RiskReversal Pod

Jensen Huang's resignation and AI agent disruption reshape tech valuations and ad models

1h 10m Listen to the original
NvidiaJensen HuangAI AgentsDigital AdvertisingSearchTech ValuationsCorporate GovernanceAnthropic

Summary

Dan Nathan and Guy Adami discuss Jensen Huang's resignation from Nvidia, citing espionage concerns, margin pressure, and customer concentration risks. Joe Marchese joins to argue that AI agents will disrupt traditional search and advertising by prioritizing fiduciary duty over ad revenue. The speakers contrast the resilience of physical experience-based businesses with the vulnerability of digital 'toll takers.' They also analyze the competitive positioning of major tech giants like Microsoft, Google, and Meta in the emerging AI landscape.

Key takeaways

  1. 1
    Dan Nathan points out that Nvidia's gross margins have been in the mid-70s, significantly higher than the typical 50-55% for hardware companies, implying extreme pricing power that is now under threat as major customers like Google and Amazon develop custom silicon.

    Dan Nathan · 11:26

Episode sections

  1. 0:00 – 3:15

    Sponsor Reads and Intro

    Guy Adami reads ads for CME Group futures, Fidelity Trader Plus, and Apex Fintech Solutions. Dan Nathan introduces the episode topic: Jensen Huang's resignation from Nvidia and the broader AI ecosystem.

  2. 3:15 – 9:47

    Jensen Huang's Leadership and Espionage Concerns

    Dan Nathan and Guy Adami discuss Jensen Huang's promotional style and the 'cult of personality' surrounding him. They cite Gary Marcus's critique that Huang lobbied for Chinese AI sales, creating espionage vulnerabilities and conflict of interest issues.

  3. 9:47 – 13:01

    Resignation, Interim CEO, and Legal Risks

    Dan Nathan notes Huang's resignation and the appointment of CFO Colette Kress as interim CEO. They discuss the potential for class action lawsuits, congressional investigations, and the liability Kress faces given her 13-year tenure signing financial reports.

  4. 13:01 – 16:27

    Nvidia's Valuation Discount and Margin Pressure

    Guy Adami argues that Nvidia's stock was trading at a discount (14-15x forward P/E) before the news, suggesting the market already priced in risks. Dan Nathan highlights customer concentration (Microsoft, Amazon, Supermicro) and the pressure on Nvidia's 70%+ gross margins as customers diversify to custom silicon.

  5. 16:27 – 21:44

    Singapore Sales Mystery and Megaspeed

    Dan Nathan reads from a Wall Street Journal article stating Nvidia's Singapore sales (22% of revenue) disappeared from disclosures after Megaspeed, a company linked to Huang's partner, was placed on an internal block list. Guy Adami notes the Commerce Department was investigating Megaspeed before the probe was halted.

  6. 21:44 – 24:33

    Ecosystem Impact and Neo-Clouds

    Dan Nathan discusses the 'neo-cloud' ecosystem, where companies like Nebius buy GPUs from Nvidia to rent to major tech firms. He questions the sustainability of this circular financing if the new leadership pulls back on support strategies.

  7. 24:33 – 27:41

    Market Outlook and Dot-Com Comparison

    Dan Nathan compares the potential fallout to the post-dot-com fraud era (Enron, WorldCom), suggesting the current AI trade is a mashup of dot-com and financial crisis dynamics. He notes Nvidia's stock has fallen 30% from highs but remains volatile.

  8. 27:41 – 34:13

    Joe Marchese Interview and Book Promotion

    Dan Nathan introduces Joe Marchese, author of 'Human.ai', and promotes a giveaway. He highlights the book's endorsements and its focus on the attention economy and AI's role in monetizing user attention.

  9. 34:13 – 38:00

    AI Agents vs. Traditional Search and Feeds

    Dan Nathan and Joe Marchese discuss how AI-powered search and feeds are disrupting traditional advertising models. Marchese argues that AI represents a shift from AI being used on people (ads) to AI being used by people (agents), which could displace current ad-based revenue streams for companies like Google and Meta.

  10. 38:00 – 40:00

    Fiduciary Duty and the Anthropic Advantage

    Marchese explains the concept of fiduciary duty in AI agents, contrasting them with traditional search engines that may be biased by ad revenue. He highlights Anthropic's promise of no ads as a potential competitive advantage, creating a 'prisoner's dilemma' for competitors like OpenAI who have ad businesses.

  11. 40:00 – 44:00

    Amazon's Blocking of AI Agents

    The speakers discuss Amazon's decision to block Anthropic's Muse agent, contrasting it with Walmart and Target's openness. Marchese suggests this is a short-term defensive move by Amazon to protect its ad profits, but notes that agents can bypass large platforms by accessing smaller retailers and Shopify stores directly.

  12. 44:00 – 46:30

    Publishing Industry and Content Creation Models

    Marchese warns that AI agents scraping content without compensating publishers could destroy the business model for digital publishing. He cites Reddit's move to require login and charge for API access as an example of publishers trying to protect their data, noting that without a new model, there will be no incentive for humans to create review content.

  13. 46:30 – 49:00

    Travel and 'Toll Taker' Disruption

    The discussion turns to travel intermediaries like Expedia and United Airlines. Marchese argues that AI agents will disrupt these 'toll takers' by optimizing for the best price and experience, breaking the 'breakage' model where humans pay more due to laziness or lack of comparison. He suggests agents will act as tireless personal shoppers.

  14. 49:00 – 1:10:19

    Future Business Models: Subscriptions vs. Ads

    Marchese proposes that AI agents might be paid via subscription or a cut of the savings they generate for users, rather than through ads. He compares this to Amazon Prime, where users pay for convenience and savings. He believes this model could eventually replace extractive advertising if agents deliver tangible value.

Assets discussed

AssetKindStanceContext
NVIDIANVDAequityneutralEverything that's going on. in and around NVIDIA and your main man, Jensen Wong.
Meta PlatformsMETAequityneutralGoogle, Meta, Amazon, they're obvious ones, right?
AmazonAMZNequityneutralGoogle, Meta, Amazon, they're obvious ones, right?
AlphabetGOOGLequityneutralLook at Google trading down to half trillion dollars in a day.
AppleAAPLequityneutralI mean, look, Apple already did this. Like Apple could you could buy all your Apple products online, but have you been in an Apple store recently?
MicrosoftMSFTequityneutralTesla out of there because that that is not particularly interesting to me. Microsoft, uh, and am I missing one?
NETFLIX INCNFLXequityneutralAnd we all talk about Netflix and the

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Jensen Huang resigned from Nvidia, encouraged by the board, and Colette Kress was named interim CEO.

Unidentified speaker (UNKNOWN) ·9:47

How to verify: Nvidia press release or SEC Form 8-K filing for resignation and officer appointment. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Colette Kress has been CFO of Nvidia for 13 years and signed off on every quarterly report and financial statement.

Dan Nathan ·10:07

How to verify: Nvidia annual proxy statement (DEF 14A) or 10-K filing listing executive compensation and tenure. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Gary Marcus stated that Jensen Huang 'enabled and even lobbied for Chinese AI' and 'subverted US policy'.

Dan Nathan ·6:06

How to verify: Gary Marcus's Substack post dated around the episode release (Nov 13, 2026) or public statements. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Nvidia's Singapore sales, previously 22% of revenue, disappeared from financial disclosures after Megaspeed was placed on an internal block list.

Dan Nathan ·16:27

How to verify: Wall Street Journal article cited by Dan Nathan and Nvidia's 10-K/10-Q filings showing geographic revenue breakdown changes. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Megaspeed was set up by Jensen Huang's partner and was under investigation by the Commerce Department before the probe was halted.

Guy Adami ·17:59

How to verify: Commerce Department press releases or SEC filings related to export control violations and Megaspeed. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Nvidia's gross margins have been in the mid-70s, compared to a typical 50-55% for hardware companies.

Dan Nathan ·11:26

How to verify: Nvidia quarterly earnings reports (10-Q) showing gross margin percentages. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Nvidia was expected to earn ~$9 per share for fiscal year 2027, with revenue expected to be nearly $400 billion.

Dan Nathan ·19:22

How to verify: Consensus analyst estimates from Bloomberg or FactSet for Nvidia FY2027 earnings and revenue. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Nvidia's stock was trading at 14-15x forward P/E prior to the resignation news.

Guy Adami ·13:21

How to verify: Historical stock price data and forward earnings estimates from financial data providers. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Microsoft accounts for 22% of Nvidia's revenue, Amazon 14%, Supermicro 10%, Alphabet 5%, and Meta 4%.

Dan Nathan ·21:44

How to verify: Nvidia 10-K filing disclosing customer concentration and revenue by customer. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

The seminal AI paper was titled 'Attention Is All You Need'.

Dan Nathan ·32:17

How to verify: Publication of the paper 'Attention Is All You Need' by Vaswani et al. on arXiv. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Google's market cap dropped to half a trillion dollars in a single day.

Unidentified speaker (speaker_2) ·34:47

How to verify: Check Google (Alphabet Inc.) stock price history and market capitalization data for the date referenced in the episode (Nov 2026) or historical data if referring to a past event mentioned. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Anthropic is about to go public and has promised no ads.

Unidentified speaker (UNKNOWN) ·39:05

How to verify: Check SEC filings for Anthropic's S-1 registration statement or official press releases regarding their IPO plans and ad policies. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Reddit required users to log in and charged for API access to protect content.

Unidentified speaker (speaker_2) ·45:52

How to verify: Check Reddit's official blog posts or API documentation updates from late 2026/2026 regarding API pricing and login requirements. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Meta's stock was down 30% from its all-time high earlier in the year.

Dan Nathan ·56:22

How to verify: Check Meta Platforms (META) stock price history for the period leading up to November 2026 to verify the percentage decline from its peak. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Anthropic's CEO is Dario Amodei.

Unidentified speaker (UNKNOWN) ·39:05

How to verify: Verify Anthropic's leadership team on their official website or SEC filings. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

OpenAI's 'Attention is all you need' paper was written by Google engineers in 2017.

Unidentified speaker (UNKNOWN) ·57:36

How to verify: Check the author affiliations on the arXiv paper 'Attention Is All You Need' (arXiv:1706.03762) and Google's corporate history or press releases regarding employee departures to OpenAI in 2026-2026. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

YouTube is the biggest streamer by far, larger than Netflix.

Unidentified speaker (speaker_2) ·58:00

How to verify: Compare monthly active user (MAU) or watch time data from independent analytics firms like eMarketer, Sensor Tower, or Statista for YouTube vs. Netflix. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Google did not launch the 'Attention is all you need' paper because its ad business lacked financial incentives to enter the AI arms race at that time.

Unidentified speaker (UNKNOWN) ·57:36

How to verify: Review Google's public statements and product launches from 2026-2026 regarding AI and machine learning, and compare with internal memos or industry analysis from that period. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

The speaker served on the board of Clear Channel for seven years.

