing in those kinds of companies.

41:08Lisa is doing it in property management. Lagora and Evenup are doing it in the legal field. Bridge is doing it in healthcare. It's just an extraordinary group of businesses.

41:20Ed Ludlow:Samir Dallakia, partner at Bessemer Venture Partners, back on Bloomberg Tech. Thank you very much indeed. Coming up very quickly, Tesla is trailing in its AI spending compared to its big tech peers. And that's worrying investors. They want to see the EV maker put its money where its mouth is. We'll get the preview next. This is Bloomberg Tech.

41:45Ed Ludlow:As big tech rushes to pour trillions into AI investments, Tesla is lagging behind. Despite setting lofty goals, the company just spent just$2.5 billion of the$25 billion in 2026 capex. And investors are growing a little bit concerned. Joining us, Bloomberg's Jordan Fitzgerald. This is the mindset going into Tesla's earnings this week. They want to see Tesla spend money. They do. This is a bit of a show me the money situation here. In prior years, Tesla spent about a little under$10 billion annually in CapEx, which is in line with more traditional automakers like Ford and GM. But as Tesla rebrands as not a car company, but an AI company, the market really wants to see it spend like one.

42:30Ed Ludlow:2026 capital expenditures, we should point out, at$25 billion are way beyond anything Tesla's even attempted before. What's the metric that we judge then? We just want to know that spending in this quarter is tracking ahead of that, in line with that. Yeah, we just want to see progress here. If Tesla's not spending, then investors have reason to doubt that they're making progress on their robo-taxis and robots and other AI-linked projects. And given that Tesla's valuation is priced as if it's a growth stock, an AI stock, that it really needs to start delivering on those AI benchmarks. There's also added pressure here since SpaceX went public.

43:08Like now investors, if they're not happy with Tesla, they have another way into Elon Musk. And so Tesla really, really needs to start delivering on those promises or at least indicating that it's on the way there.

43:21Ed Ludlow:Jordan, really quick, TSR, when are Tesla earnings? What else are we looking for? Tesla earnings tomorrow after the bell. I'm sure you and I will both be head down on that. Other than that, commentary from Elon Musk is always the most important thing in any Tesla print. But the past few quarters, we've seen the print come out and the shares initially rise post-market. And then Musk starts talking on the call and we see a little bit of a dip. Things really, really move off of his tone, off of the words he uses. There's a lot that's not in the direct report that we can get a hint on from Musk. And I'm sure many are hoping for any sort of color he can give on the potential for a merger.

43:59Ed Ludlow:I am braced. I am going to participate. And hopefully my voice will last that long. Blumos Jordan Vichderald, thank you very much indeed. Indeed, that does it for this edition of Bloomberg Tech. Again, kind of tech stocks at session highs or just off them after the President of the United States was speaking. But generally, the rebound in chip stocks continues to gather momentum. Some really top conversations across public and private markets today. Recap them on the podcast. You know where to find it. It's on the Bloomberg Terminal as well as online on Apple, Spotify and iHeart. This is Bloomberg Tech.

44:37Healthcare doesn't always work great. If you've ever waited on a refill or couldn't schedule an appointment, you get it. That's the kind of stuff Optum is changing. They're using data and technology to integrate patient care, pharmacy, and everything else. So healthcare is connected, not complicated. What's that look like? Cheaper prescriptions that are easier to get and care that looks at the whole person how you need it. Optum is helping make healthcare work as one for everyone. Learn more at business.optum.com.

45:33in Lacqua on Bloomberg Television or wherever you get your podcasts.

From the publisher

Bloomberg's Ed Ludlow breaks down Nvidia's latest AI chip rollout, as the company says its next-generation processors are now shipping to customers and entering full production. Plus, a judge hits pause on Paramount's merger with Warner Bros. Discovery, throwing the blockbuster media deal into limbo. And, Sameer Dholakia of Bessemer Venture Partners joins to discuss the AI race's next phase, with Washington promising to scrutinize Chinese models as investors hunt for the next big winners.

See omnystudio.com/listener for privacy information.

More from Bloomberg Tech

All 343 episodes
Nvidia's AI Push, Paramount's Merger Pause & China's AI RaceBloomberg Tech · 39 min
Listen in VO