Unidentified speaker (speaker_2) ·1:01:11

How to verify: Check Clear Channel Holdings' SEC filings (DEF 14A or 10-K) for board member lists and tenure during the relevant period, or verify via the company's investor relations history. For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Clear Channel was taken private by Mubasta.

Unidentified speaker (speaker_2) ·1:01:11

How to verify: Search SEC filings for Clear Channel Holdings' delisting or private equity buyout announcements, or news archives for 'Mubadala Clear Channel acquisition' (note: ASR likely misspelled Mubadala as Mubasta). For recent events, verify 2026-08-13 through 2026-11-13. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Review Nvidia's Q3 2024 10-K and 10-Q filings for revenue recognition policies and customer concentration disclosures.
  • Verify Gary Marcus's public statements and any available evidence regarding lobbying activities for Chinese AI sales.
  • Check the Wall Street Journal article referenced regarding Megaspeed and Nvidia's internal block list status.
  • Analyze Nvidia's gross margin trends against historical hardware company averages (50-55%) to assess pricing power erosion.
  • Evaluate Anthropic's 'Claude' and 'Muse' product roadmaps for explicit fiduciary duty commitments versus ad-supported models.
  • Research Amazon's policy changes regarding AI agents like Anthropic's Muse and their impact on third-party developer access.
  • Monitor class action lawsuit filings against Nvidia following the leadership change.
  • Assess the revenue mix of major tech companies (Google, Meta, Amazon) to determine exposure to AI-driven ad displacement.
Full transcriptexpand
0:00
Guy Adami: Balance your trading strategy by adding futures. CME Group helps you manage risk and capture opportunities in all market environments. Capitalize on around-the-clock access to highly liquid global futures and options market across all major asset classes. Just visit your online broker and get started. Plug into
0:20
Guy Adami: valuable educational materials and trading tools and see what adding futures can do for you at CME cmegroup.com Group.com.
0:27
Guy Adami: slash on the tape.
0:31
Guy Adami: This episode of the Risk Reversal Podcast is presented by Fidelity Investments. When timing is everything, you need a platform that can meet the moment. The all new Fidelity Trader Plus can help you stay close to the trade with powerful tools and charts that show up consistently across all your devices. So you can build a trade at home get
0:51
Guy Adami: real-time data from your phone, and complete a saved order from your laptop without starting over. Fidelity Trader Plus can help you make smarter, faster decisions wherever you trade with free third-party single score analyst ratings and custom alerts. Try it today at Fidelity.com/slash
1:11
Guy Adami: Trader Plus. When you open your investing app and hit buy, do you ever wonder what actually happens next? Behind that tap on your screen is a chain of moving parts, clearing, settlement, and custody that has to work flawlessly every single time. There is a good chance that's Apex
1:31
Guy Adami: Fintech Solutions working behind the scenes with millions and millions of brokerage accounts across 200 plus investment platforms. Apex is one of the largest and most trusted custodians in the business, providing the cloud native wealth tech that modern investing runs on. Visit ApexFintechSolutions.com.
1:52
Dan Nathan: Welcome to the Risk Reversal Podcast. I am Dan Nathan. That is Guy Adami.
1:57
Guy Adami: Hello, Dan.
1:57
Dan Nathan: Yeah, well, we have a very special podcast for you. We're not going gonna to kind of waste any time. Everything that's going on. in and around NVIDIA and your main man, Jensen Wong. So we're gonna break it down. We might have a few of hot takes. After that conversation, we have one of our very good friends, Joe Marcazi. He is the general and build partner at Human
2:17
Dan Nathan: Ventures. And he wrote a book, and
2:19
Unidentified speaker (UNKNOWN): we're gonna
2:19
Dan Nathan: give a bunch of them. Yep, we're gonna Going give to a bunch of them away. Human.ai,
2:24
Dan Nathan: how the fight for your attention became the world's biggest business. And I think a lot of what Joe wrote in this book is really applicable to what's going on in and around AI. As a lot of folks think about how they're going to monetize these what will be trillions of dollars investments. The old model was advertising. And the question is, what is the new model?
2:44
Dan Nathan: We know a lot of focus has been on enterprise and coding and a whole host of other applications, but Joe's gonna work through all of that. So stick around for that conversation. All right, guy. Um, and I know that you don't like to give yourself any credit for much. What is your line? What do you say that you um
3:04
Dan Nathan: individually take uh responsibility for bad calls, but you collectively like to take uh, you know, not that you take that sort of uh praise for? that.
3:15
Unidentified speaker (UNKNOWN): I live
3:16
Guy Adami: by the edict of I was wrong. We, the collective, were right.
3:21
Dan Nathan: Yeah. Well, I've been sitting next to you at Fast Money for nearly 17 years, and I've heard you say that. And I obvious obviously uh think that you've been very, very humble. So um, again, you know what? You've had a beat on Jensen Wong, and I think a lot of our regular viewers, a lot of our regular listeners, they know it. And it's not that you have basically said the guy's a
3:41
Dan Nathan: bullshit CEO or what they're selling is not real. It's more about the way in which he has been promotional um over the last call it nearly five years. Just kind of spell that out a little bit, what your kind of thought process has been and why your antennas have been up for so long.
3:59
Guy Adami: Arguably Argued with the most important company um over the last decade. and then you can make an argument that if that's the case, it might be the most important company of all time in terms of the scope and in terms of the market cap at its peak. So let's just get that out there. But there have been a lot of important companies over the decades, and quite frankly, many of those companies you wouldn't be able to
4:19
Guy Adami: identify the CEO uh in a lineup. And for better or for worse, I think what's happened over the last call of ten years, there's been this cult of personality thing that's taken over, and people become uh seduced by the celebrity that that it's created around it. And some people are and
4:39
Guy Adami: some people are not. And my concern around him has been it appeared as though he was one of those people that was going down that road. The leather jackets obviously the signing of women's chests, which we've seen pictures of, obviously. You know, this is sort of Mick Jagger like stuff that has historically just been under the realm of rock
4:59
Guy Adami: stars. And, you know, maybe CEOs are the new rock stars. I don't know. But when you go down that road has been my point, you sometimes take your eye off the ball and you sometimes lose the rigor that got you there in the first place. And I'm not comparing him to Sam Bankman Bankman-Fried Fried in terms of some of the corruption around that or Elizabeth Holmes, but
5:18
Guy Adami: there's an eerie similarity between the three of them in terms of the levels that society and the media took them to.
5:26
Unidentified speaker (UNKNOWN): Yeah,
5:27
Dan Nathan: and I think the media part is really important because when you think about you know how CEO like this, how a company like this kind of works into a sort of lore. Listen, in the market that we have been in for the last five years, the investment cycles or the investment capital that we have seen, the creative nature of the capital raising, the size of
5:47
Dan Nathan: some of these IPOs that are coming to market that could have only been made possible by the technology, right, that NVIDIA has sold. And he has obviously been a huge part of that. I just want to read a quote from a friend of ours, Gary Marcus, Marcus on AI. is his Substack. He's been on our podcast. He's been on Fast Money a
6:06
Dan Nathan: bunch. And you know, I think this is really important. He starts out today's Substack by saying he's met Jensen. He's clearly a great CEO, all that sort of stuff. But here's the quote: he enabled and even lobbied for Chinese AI, subverted US policy, carelessly left himself vulnerable to both blackmail and espionage.
6:26
Dan Nathan: His blindness to conflict of interest or the appearance thereof cost many people dearly. He acted to the peril of his company, his industry, and his country. He also left NVIDIA open to endless litigation and investigation, which is perhaps part of why they encouraged him to go, despite his enormous talents. And I think, listen,
6:47
Dan Nathan: that is really in a nutshell how a lot of folks are going to. end up feeling about this. But all those people in the media, in the tech ecosystem, in the analyst community, in the banking community who've just glazed this guy repeatedly, okay, for five years, paying little attention to what has gone on in his
7:06
Dan Nathan: lobbying effort to be able to sell the world's best technology despite the national security implications of. the two-way battle that's going on between us and China, has always been somewhat curious to us and very few others, though, that were willing to ask the hard questions. He's been lobbying the
7:26
Dan Nathan: administration to sell his chips to China. He thinks, and he's had this very unique view, that the world should be building their AI on his technology, regardless of the ability for China or our major adversaries, to keep pace with us. So curiously, this is something that very few people push back, even within the
7:46
Dan Nathan: administration. Once he got Commerce Secretary Lutnick to agree to a whole host of things, Trump came in and said, fine, we are going to lift certain export controls as long as we take a VIG as a country. Okay. But I guess the few questions that have been asked. about why he was so intent on doing this.
8:06
Dan Nathan: And that I guess is where this whole espionage honeypot situation comes in. And to Gary's point, leaving himself and the company vulnerable to what will be endless sort of inquiries about this company and how they operate.
8:21
Unidentified speaker (UNKNOWN): Yeah,
8:21
Guy Adami: right. And you're spot on. And if you think about it, I mean he's not alone. Tim Cook had lobbied to be able to do right. So there are a lot of CEOs that have come to the administration uh to try to get a stri a stronger hold into the Chinese markets for obvious reasons given the size of it, right? And then given the obvious revenue stream that it would create for them.
8:41
Guy Adami: So it's not unprecedented, but the questions that you raise are the right questions. Did he do it because he wanted those markets, or did he do it because somehow he was being coerced? Or, you know, he was under the auspices or under the thumb of this espionage and he knew he was compromised. So, and those will all those
9:01
Guy Adami: things at some point will come out. And he created this situation. People talking about it didn't create it. He created it by himself. And Then there's going gonna to be this sort of Damocles for a period of time. You mentioned by the way, these class actions. action suits. I think there's an inevitability to that, especially if the stock continues to deteriorate. We can talk about the stock in
9:21
Guy Adami: a minute, but if you were to see a further decline, you can almost rest assured that there will be some class action shareholder lawsuit around exactly that. So it may have sort of um in the short term uh c capitulation in the stock, we may have seen.
9:39
Guy Adami: but longer term, I think there are many more chapters to this story. Whether or not that influences the stock or not yet is remains to be seen.
9:47
Unidentified speaker (UNKNOWN): Yeah,
9:47
Dan Nathan: and you and I are recording this about one thirty Thursday. And I guarantee you by the time you're listening to this, there's probably gonna be a whole host of other uh major announcements. So he resigns this morning at the, you know, the the encouragement of the board. They name as an interim CEO. the CFO, Colette Cress of the last 13 years. I
10:07
Dan Nathan: think a lot of folks, just the way they felt about him being a CEO, feel the same way about her as a CFO. But here's the thing: the woman has had to sign off on every quarterly report, every financial statement that this company has issued in her 13 years as the CFO. If there is some sort of cover-up,
10:27
Dan Nathan: if there was some sort of, you know, um, I guess, endeavors to circumvent, right, a lot of our export controls. And he was in some way, shape, or form conflicted, right? Then there's a lot of liability. And I wouldn't be only worried about the class action suits. I mean, we just saw a
10:46
Dan Nathan: sweep, right, of Congress. We just are now in a lame. duck session of a president, right? So what sort of investigations? This is going to go on and on. In the past, when we've seen this sort of anal probing by a government, okay, it really causes companies for the most part to take their eye off the ball.
11:06
Dan Nathan: Let's say you bring in an outsider to run this company. Well, that individual is going to have no interest in being involved. right? In all of the questions about this company before all this, but the circular financing situation. It is worth noting, and you've made this point again and again. The company, a
11:26
Dan Nathan: hardware company like this that normally has, let's call it 50% gross margins, 55% gross margins, to be generous, that has been enjoying mid-70s mid 70s gross margins means that they have been taking a huge bite. out of their customers, right? Their customers have been concentrated. Their customers have gone way out of their way to
11:46
Dan Nathan: diversify away from NVIDIA, whether it's Google with their TPUs, whether it's Amazon with training. I mean, the list goes on with custom silicon, right? All that sort of stuff. And you say to yourself, this coming out now couldn't be worse. Their They're customers who have gone, you know, from spending, let's say, very, you know, these were
12:05
Dan Nathan: software companies. They had a lot of leverage. They were like basically, you know, from a capex standpoint, they were not asset heavy. They've gone to negative free cash flow, most of them, other than Microsoft, to support their CapEx at the expense of NVIDIA's ability to charge for it, as we just said. So at this point now, who has the
12:25
Dan Nathan: pricing pressure? Where are those margins going? Look at NVIDIA, which makes boatloads, right? And earnings and cash flow. They have an amazing balance sheet with $100 billion in cash, less than $40 billion in debt. So you tell me what the investment looks like from here on out. And again, you've been highlighting the fact on fast money
12:45
Dan Nathan: on here. Why is this country or why is this company, and I say country, it basically is post-nation state. Why is it trading well below a market multiple when they have so much pricing power and you know that sort of thing? So again, break that down for me.
13:01
Guy Adami: Yeah. Well, I mean, obviously, listen, I understand why the optimists would look at their valuation prior to the last week or so when it was trading around 220 and sort of languishing, trading it, you know, at some point, I think we were trying to talk about talking a fourteen. 14 and a half or 15 price to forward earnings, and you say this doesn't make any sense given their growth rate,
13:21
Guy Adami: given the margins they enjoy, given their EPS growth rate, a number of different things. And the optimists and the bulls were saying this stock is way too cheap, it should be significantly higher. I understand that argument. My counter-argument that nobody wanted to seem to make, and I'm not just labeling myself, I'm sure there are a lot of people out there, maybe there's a reason. why the company was trading at a
13:40
Guy Adami: discount. And maybe you should take a look through that lens and wonder, okay, why isn't the market rewarding them? And I I'm not suggesting it's on the back of what's transpired over the last couple days, but it was clearly on the back of something. So now we're back down here and with levels we last saw, I believe in the spring or so, or maybe even longer,
14:00
Guy Adami: you know, farther back than that. As a matter of fact, we just traded down to levels that we saw in January of 2025 when the stock at the time made an all-time high before trading down to about 90 bucks a share in the spring. So here we are. What does it all mean? Well, again, to use the term, they're going to be under the microscope for quite some time. And
14:20
Guy Adami: whoever comes in to run this place is going gonna to look under every single floorboard just in case there's something that they miss. So the level of scrutiny around this company is be significant. And you know, for you Shakespeare fans out there, we are a nation that loves to create um giants and loves to create icons, and we're equally love to
14:40
Guy Adami: knock them down. Well, we created one, now we're in the knockdown phase, and the knockdown phase can be pretty messy, and it's not gonna going to just be one person going down with a ship, they're gonna going to be others. So the chapters, there are many chapters left. There are many congressional hearings yet to be had. There are many people, I think, in the
14:59
Guy Adami: administration that will either be summarily fired or will step down over the next couple of weeks. Again, what that means to the market, I don't know, but let's put it this way it's certainly not bullish.
15:09
Dan Nathan: Yeah, well, you know, if you're talking about Shakespeare, maybe that's also Taylor Swift, the great lyric that she has, they build you up so they can tear you down. I'm just saying.
15:19
Guy Adami: So you're telling me that this Taylor Swift, who's supposed to be a generational talent, basically just steals from Shakespeare.
15:27
Dan Nathan: No, that's not what I'm saying. I'm saying she iterates, guy. And every artist iterates on something that came before them. Um, but Shakespeare died penniless, and Taylor Swift is probably a multi multi-billionaire. billionaire. So good on you, Taylor. Swift. Um, I I want to bring this to, you know, listen, we can talk about the way in which this company has run their business. Okay.
15:47
Dan Nathan: We can talk about him lobbying, you know, in his reasons for to sell chips to China. And some of them make perfect sense in a way. He's like, if they're gonna build this own AI, okay, to compete with ours in a sort of digital belt and road, we might as well have them. on our technology to understand the playing flield, right? The software,
16:07
Dan Nathan: this CUDA software ecosystem that they've created. I get it. Okay. And there's others who feel the exact wrong way. The problem here is that there appears to be, you know, some fraud. I mean, for all intents and purposes. There was an article in the Wall Street Journal today, guy. Nvidia's NVIDIA is disappearing, Singapore sales deep in mystery around mega Megaspeed. speed.
16:27
Dan Nathan: Mega Speed is the company that was set up by this woman that he was having this affair with. She was targeted, right? And the whole idea was to get access to his product for a country that he was working for, despite, okay, the export controls for national security reasons. The article, this is, and you know, I want you to pine on this.
16:46
Dan Nathan: Here's a quote As recently as the summer of last year, Singapore was officially one of NVIDIA's biggest markets, generating 22% of the chipmakers' revenues. Then, practically overnight, it disappeared from the company's financial disclosures. Investors deserve to know why. And NVIDIA needs to start explaining with or without Jensen at the
17:06
Dan Nathan: helm. The timeline indicates that NVIDIA altered its methodology for reporting the geographic breakdown of its revenues shortly after placing Megaspeed on international
17:17
Dan Nathan: on an internal block list. It isn't clear if the two are related, but the shift um had the effect of removing visibility and visit uh in NVIDIA sales tied to Singapore.
17:28
Guy Adami: Are you uh are you familiar with the the the the actor Desi Arnaz by any chance?
17:34
Dan Nathan: Sure. Yes. Uh from the um I Love Lucy franchise of
17:37
Unidentified speaker (UNKNOWN): the 19th 1950s. century.
17:40
Guy Adami: Loved it. Four people, same apartment. Brilliant, actually. But he he would famously say, Lucy, you have some splaining to do. And that was his sort of broken English. And I think that's where we find ourselves. There's some splaining that needs to be done. And in terms of this mega speed, which the commerce department, this is all documented, was
17:59
Guy Adami: investigating an unknown an unnamed um
18:03
Guy Adami: personnel from NVIDIA apparently went to commerce and sort of had it kiboshed. So that investigation went away. One has to wonder where was the FBI in all of this? Was the FBI involved or were they told to stand down for whatever reason? So at some point, I think Cash Patel has some answering to do. Clearly, Howard Lutnick has some
18:23
Guy Adami: answering to do. Um, And to your point about it mysteriously falling away. yeah, there are a lot of questions to be asked. And, you know, in terms of the revenue it created and generated, you know, what's behind the curtain and whole thing? And nobody knows the answers. I tell you who does know the answer to these things, it's Jensen Wong. And he will at some
18:43
Guy Adami: point, whether it's by subpoena or he goes up uh on his own accord, he's gonna have to answer a lot of difficult questions. And not because there's a bullseye here, it's because of the national security implications around the company that he founded. Front full stop. That's it. If it was a hamburger company, I don't think anybody would care.
19:02
Dan Nathan: I would care. I love hamburgers. Um here here, let's talk about the stock. Let's talk about the ecosystem. Let's talk about the expected IPO of anthropic in OpenAI, why this is so important. Okay, and then we'll kind of close out. We'll get to uh my main man uh Joe Marcese. For this current fiscal year 2027, okay, this company,
19:22
Dan Nathan: NVIDIA, was expected to earn nine dollars in uh nine dollars ish in earnings. This is a company back in 2022 that was earning 40 cents in earnings, okay. From a revenue standpoint, this year they're expected to do nearly 400 billion dollars in revenue, okay. Going back to 2022, this is the year that
19:41
Dan Nathan: ChatGPT came out, they did uh $27 billion in revenue. So this year, expectations of 90% earnings and sales growth decelerating to 70% next year. Truly astounding. This is a company, again, that's supposed to go from $400 billion in revenue fiscal this year,
20:01
Dan Nathan: basically the fiscal 20%. 27 or excuse me, 28 at nearly $700 billion in revenue. Okay. 70% plus gross margins. Trading this year at nearly 18 times earnings. Next year, 10 and a half times earnings. Okay. Insane, right? You'd say this is an unusual
20:21
Dan Nathan: value. And my question to you guy is why does it trade at 11 times 10? at a
20:27
Unidentified speaker (UNKNOWN): time? Well, and
20:28
Guy Adami: that's a question we were asking a few months ago, and a question that w did not need to be answered because of the optimism around the stock. In retrospect, I think it was the right question to ask. And now everybody is seemingly jumping on board. Well, maybe there was a reason it was trading at a discount, not only to the broader market,
20:47
Guy Adami: but a significant discount to its peers. So there will be a point in time where the underlying business of NVIDIA is such that the evaluation is too compelling to pass up, regardless of all the noise around it and surrounding it, and regardless of whoever steps in at the helm. I think that that's a fair statement to
21:07
Guy Adami: make. I think the reality is there was inevitability to their margin starting to decline prior to this. happening. I think what we've seen over the last couple days is that inevitability, the timeline has been going to be sped up in a meaningful way.
21:24
Dan Nathan: Yeah. Um, and just again at the opening, I think the stock was down like 15% Thursday morning. The stock was down 40% in less a little more than a month from its all-time highs, right? It's recovered a bit, it's down only 5% right now, um, which plays. which places it down 32% or something like that from the all-time highs. Um, on the year, it is
21:44
Dan Nathan: down 11%. Pretty um shocking. Let's just look at the customer concentration as we've talked about. Okay. So as a percent of revenues, Microsoft 22%, Amazon 14%, Supermicro, that makes the boxes um about 10%, Alphabet 5%, Meta about 4%. So you you can see what's going on. here, right? This is the
22:04
Dan Nathan: entire ecosystem. All of those folks, for the most part, like we said, are trying to diversify away. Now, maybe that is some of the sort of reason why you see that sort of discount. On the analyst front, okay, you basically have, let me just pull it up here from my fact set That um machine guy, Dami. You have 84%
22:24
Dan Nathan: of the analysts who cover the stock rated a buy. 14% hold, and there's only one sell. Okay. So we know that basically the sentiment towards the stock has been off the charts. But the fact that you've seen their major customers catch a bid guy in this period, it it I think it's misplaced to some degree. You know, the
22:44
Dan Nathan: the idea that they're going to be able to have some pricing power at a time where it's going to be okay. what are we doing here? Is it the circular financing that's been propping it up? The Neo clouds, and you know what they are. I mean, Nebius is down 15% today. These are stocks that have gotten absolutely killed. Some of these neo clouds basically exist
23:04
Dan Nathan: to buy GPUs from NVIDIA, okay, put them into data centers and rent them out to NVIDIA's major customers, right? Who are already building their own infrastructure. So it's so at some point, if NVIDIA has been the one doling out, let's say tens of billions of dollars to support the ecosystem, and the new CEO, let's forget the
23:24
Dan Nathan: CFO interim, comes in and pulls back on that whole strategy because they don't want to be building on the strategy that's been in place for five years in case there's some fugazi going on there, right? Like, does that all make sense? So then the question is: what happens? to the ecosystem, especially in front of Anthropic, where NVIDIA was meant to be ten billion dollars
23:43
Dan Nathan: of a seventy or eighty billion dollar capital rate.
23:46
Guy Adami: Yeah, long conversation, but I'll try to give a short answer. I mean, I think in terms of what's happening with these other stocks, this is the shell game uh period of of the of the market where money's obviously coming out of one name and and trying to find a home somewhere else, as people say, well, it's gonna be it's gonna be basically binary, right? Somebody's gonna win and
24:06
Guy Adami: somebody's gonna lose. So obviously, right now NVIDIA's losing, and now they're putting their chips on who they think is gonna win. And in the short term, that could be the right strategy. But I think to your point, in the longer term, I think it's the wrong, I think they're looking at it the wrong way because if the structure of this trade is cracking, which is something we said prior to all this news coming out, and something I
24:26
Guy Adami: still believe.
24:27
Guy Adami: then they're not going to be anybody left. There will be more losers than winners, let's put it that way.
24:33
Dan Nathan: Yeah. And um, you know, the stocks, which obviously, you know, NVIDIA has been a huge driver of, again, because of a lot of the ecosystem in general. We've seen bids to a lot of the competitors. Um, you know, it's down 30% from the all-time highs uh made in June. I don't think either one of us are saying run for the hills and sell all your cents. semis and, you know, that sort of thing, because again, a lot
24:53
Dan Nathan: of this stuff has gotten pretty compressed, and video's volume over the last week or week and a half has been out of control. You might ask yourself, um, who's left uh to sell it? But but again, I just feel like if this has been a big driver of the entire narrative and this is going to be basically pulled back to some degree, then you have to ask yourself, what are the major catalysts? If the major
25:13
Dan Nathan: catalysts are huge capital. raises from anthropic and OpenAI, I'm not so sure those are the drivers of the market from here on out. So again, I mean, you know, you and I, we didn't come on today to kind of say things are about to get crazy with the stock market, with the stocks, with this name individually, because a lot of
25:33
Dan Nathan: damage has been done, even if it wasn't done because of this one thing. But if you would had ask asked me a year, or two ago, when the whole market was basically driven by NVIDIA's performance, in my opinion, dragged so many other parts of technology and then outside of technology up because people were so geeked up about the numbers that they were printing and the guidance that they
25:53
Dan Nathan: were giving. Well, then there has to be a discount on that. I don't know where that discount bottoms out. But to some degree, I also would say if there is fraud here, if there is some sort of national security implications, some cover up in this and that, whatever, this is going to make the post dot com fraud, whether it was, you know,
26:13
Dan Nathan: global crossing and and worldcom and Enron and all that sort of stuff, it's going to make it look like a walk in the park. That's my only point. And you know, that was the last period where we had a protracted bear market. People have been, you know, comparing. to what's been going on at AI to the dot-com. You and I have basically said this is going to be a mashup
26:33
Dan Nathan: when it comes unwound of the dot-com and the financial crisis. And we haven't even talked about the financing in and around it outside of the circular financing about NVIDIA, but we know that Jensen has been in the middle of this consortium of bankers and and the like, willing to backstop, willing to basically create. asset classes in and around their
26:53
Dan Nathan: main product, which are GPUs, which depreciate, which are not something that is in, you know, perpetuity. And so there's a lot more chapters to be written about this. So again, um, you and I, your antennas in particular, have been up about this guy in general. The fact that he is the guy that is gonna be the uh basically on
27:12
Dan Nathan: the hook for all of this, I think is. pretty unique. And I think this is probably, I don't know, first or second inning.
27:20
Unidentified speaker (UNKNOWN): Well,
27:20
Guy Adami: we're gonna find out. And it it is I I don't think enough people are doing the deep dive as to what it means around around the edges of a trade that's clearly had fundamentals in place, but more the optimism around it is really what's driving it. And you see how quickly optimism turns to pessimism, and I think we're in that portion. of the uh uh trade
27:40
Guy Adami: right now.
27:41
Dan Nathan: Guy, what is this thing? You see what this this this is? You see how it's like it's kind of like a rectangle and you're
27:46
Unidentified speaker (UNKNOWN): human.ai.
27:47
Guy Adami: It's a book.
27:48
Dan Nathan: It's a book. You know who else has been writing a book for years is Guy Christopher Adami. Um and so Guy maybe just needs a little bit of an impetus, and maybe our friend Joe Marcese, Marchese, who just wrote this book, which we are gonna give away a hundred of, maybe more. You want more than that? We're gonna give away. We're gonna tell you how to do that. But on the back of this book, and if I am not
28:08
Dan Nathan: writing one of these little blurbs on your book, okay, right now he's got, look at this. He's got Scott Galloway, Esther Perel, Derek Thompson, Reed Hoffman, Matthew Prince. I mean, come on, they're all endorsing this book. Gayadami and Dan Nathan are endorsing it. Stick around for my conversation with our. dear friend Joe
28:27
Dan Nathan: Marcese. We're going to talk about some of the stuff in the book. We're going to talk about how you can get one. Uh, and we appreciate you being here. Thanks so
28:33
Unidentified speaker (UNKNOWN): much.
28:38
Guy Adami: Balance your trading strategy by adding futures. CME Group helps you manage risk and capture opportunities in all market environments. Capitalize on around the clock access to highly liquid global futures and options. market across all major asset classes. Just visit your online broker and get started. Plug into
28:57
Guy Adami: valuable educational materials and trading tools and see what adding futures can do for you at cmegroup.com/slash on the tape.
29:08
Unidentified speaker (speaker_3): This episode of the Risk Reversal Podcast is presented by Fidelity Investments and the all-new Fidelity Trader Plugin. Plus platform. Want to access your preferred trading tools, charts, and saved orders from anywhere without interruptions? Try Fidelity Trader Plus, Fidelity's most powerful trading experience yet. With seamless connectivity from your desktop to your
29:28
Unidentified speaker (speaker_3): laptop to your mobile device, you can make smarter, faster trade decisions when it matters with Fidelity Trader Plus and get free third-party analyst ratings and custom alerts. Unlock your edge. with the next generation of advanced trading at Fidelity.com/slash Trader Plus.
29:44
Guy Adami: This risk reversal podcast is sponsored by Apex FinTech Solutions. If you've ever bought a stock on your phone, there's a good chance Apex made that possible. Apex provides the tech-forward AI-powered Powered custody clearing and wealth tech infrastructure that powers modern investing. Discover the company driving the digital
30:04
Guy Adami: wealth transformation. Learn more at apexfintechsolutions.com.
30:12
Unidentified speaker (speaker_2): Welcome back to Risk Reversal Podcast. I'm still Dan Nathan. That is my great friend Joe Marcese. We already teased him. We teased the book. We got a lot to go in through. Um, Joe, welcome to the podcast. Thank you for having me. Yeah, not just one of my favorite humans in the world, but also you make one of my favorite luxury tequilas in the world. That would be Como's tequila. I think a lot of our listeners, a lot of our viewers already know that.
30:32
Dan Nathan: Well, just to be clear, I don't know how to make anything.
30:34
Unidentified speaker (UNKNOWN): No, our main man, Richard Bad.
30:36
Unidentified speaker (UNKNOWN): Yes.
30:37
Unidentified speaker (speaker_2): Richard's been on the pod too. All right. You are here for multiple reasons. Um, you've been on the pod a bunch over the years. As I said to Guy right before we went on here, you've been one of our great, great partners, great sponsors uh of us. So we really appreciate you being here. Primarily you're here because you're having a bit of a whirlwind. This is this is the book. It's human.ai is D-O-T.
30:57
Unidentified speaker (speaker_2): How about that? Beautiful cover there. And it's very digestible, by the way. And it's one of those books that I think is important for our listeners, our viewers who span from private market operators, obviously just warriors about markets, public markets. I think there's so much here, especially given what we just have seen in in. Q3 earnings and Google's another one. And
31:17
Unidentified speaker (speaker_2): we're going to get into that. But let's talk about the book, why you wrote it. And by the way, we're going to give away at least a hundred. So uh all you got to do is go to your favorite podcast store, look at Market Matrix. That's M-R-K-T matrix. Subscribe, leave a comment, and then send a screenshot to Amanda at riskreversal.com. RiskRiversal.com. She's going to send you a link to buy it. And then when you get the book
31:37
Unidentified speaker (speaker_2): and you read it, leave a review. on Goodreads, on Amazon, wherever. Okay. Because we want to help our main man get this out to the masses. Um, all right. Why did you write the book? And when did you start reading the book? What was the original reason for it? And how has that changed over the last couple of years?
31:53
Unidentified speaker (UNKNOWN): Uh,
31:54
Dan Nathan: spite. I
31:55
Unidentified speaker (speaker_2): I
31:55
Unidentified speaker (UNKNOWN): really I wanted to write the book. Spite book?
31:57
Dan Nathan: Pure spite. Um So uh COVID had happened and we're post COVID. and social media is continuing to explode. We'll dis decide whether or not it really is social media, TikTok, meta, et cetera. Everything accelerated and everyone was talking about the attention economy. Everyone was saying, you know, as if we had just discovered the attention economy. And I was
32:17
Dan Nathan: uh, you know, complaining to my wife Christy, who you know well, who's an entrepreneur, and people who I'd worked with, like Jonathan Bing, my co co-author, author, that I'd been talking about the attention economy since two thousand 2007. seven. And it's driving the market. And I have the receipts. I've got articles from 2007. And they just, to shut me up, they said you should just write a book. And I was like, no, no, no, I'm not going to do that. And then people kept talking about the attention
32:37
Dan Nathan: economy. And then AI comes along, and the seminal paper that created AI, the paper was called Attention Is All You Need. And then it went even to a fever pitch of people realizing we all live in this attention economy. And so I wanted to write the book to talk. about one, how we got here and then where it goes next with AI.
32:54
Unidentified speaker (speaker_2): Yeah. And by the way, um, for our listener, our viewer, I used to like I kind of tweak you a little bit. Anytime I'd see something written that I was reading about markets, about anything, and it said the attention economy, and I would forward it to you and I'd say, Didn't you kind of uh just to piss you off? So I actually think that I do deserve some responsibility for maybe getting you to write the book. One one of the things I I just really I
33:14
Unidentified speaker (speaker_2): thought was really important, it kind of reinforces that. And Derek Thompson, who I think, as you know, you introduced me to him, brilliant, brilliant writer, thinker, plain English podcast. He writes his own Substacks, so go check it out. He used he was at the Atlantic for years, right? Yeah,
33:27
Unidentified speaker (UNKNOWN): he was at the Atlantic for a
33:28
Dan Nathan: long time. He is he is one of the clearest thinkers on on such a wide range of subjects. I highly recommend having
33:33
Unidentified speaker (UNKNOWN): to do that. Yeah,
33:34
Unidentified speaker (speaker_2): so so this is on the back of the book, and my little quote must have been lost in the
33:39
Unidentified speaker (speaker_2): Joe Marchese has been on the front lines of advertising revolution for decades, always seeing the next uh big thing and extraordinary clarity. And you know, I I that actually sums up exactly what I take away from the book. So let's talk about like what you think for our viewers, our listeners, what are the most important things if you're focused on public markets and you're looking at all the disruption
33:59
Unidentified speaker (speaker_2): from AI, how the models are changing, and your belief how the build-out gets kind of continued, I don't know, tailwinds, if you will, because I think there's a lot of overlap between your thoughts about the last 20 years in attention and what the next 20 years look like.
34:13
Dan Nathan: I think if you trade in the public markets and you do not understand how much of the public markets rely on advertising as their rest source of revenue, like
34:22
Unidentified speaker (UNKNOWN): uh
34:22
Dan Nathan: Google, Meta, Amazon, they're obvious ones, right? But everything else on
34:27
Unidentified speaker (speaker_2): like Walmart, Uber Eats, Instacart, they're all in the advertising business. A lot of them wouldn't be profitable without being in the advertising business. Why? Because attention has gotten even more valuable as content has gone up. So like if you trade the markets and you don't understand how advertising fits in, you're missing a huge piece. Look at Google trading down to half trillion dollars in a
34:47
Unidentified speaker (speaker_2): day. But then the knock-on knock on effects, everything's built on top of. search, right? There's only two, there's only two mechanisms in uh the internet that brings information to humans, search and feeds, right? Search is Google, and then it's also Amazon, and then it's also YouTube, but it's also any any website with a search box. Feeds are meta and TikTok.
35:06
Unidentified speaker (speaker_2): That's both of those things are actually AI getting us information, right? And the new wave of AI, if it's anything, it's better search. Google's mission was to organize the world's information and make it universally accessible and useful. Who would argue with me that AI results are not more useful results? AI is better search.
35:26
Unidentified speaker (speaker_2): That has an impact on Google. Google's going to be fine. They sell chips, they sell, they have Kawa, they have YouTube.
35:33
Unidentified speaker (UNKNOWN): But
35:33
Unidentified speaker (speaker_2): if you change search, everything downstream of search, it it it changes everything. So every public company that really relies on search. And so that's the second part. And then the third part is you can't understand what the business model of AI is going to be, right? Like the current business model is going to displace a lot of those advertising businesses I talked about. And we don't know what it's going to replace it with yet. And so like
35:53
Unidentified speaker (speaker_2): if you're trading markets and you don't understand that, you're missing out.
35:56
Dan Nathan: Yeah. I mean, some of the um early indications of that, you know, we've seen you just mentioned Google in that quarter they just reported and the slowdown. in digital ads. I mean, it was it was kind of a a pretty important sort of data point there because a lot of folks were looking two years ago and saying there's this existential risk, right? If a lot of if if people are basically displacing the way in
36:16
Dan Nathan: which they had searched the internet for the prior twenty five years or so, and now, you know, there's AI mode in Google. A lot of folks thought that that was going to cannibalize a good bit, and it has. It just started doing that. And the stock was in the penalty box for a very long time. You just mentioned the vertical integration of a Google, you know, with TPUs and you
36:35
Unidentified speaker (UNKNOWN): know all
36:35
Dan Nathan: and and again, they have the distribution. And so help me think about, you know, as far as distribution, because you just mentioned feeds. That's hugely important for what Meta
36:46
Unidentified speaker (UNKNOWN): has
36:46
Dan Nathan: been able to do. They were a little behind in their build out, but
36:50
Unidentified speaker (UNKNOWN): that
36:51
Dan Nathan: business is all about advertising
36:53
Unidentified speaker (UNKNOWN): and
36:53
Dan Nathan: the distribution of it. And they were early in monetizing it within their own business, how they served ads. Right. And that was helping them. But it seems like they've hit a bit of a wall. And maybe it's kind of Google-esque in in many ways, or no?
37:05
Unidentified speaker (speaker_2): It's different than Google because Meta, since it's a feed base, it's an entertainment product. Like uh uh Mark Zuckerberg would say this. Like it used to be a a social utility, a place for friends. Like that's why some people think we should take the social media moniker off of social media. So so there's a big difference between Meta and Google, which is meta monetizes not through understanding demand or or
37:25
Unidentified speaker (speaker_2): having the search box monetizes because I put in, I want to buy shoes, I want to go on a trip, I want to take Meta gleans that information from the things you engage with in your feed. So as long as you stay engaged in the feed, meta still has a very strong. ad business in its core ad business, which allows it to not give its Muse its AI products. It doesn't need to have an ad incentive right
37:45
Unidentified speaker (speaker_2): away. Google's diversification is different. It has YouTube, it has other businesses. But just keep in mind what we've been talking about search, feeds, the TikTok algorithm, that has always been AI, right? It's just been sorting through content to keep you engaged and make ad revenue.
38:00
Unidentified speaker (UNKNOWN): This
38:00
Unidentified speaker (speaker_2): new wave of AI is instead of AI being used on people, it's AI being used by people, right? So
38:06
Dan Nathan: that's the personal agent front.
38:08
Unidentified speaker (speaker_2): That's the personal agent front. And that's and like again, here's another one for for your listeners who understand like this term, fiduciary duty.
38:15
Unidentified speaker (UNKNOWN): There
38:15
Unidentified speaker (speaker_2): is a term, like people have been throwing around the word agent for AI, right?
38:19
Unidentified speaker (UNKNOWN): If
38:19
Unidentified speaker (speaker_2): you're an agent for someone, you have When a fiduciary duty. Like if I'm your talent agent, if I'm your real estate agent, I can't be getting bribed by the other side. on what to pick for you. If I start recommending a house for you because I'm getting a bigger kickback on that house than the other house, like that, there's a legal definition of that fiduciary. Right now, with agents, because they don't have ads
38:39
Unidentified speaker (speaker_2): inside of them and they shouldn't be swayed by it. You know what no human has done in the last five years? Gone to page three of the Google results or page four or page 10. ten. You know what?
38:48
Unidentified speaker (UNKNOWN): They
38:48
Dan Nathan: don't even get to the fourth link. I
38:50
Unidentified speaker (UNKNOWN): don't even know if they go. below the fold. Right. It's
38:52
Unidentified speaker (speaker_2): all that. But that's where the monetization occurs. You know what your agent will do? Go to page 100. It'll find a needle in a haystack theoretically for you, right? Because
39:01
Dan Nathan: of the fiduciary responsibility to deliver the best price of the
39:04
Unidentified speaker (speaker_2): Because it's tireless. Yeah.
39:05
Unidentified speaker (UNKNOWN): Right.
39:05
Unidentified speaker (speaker_2): It doesn't, it doesn't, it's not lazy. It's not swayed by like what's like it should give you the best utility. And as of right now it will. And there's a very big competitive dynamic. Anthropic's about to go public, right? Anthropic has promised no ads. If anthropic keeps it that way. So in the past, Corey Doctorow had this uh thesis of in shittification, meaning like you get people
39:25
Unidentified speaker (speaker_2): hooked on a platform, then you slowly make it shittier with ads. I tend to agree. That's how cable broke, but we'll come back to that.
39:32
Unidentified speaker (UNKNOWN): If
39:33
Unidentified speaker (speaker_2): anthropic stays with no ads, it creates a real prisoner's dilemma for the other guys. Because if you trust Anthropic's agent more than I trust OpenAI's agent, because OpenAI's agent might be going with the best ad rather than the best product for me. That's a big problem. That's a multi-trillion dollar problem in the markets. So I think that is a
39:53
Unidentified speaker (speaker_2): great opportunity for consumers because then we reset that whole ecosystem.
39:56
Dan Nathan: All right. So let's talk about an agent, the fiduciary responsibility. Um, you would think that if it's a it's a search as a relationship. it relates to something um that you want to purchase, right? Normally you would have gone directly to Amazon. Right. Normally, if you had searched Google though, you might get a top sponsored link,
40:13
Unidentified speaker (UNKNOWN): right,
40:13
Dan Nathan: from Amazon. So right now, Amazon's got this ad business. Uh
40:17
Unidentified speaker (UNKNOWN): I
40:17
Dan Nathan: know that's something you've been talking about for a while. I'm sure, you know, if you were sitting, you know, up there in uh, you know, Redmond, Washington, where are they? in Seattle, Washington, like seven years ago or something like that. you would have said, hey, listen, you guys are missing a huge opportunity, right? Um, so now Amazon wants to block, let's just, they're blocking Muse, right? And so the idea would be if you went directly to
40:37
Dan Nathan: Amazon, they're gonna have the sponsored sort of links. And and again, they're good at serving that. You already went there thinking about buying that thing. Now they have third-party third party sellers. How do you think about a huge platform like Amazon? blocking personal agents? And will they be able to do it? And what are the implications for their ad business right now?
40:57
Unidentified speaker (speaker_2): It could not be bigger. Look, again, Amazon is a diversified business, has web services, has, but advertising makes up a disproportionate amount of profits, right? And all those third parties you sell, like when you go in and you search for something and there's tons of results, like all you need is the best one. And that's what your agentic agents should do. It's why they blocked Muse to stop. start. I don't know where they stand today.
41:17
Unidentified speaker (speaker_2): But then they have a problem. Walmart has said, absolutely, Muse, come on in. Target. Like these places where they were at a disadvantage before in a web-only world are at a huge advantage with physical space. I mean, look, Apple already did this. Like Apple could you could buy all your Apple products online, but have you been in an Apple store recently? Like crowded,
41:36
Unidentified speaker (speaker_2): experiential physical space in the world. That is now an advantage for Target and Walmart as they think about those things. And then they let agents in. Now they do have advertising businesses, what they call retail media. And if you're you're looking at stocks and looking at the profits, like that's the line. So there is some risk there, but they'll trade that for better consumer experience.
41:56
Dan Nathan: So e commerce though, when you think about, you know, again, if you're searching for something, getting a contextual answer, again, does that turn everything upside down? Because that seems to be a little bit of what, or a lot of bit, of what Google's dealing with, right? But what about every sort of an Amazon has their own agent internally, right? And so, but the behavior still has to be that you go
42:16
Dan Nathan: to Amazon first and then you use that personal agent. So, how should we think about that as a consumer right now? You know, yes, there is Muse, and that's going gonna to be the one that actually is most widely disseminated. There's things like Instinct. I'm sure there's gonna be a bunch of others that pops out. But at the end of the day, duet is coming. OpenAI is has dots, not
42:35
Dan Nathan: widely distributed. Um, you know, Siri at some point will
42:39
Unidentified speaker (UNKNOWN): have
42:39
Dan Nathan: something, you know, that sort of sort
42:40
Unidentified speaker (UNKNOWN): of
42:40
Dan Nathan: works. So again, you know, how are you as a consumer gonna gravitate towards a personal agent? Will you have multiple personal agents?
42:49
Unidentified speaker (UNKNOWN): The
42:49
Dan Nathan: prisoner's dilemma that you just mentioned.
42:51
Dan Nathan: about the ones that are going to monetize and OpenAI has this ad business that is, you know, growing, I
42:57
Unidentified speaker (UNKNOWN): think,
42:57
Dan Nathan: uh pretty fast. I mean, it seems like we're about, I think 2027 is going to be the world where we're getting confused about which agents are going to do what for us, which ones are going
43:07
Unidentified speaker (UNKNOWN): to be
43:08
Dan Nathan: compromised in some way, shape, or form. And do they just get kind of blocked out to some degree of some of the largest platforms? Amazon being one of them?
43:16
Unidentified speaker (speaker_2): Right. Yes. I think that one
43:19
Unidentified speaker (speaker_2): switching cost between agents is very low right now. Right. And I think people will get comfortable with an agent. And then it will be if trust is broken. Like again, it's like you can think about it like real world. If your trust with your agent is broken, you're probably going to fire the agent and go get another one because they'll want your business. And these large platforms that block them are going gonna to have a problem because. their
43:39
Unidentified speaker (speaker_2): agent can go to a ton of small businesses. They can go to all the different Shopifies and they can solve your needs. Like they're the value of scale and aggregation doesn't exist if my agent doesn't mind going to 13 different checkouts and doing it for me. Like that was the beauty, was that there was one checkout in one place. So I I think we are in for a
43:59
Unidentified speaker (speaker_2): lot of disruption. I think that the the unanswered question though is. in the long run, our agent will do it for us. It's basically well machines working on our behalf. They'll skip through the advertising, they won't be swayed. But in the short run, what is the business model of the internet going to be? We don't if we're not sending traffic to commerce sites and we're not sending traffic to publishers who create content,
44:19
Unidentified speaker (speaker_2): and this is one I think people don't think enough about. Okay, right now, a AI has been trained on all the information in the world. Famously the there's a lawsuit happening with the New York Times. There's some other publishers have signed on, but that's already happened.
44:32
Unidentified speaker (UNKNOWN): But
44:32
Unidentified speaker (speaker_2): if those business models are under threat and no one's getting paid to make running shoe reviews and restaurant reviews, and how is the agents going to know what's good? How are they going to get information about what people like and don't like in the future? And so I think the idea that we are killing the publishing industry or seriously impairing it, like we already had Facebook zero, meaning no traffic. was going from Facebook to
44:52
Unidentified speaker (speaker_2): publishers. We're getting very close to Google zero because those AI overviews are stuck. No one's going to read the article on the site. The business model of those publishers is gonna be why should I publish just for AI to read it and make a recommendation? So that is something that we should be concerned about over the next six months.
45:09
Dan Nathan: Yeah, let's talk about publishing because again, that was one of the the I I think the earliest sort of you know, uh conflicts,
45:16
Unidentified speaker (UNKNOWN): if
45:16
Dan Nathan: you will, as soon as you know we started seeing ChatGPT, and then obviously Claude. You know, I go back 20 some years, and I think a lot of publishers, and you know, specifically news, had a really hard time figuring out what that business model was going to be going forward, right? You know, digital ads and then no paywall. You know, they they they took a lot of pain when they put the paywalls up, right? And
45:36
Dan Nathan: but they figured it out. I mean, the New York Times, I don't think they're digital. their distribution has ever been higher, right?
45:41
Unidentified speaker (UNKNOWN): So
45:42
Dan Nathan: how how do you think about that? Because if it really is something that's going to be a wall that a lot of these publishers, once again, you know, kind of like drive right into, like, how do they fix that?
45:52
Unidentified speaker (speaker_2): Again, public market. Look, look what happened. It right when everything was happening with Google with the stock price, Reddit had to turn on. You had to be logged in to to go.
46:01
Unidentified speaker (UNKNOWN): Last week
46:02
Dan Nathan: the stock went, but the stock went down 30% for them to kind of go out and say, hey, this is the new
46:06
Unidentified speaker (UNKNOWN): Because they're going to have to
46:07
Unidentified speaker (speaker_2): take short term pain. And now they've announced uh Reddit reviews. So now you just go to Reddit and search, right? And we've moved the search box from one giant search box that goes everywhere to now inside of our our garden where we have humans talking to each other about reviews. Because otherwise if you're not logged in, they don't know if you're a real person. You get spam, you get bots, you get AI
46:27
Unidentified speaker (speaker_2): slop, right? So I think they become more valuable in the long run, but in the short run, there's not a business model to support it. So like there's gonna be a gap. All
46:34
Dan Nathan: right, let's talk about toll takers, right? So, you know, Expedia is a great example. Yeah. Right. And so some of these stocks, as soon as like Muse came out, they just got destroyed for a
46:43
Unidentified speaker (UNKNOWN): little bit,
46:43
Dan Nathan: right? And and there was other business models and the financial services company. You know, in in Expedia, um, well, like yesterday, for instance, United. said they're they're blocking personal agents coming in. But again, it's not like they have an advertising business, they have a business of selling tickets, right? Expedia has an advertising business and they also get what fees, you know what I mean, when they're selling um, you know, a
47:03
Dan Nathan: United flight, that sort of thing. Two very different things in a way. So let's talk first about the the middlemen, the toll takers,
47:09
Unidentified speaker (UNKNOWN): yeah,
47:09
Dan Nathan: and then let's talk about some individual
47:11
Unidentified speaker (UNKNOWN): brands.
47:11
Dan Nathan: Yeah.
47:12
Unidentified speaker (speaker_2): So so the toll takers, I I you know, you would have thought they would be the first to go, right? The price line, Expedia, TripAdvisor, picket. But in reality, what you want is to trust the source of who's doing it. So yes, I could have my agent just go book everything direct, but then I'm on my own. I have no customer service. I have no one, no recourse on it, right? Whereas it
47:32
Unidentified speaker (speaker_2): if Expedia's This new function in the world is to provide a layer of I can stitch together your whole trip like a travel agent would. They are OTAs, online travel agents, and I will be the customer service for you. I will be that human type, like I will be the thing that ensures. Because the thing with travel, I've always I like that it's the example of every AI company tries to give an example of like booking better travel. It's
47:52
Unidentified speaker (speaker_2): because we can't get our time back. Like, sure, I can get my money back, but that week is gone if you travel Traveling and you don't have a good experience. And so like the the cost is very high. So there is, there's some benefit there. The United one is a little bit different, right? For United, if you had an assistant that was tireless and would stay up 24 hours and just
48:11
Unidentified speaker (speaker_2): constantly look for the best price and then, oh, and then a better price could switch over to that one. And actually this happens in the financial services. If you had someone that was constantly just moving your money around to slightly higher yield accounts all the time, it would break the system because the system is based on people, like some some amount of breakage. and humans just being lazy
48:29
Unidentified speaker (UNKNOWN): is probably the easy
48:30
Unidentified speaker (speaker_2): way
48:30
Unidentified speaker (UNKNOWN): to
48:30
Unidentified speaker (speaker_2): say it. Well, now they have to change that. It's changed their whole model. Like predictive models of who's willing to book first class at this time and who's willing to book coach and how like how many upgrades should we have have all broken because your agent is now optimizing for the things that they can get for free. It's like it's as if everybody has a points guy in their pocket, right? If you you know you have that one friend who's an absolute
48:50
Unidentified speaker (speaker_2): genius. at using points to get to Tokyo for like 50 bucks and a and a bag of peanuts, like somehow trading points. Everyone has that now. Well, everyone can't do that. So now what do they do?
49:01
Dan Nathan: Yeah. How how do the agents outside of let's say uh amuse, like how how do they what's instinct's business model? Do you
49:09
Unidentified speaker (UNKNOWN): know
49:09
Dan Nathan: what I'm saying? We
49:10
Unidentified speaker (UNKNOWN): don't
49:10
Dan Nathan: know just yet. You know, they just raise capital, they need to
49:12
Unidentified speaker (UNKNOWN): kind of
49:12
Dan Nathan: build out compute, they're literally gonna be subsidized by VC,
49:16
Unidentified speaker (UNKNOWN): right?
49:16
Dan Nathan: To get to a certain scale. Obviously, they also need a better mousetrap, you know, competing with something like Muse. You're gonna have to differentiate from really what the result standpoint is or the ease of use or speed
49:27
Unidentified speaker (UNKNOWN): or
49:28
Dan Nathan: something like that.
49:28
Unidentified speaker (speaker_2): Yeah, I mean, my hope would be that the business model is people value their time and attention, just like we said, and they value savings, like the how much money have you saved me. If there was some way that the agent could get a cut of the time it got you back and the money that it saved you, right, that would be a fair economic trade. And it would might be worth much
49:48
Unidentified speaker (speaker_2): more than the advertising dollars were. Because in theory, advertising dollars are paying for transactions. They're paying for you to be a customer of them. That margin could just go to you in some ways. And the advertisers could move back into the things I love, which is advertising. to humans, billboards. I mean, we just I don't know if you're here over the weekend for the New York City Marathon like
50:08
Unidentified speaker (speaker_2): a week ago.
50:09
Dan Nathan: You cracked what what was your number?
50:10
Unidentified speaker (speaker_2): I I did run in two fifty seven. But it was the hottest year on record. I should have been faster.
50:15
Dan Nathan: Three hours, man. That is absolutely sick.
50:18
Unidentified speaker (UNKNOWN): I
50:18
Unidentified speaker (speaker_2): just wanted to be done. Yeah.
50:19
Unidentified speaker (UNKNOWN): So that's
50:19
Unidentified speaker (speaker_2): a lot of running.
50:20
Dan Nathan: Did you take shots at Comos along the way? I remember a few.
50:21
Unidentified speaker (UNKNOWN): years ago. I did that next
50:22
Unidentified speaker (speaker_2): the next year when I was pacing someone.
50:24
Unidentified speaker (UNKNOWN): So great
50:25
Unidentified speaker (speaker_2): it's a great way to
50:25
Unidentified speaker (UNKNOWN): do it.
50:25
Dan Nathan: Yeah.
50:26
Unidentified speaker (speaker_2): Um but TCS sponsors at Marathon, Day of Joy. Like, it's an amazing use for advertising. Uh, run clubs, New Balance has has signed on as the title. It's an amazing sponsorship opportunity to like like
50:38
Unidentified speaker (UNKNOWN): get
50:38
Unidentified speaker (speaker_2): humans to try your product in the real world, subsidize experiences. Like, it's like I don't want people to get me wrong. Advertising is a great benefit when done correctly. It's when it becomes extractive that these agents might push back against and like that. would be a very good reset for a lot of the markets. All
50:53
Unidentified speaker (UNKNOWN): right.
50:53
Dan Nathan: So if you're not um monetizing an agent through uh ads,
50:57
Unidentified speaker (UNKNOWN): okay,
50:58
Dan Nathan: do you think there's an appetite on the scale of let's say hundreds of millions of people on this planet to
51:03
Unidentified speaker (UNKNOWN): actually
51:03
Dan Nathan: pay a subscription fee for these things?
51:05
Unidentified speaker (speaker_2): I do. I I do if the utility gets them what they want at the right price. Like if it is optimized, like think about how much money you spend on all these different streaming services, subscriptions you've forgotten to cancel, you know, booking a flight for more money than you don't necessarily need. If if your life is optimized, there's a lot of savings. Now there's a lot of work to
51:25
Unidentified speaker (speaker_2): do to like show people that how much benefit they're getting because people like free. Like it just we have said there's a great South Park episode, a little inappropriate, but great South Park episode where um uh what Steve Jobs is was trying to get people to read the terms of service. And he's like, Why won't you read? It's the same thing with when people like if you're getting something for free, someone's making money somewhere and they're
51:45
Unidentified speaker (speaker_2): probably making it off of you somehow. But like there is there's a great business in advertising uh to people's attention rather than buying their transactions. You
51:55
Dan Nathan: know, it's funny, and and I think about this again. A lot of our people who listen to us, well, you know, daily, weekly or whatever, they know I come up with some like stupid idea. ideas here, like
52:03
Unidentified speaker (UNKNOWN): they
52:03
Dan Nathan: just kinda pop in my head. But if you think about like Amazon Prime, yeah is a great example, right? They charge what, ten bucks a month or something like that. But they know that the average Prime member spends like two and a half X or
52:16
Unidentified speaker (UNKNOWN): something
52:16
Dan Nathan: like that. They're not so they're obviously getting value, they're paying for it, but then they're basically, you know, you get uh music and and videos and I just got an email from Uber Eats about Prime,
52:26
Unidentified speaker (UNKNOWN): you know,
52:26
Dan Nathan: they're doing some integrations there. You're gonna save delivery fees. Well, you know what? And the way in which you you and I use Uber Eats,
52:33
Unidentified speaker (UNKNOWN): you know, that that
52:33
Dan Nathan: probably gets that probably pays for Prime in two months or something like that.
52:36
Unidentified speaker (UNKNOWN): I feel
52:37
Unidentified speaker (speaker_2): like I'm making money on my Uber One subscription given how much.
52:40
Unidentified speaker (UNKNOWN): Well
52:40
Dan Nathan: so so my my question if if I'm Amazon,
52:43
Unidentified speaker (UNKNOWN): why
52:43
Dan Nathan: don't I just buy instinct, you know what I mean? And and if
52:46
Unidentified speaker (UNKNOWN): I'm
52:47
Dan Nathan: competing with uh you know if
52:48
Unidentified speaker (UNKNOWN): I'm gonna
52:48
Dan Nathan: block Muse, you know what
52:50
Unidentified speaker (UNKNOWN): I mean?
52:50
Dan Nathan: Uh like and again, it's build versus buy, it doesn't really matter
52:53
Unidentified speaker (UNKNOWN): in a way.
52:53
Unidentified speaker (speaker_2): You might, but then you have to just decide is your is this agent gonna be fully functional on Amazon and not the rest of the web? Will it will it compare products on our site equally to a product on Walmart's site, right? You probably wouldn't want to if you owned it. I also think I was having a conversation, I don't know if you know John Battel. Uh he he yeah wrote the books on search.
53:13
Unidentified speaker (UNKNOWN): And I did the
53:13
Dan Nathan: pod with him like years ago.
53:15
Unidentified speaker (UNKNOWN): He
53:15
Unidentified speaker (speaker_2): he's one of the clearest thinkers because he saw search come up. And like what's happening now? Here's some here's another parallel to draw for investors. We are speed running the the search engine wars of 2000s, right? SEO, remember white hat, black hat, SEO? No different right now with a GEO and AEO and all the companies promising to get to you to show up in the
53:35
Unidentified speaker (speaker_2): in the results. And I was talking to John, and and one of the things we kind of came to is what if there are a lot of small language models like Instinct popped up overnight. It's only going to get better. It's only going to get smarter. Not just Instinct, but the person who comes up with the next agent and they could use open source models. And these small models will have no advertising incentive. Their job will just be to work on
53:55
Unidentified speaker (speaker_2): your behalf to optimize these things. So there is a lot of disruption coming that's not yet baked in because we haven't seen it yet. Yeah.
54:02
Dan Nathan: the perfect example might be iOS and Apple.
54:05
Unidentified speaker (UNKNOWN): Think
54:05
Dan Nathan: about that. So it's kind of the uh printer versus the ink sort of situation. And you know, they build these small models, they live on the hardware, right? They're doing things that keep you, you know, much further engaged with their product. That means they're probably gonna end up selling more services, right? Through the um, you know, through the app store or that sort of thing. And so I think it's it right now from a
54:25
Dan Nathan: public investor standpoint.
54:27
Dan Nathan: I think we're so focused on disruption right
54:30
Unidentified speaker (UNKNOWN): now.
54:30
Dan Nathan: We're focused on monetization of you know some
54:33
Unidentified speaker (UNKNOWN): of
54:33
Dan Nathan: these models, the headwinds. If you're like a Google, um, that sort of thing. The companies are actually getting the return on the investments that they've made so far.
54:42
Unidentified speaker (UNKNOWN): So
54:42
Dan Nathan: when I think about all of that, let's just like focus on the Mag 7 seven for
54:45
Unidentified speaker (UNKNOWN): a
54:46
Dan Nathan: second.
54:46
Unidentified speaker (UNKNOWN): Okay.
54:46
Dan Nathan: So we'll start with Apple, we have Amazon, we
54:49
Unidentified speaker (UNKNOWN): have
54:49
Dan Nathan: Google. Um, let's throw test. Tesla out of there because that that is not particularly interesting to me. Microsoft, uh, and am I missing one?
54:56
Unidentified speaker (UNKNOWN): Nvidia?
54:57
Dan Nathan: Um, NVIDIA. That's a good one. Um, but like let's just start with uh Microsoft Azure, AWS for Amazon, Google Cloud. Now, those are like the cash cows right now for them, but they all have different distributions. Distribution.
55:09
Unidentified speaker (UNKNOWN): Well, cash
55:10
Unidentified speaker (speaker_2): cows, but still dwarfed inside of Google with ad dollars. Not in Microsoft, right? Like it's it's more it's more important than Microsoft. It's actually growing higher in Microsoft, and Microsoft is an enterprise business. Of Google Meta, Amazon, Microsoft, of those four, Microsoft's the only one that isn't highly reliant on advertising for the provided.
55:29
Unidentified speaker (UNKNOWN): And as an ad
55:30
Dan Nathan: guy, do you like that better? Do you think that's like remote?
55:35
Unidentified speaker (UNKNOWN): It
55:35
Unidentified speaker (speaker_2): is a better position to be in
55:37
Unidentified speaker (UNKNOWN): in a
55:38
Unidentified speaker (speaker_2): world where we're we don't know what the ad model is going to be. Now, uh
55:42
Unidentified speaker (speaker_2): the other guys can compete on a lot of the same fronts, like you just mentioned the cloud services and Google has so many more things going on with Waymo and everything else. So and and TPS or the chips. But in a world where we do not know if those cash cows of advertising, it would it's nice to not be reliant on it for the moment. Um
56:02
Unidentified speaker (UNKNOWN): they
56:03
Unidentified speaker (speaker_2): could figure it out or or Google could say, Great, we found a brand new monetization model at the point of search in Gemini. The only thing about that is
56:10
Unidentified speaker (UNKNOWN): it
56:10
Unidentified speaker (speaker_2): took 25 years to come up with the 10 blue links and create a whole marketplace of clicks to send traffic, and here's how we price it. Like, I don't doubt that Google could figure out how to monetize Gemini answer engines. I just don't think it happens overnight.
56:22
Dan Nathan: Yeah. So if you think about this, at one point Meta was down 30% thirty percent from its all-time all time high. highs earlier this year, right? We're talking about hundreds of billions of dollars of market cap. We just saw the same thing with Google,
56:33
Unidentified speaker (UNKNOWN): right?
56:33
Dan Nathan: And until Muse came out, I mean, you know, Microsoft Meta was just kind of dead in the water. Okay. Now, you know, it's given a lot of that back since their earnings. But when I think about those two companies, I think about your knowledge and your experience over the last 20 years. You built in that business, you sold it to Fox. You you you kind of ran a business at a big, big, you know, network across lots of
56:53
Dan Nathan: different
56:53
Unidentified speaker (UNKNOWN): platforms.
56:54
Unidentified speaker (speaker_2): Big by TV standards, three
56:55
Unidentified speaker (UNKNOWN): and a
56:55
Unidentified speaker (speaker_2): half billion dollars.
56:56
Unidentified speaker (UNKNOWN): Not crazy.
56:57
Unidentified speaker (speaker_2): Apparently small by
56:58
Unidentified speaker (UNKNOWN): uh
56:58
Dan Nathan: yeah, but I mean it was at a very formative time because the you know you you would have had a a lot of foresight to realize that it was there was gonna be businesses that were like a thousand X, that sort of thing. Okay. So if you think about that, Google's down 30%, and you look at the headwinds that you know of and your view about things. are those stocks like, are they in no man's
57:18
Unidentified speaker (speaker_2): land right now
57:18
Dan Nathan: down 30% from those recent highs?
57:20
Unidentified speaker (speaker_2): I mean, I would let you and and Guy and the other experts talk about the stock price versus value. What I would say is there is competitive pressure from people who are not like anthropic in OpenAI. Remember, OpenAI, again, that that paper that I mentioned earlier, Attention is all you need, it was all written by Google engineers.
57:36
Unidentified speaker (UNKNOWN): Yeah. But they
57:36
Unidentified speaker (speaker_2): didn't like launch
57:37
Unidentified speaker (UNKNOWN): it. Back in
57:37
Dan Nathan: 17.
57:38
Unidentified speaker (UNKNOWN): Yeah, back in. 17. And
57:40
Unidentified speaker (speaker_2): it was a nonprofit in OpenAI that had to launch it because no one with the financial incentives of the ad business wanted to launch into this arms race that we're all in right now. But once it happened, everyone had to be in it. And so having this competitive pressure from Anthropic and OpenAI who are just raising money and coming after these businesses is a pressure
58:00
Unidentified speaker (speaker_2): point. That said, I think they have all the rights. to win because of how much consumer engagement and distribution they have. Microsoft, the enterprise, Google with both search and YouTube. I think YouTube is a giant that people don't talk about. Like YouTube on the TV on your wall, not YouTube TV, is still the biggest streamer by far. And we all talk about Netflix and the
58:19
Unidentified speaker (speaker_2): streamers and the paramount deal that just got done with Warner Brothers. YouTube is just a monster of a business and has so much insight. So like the ecosystem. and Gmail isn't monetized in that way. There's so much there. So the pieces all fit. Meta, I think, is in the most interesting place in that the feed is still the feed. You're still going to get entertainment from that, the creator
58:39
Unidentified speaker (speaker_2): economy that sits on Instagram. I know they're going to the TV on the wall. But they have all the room to play with open models. They're not risking a search business. They're not risking a search advertising business. They
58:50
Unidentified speaker (UNKNOWN): can
58:50
Unidentified speaker (speaker_2): put play disruptor and do do things on an open basis. And that's why Muse got so much disruption. My mom loves Muse. It is giving her I and I love it for her because it's sounding her notes. Don't open that text you just got. It's probably a scam. That's wonderful. Like I think AI can be magic for people. And that's the other thing. This is a very unique piece of technology. Normally the kids adopt
59:10
Unidentified speaker (speaker_2): it first, or at least the tech forward people. You have a lot of like seniors, uh and and older generations that are adopting AI much quicker because it's not technical. You're just talking to machines in normal language.
59:21
Dan Nathan: I've been asking, um it's really interesting that you bring it up. I've been asking young people, okay, and and I think a lot of our listeners maybe they remember this. About a month ago, Gay Dami had twenty five uh finance majors, seniors in college at a school down in North Carolina, come to the Nasdaq before fast money. The guy ends up like not being in town. So
59:41
Dan Nathan: I had
59:42
Unidentified speaker (UNKNOWN): the opportunity to sit down and speak to them.
59:44
Dan Nathan: Um but we were talking about like how to break into, you know, finance and and that sort of thing. And we we got to, you know, LLMs, we got to agents, and I asked them how many of you are using you know personal agents? And
59:57
Unidentified speaker (UNKNOWN): this was
59:57
Dan Nathan: a month ago.
59:58
Unidentified speaker (UNKNOWN): Yep.
59:58
Dan Nathan: Okay. Yep. Not one of them raised their hand.
1:00:00
Unidentified speaker (UNKNOWN): Yeah.
1:00:01
Dan Nathan: Not one I I was absolutely
1:00:03
Dan Nathan: you know, shocked. So it's interesting that you're seeing adoption, you know, from adults rather than kids. You know, and it brings me back to the late 90s where the idea of getting an adult to put their credit card on the internet, you know, this in 1998, nine or whatever,
1:00:18
Unidentified speaker (UNKNOWN): was a r
1:00:19
Dan Nathan: it was a really hard thing to do. You know what I mean?
1:00:21
Unidentified speaker (UNKNOWN): And this just
1:00:21
Unidentified speaker (speaker_2): shows you this is it's that's why it's so different now. I I did a very similar thing. I just talked to a group of marketing students. like uh who all want to get into the marketing advertising field. And some of them have moral objections. What I kind of said to them is this train is going, like you have
1:00:34
Unidentified speaker (UNKNOWN): to
1:00:34
Unidentified speaker (speaker_2): figure it out.
1:00:35
Unidentified speaker (speaker_2): But here's the good news. The good news is if you had said six months ago, what do you have to hire for? A lot of people would be like, AI skills. You should hire for AI skills. I don't think so, right? Because six months ago, to set up an open claw, club, you'd have to get a Mac mini. Remember there was a run on Mac minis mini,
1:00:50
Dan Nathan: and you had
1:00:50
Unidentified speaker (UNKNOWN): to.
1:00:51
Unidentified speaker (speaker_2): go, you'd set it up and you had a server and you could s create an agent architecture to like have people. Now I just sign up for Muse. Like the tech is coming to us. Like the tech is coming to humans and just asking like a genie, what do you want me to do? Like and you're just talking to it. So that leaves the human skills. Everything machines can't understand is where we have the advantage, right? Human relationships
1:01:11
Unidentified speaker (speaker_2): and trust, like I physical space and gathering. You know, I I use this one. You you'll appreciate I was on the board of Clear Channel for seven years. I am no longer on the board. It was taken private by Mubatala. So I can say this without a horse in the race. They're building trillions of dollars of machine gods in Silicon Valley. And what what are they buying to tell you about it?
1:01:30
Unidentified speaker (speaker_2): Billboards, subway advertising, physical space in the real world, Super Bowl ads. It will be one of the biggest categories of advertising. during the Super Bowl. Why? These are cultural events where people are and humans are there. So I think those things become more valuable. And we're good at that.
1:01:50
Dan Nathan: And I I, you know, benefit from that. I mean, you've, you know, this is uh hold on, people, human
1:01:56
Unidentified speaker (UNKNOWN): or nothing. I mean, this is
1:01:57
Dan Nathan: on your head right
1:01:58
Unidentified speaker (UNKNOWN): there.
1:01:58
Dan Nathan: And you know, your business is very unique, you know, human ventures in your
1:02:02
Unidentified speaker (UNKNOWN): business.
1:02:03
Dan Nathan: general, because you know, you start with who are these people? You know what I mean? Like, what are they all about? You know what I mean? What is their track record? What is their vision? And then the thing that they're building has to be intermeshed. Is that fair with all of those sorts of qualitative sort of judgments? And you know, I think it is something when you talk about these digital gods, that sort of thing.
1:02:23
Dan Nathan: I don't believe that the people that are building these things out there, when I say the people, we all know who they are. These are the you know, trillion-dollar, you know, market cap people or whatever. I don't believe that they have our best interest in mind. And I'm just curious, like, how that kind of, you know, like how does that work into the
1:02:39
Unidentified speaker (UNKNOWN): way
1:02:39
Dan Nathan: you think about this business and how it's monetized going forward?
1:02:42
Unidentified speaker (UNKNOWN): Well
1:02:43
Unidentified speaker (speaker_2): well, I think that's I'll take it, I'll take that in another way, which is San Francisco has begun to realize, right, Silicon Valley, that taste, trust, brand, media are all gonna matter, right? And so they've been investing in it. They've been investing in their direct to consumer. They want to go direct, right? And yet New York really represents
1:03:02
Unidentified speaker (speaker_2): media, advertising, fashion, culture, hospitality. Like these are the the human businesses. And then so we think about it from an like a very New York centric. The front of this hat says human or nothing. The back says New York or no. or We made these. in partnership with our friends there. And it's not, again, human or nothing is not a anti AI. I like I said, AI is
1:03:22
Unidentified speaker (speaker_2): absolute magic for people. It could get us our time back, it could get us our attention back. It could help us with get better deals. It can help my mom and my dad. So love AI for humans. I love AI that gets people back to talking to humans. But the realization that San Francisco's having is that taste brand trust. these things are going to matter again, right? Media matters. And New York
1:03:42
Unidentified speaker (speaker_2): is just better at that. We have been very good at storytelling and very bad at the tech side of things. Well, if AI distributes tech and tech isn't the hard part anymore, now the hard part is trust brand. We're pretty good at that here. So we tend to think about what's going to be the hard thing. We like investing in things that end up in the physical world.
1:04:02
Unidentified speaker (speaker_2): It can be a tech, it could be AI, just ends up in the physical world, or it covers the world of hospitality, like our friends at Caper.
1:04:09
Dan Nathan: Let's finish with this. And when I think about what's going on today in the markets, right? It is just a great microcosm of what was, you know, the debates that are going on right now. Like the the NVIDIA, and you just heard Guy and I talk about it for, you know, 25 minutes or something like that. You know, that's something that's not ending today, right? Like so that,
1:04:29
Dan Nathan: but that you could say is is a very like, you know, idiosyncratic sort of situation. Right. Ultimately, there will be knock-on effects, right? Um, of like the whole ecosystem, uh, you know, anthropic's coming. There's gonna be a ton of scrutiny again on how that company gets to profitability, what their burn is, how much money they're losing, that sort of thing. But when I think about the
1:04:49
Dan Nathan: disruption that's happening on all the toll. takers, the disruption that's going to be obviously within the publishing industry, within e-commerce. I mean the list goes on and
1:04:57
Unidentified speaker (UNKNOWN): on.
1:04:57
Dan Nathan: What is one thing? Let's assume that things get we get a little mired here. There's a digestion phase, you know, the way businesses, consumers, you know, all think about AI. Everything that you just mentioned as far as humor interaction and taste and culture and all those sorts of things, they're not going away. No. Okay. No,
1:05:15
Unidentified speaker (speaker_2): no, they're their prices are exploding. Look at sports teams. Look at league look at league rights. Look at like, you know, it's it's almost it's almost to the detriment of the league how much the the price of teams are going up, right? The Super Bowl ads are ten million dollars a spot. Uh you have Barry Diller and Ari Emanuel like investing in the experience economy. I think they know what they're doing, right? Like these are not anti AI bets. Like,
1:05:35
Unidentified speaker (speaker_2): you know, Apollo is doing tons of AI investment, but they also just were part of the team that brought Soho House private, a global network, a physical space with like like interesting people as members. So I think that that's your barbell to be playing with. Understand the AI trade, understand how it's going to affect advertising and how the knock-on effects of that to every other business. And then what are the
1:05:55
Unidentified speaker (speaker_2): businesses that are, let's call it AI resilient, because it's not, it's not anti, right? It's in theory, it should be helped. Like we have a company in our portfolio breakaway. largest touring music festival.
1:06:05
Dan Nathan: I love that.
1:06:06
Unidentified speaker (speaker_2): Amazing festival.
1:06:07
Unidentified speaker (UNKNOWN): We went this
1:06:07
Dan Nathan: summer uh under the K-Bridge
1:06:08
Unidentified speaker (UNKNOWN): for the U.S. Under the K Bridge for Town. That was
1:06:10
Dan Nathan: awesome. It was good. Awesome.
1:06:12
Unidentified speaker (speaker_2): 400,000 kids every summer wearing breakaway hats, breakaway t-shirts. It's a it's a thing that they do once a year and they they they look forward to it the next year, right? That human experience is more valuable than it's being monetized right now. And I think you'll see a lot of money. flowing into that side.
1:06:27
Unidentified speaker (UNKNOWN): Right.
1:06:28
Dan Nathan: So so and we'll just end right here. So when when I think about everything that you just said, I get it. You
1:06:34
Unidentified speaker (UNKNOWN): just
1:06:34
Dan Nathan: said some specific things that differentiate New York
1:06:37
Unidentified speaker (UNKNOWN): from
1:06:37
Dan Nathan: Silicon Valley. Um, I think a lot of public market investors, they only they they think about AGI, superintelligence, you know, all all the different things that are gonna turn a four trillion dollar market cap company into the first 10 trillion. ten dollar or whatever, you know, that sort of thing. You know, when I listen to everything that you just had to say, right, you know, is
1:06:57
Dan Nathan: there value that will be created that's being underappreciated right now in a digestion phase? Let's just say the Nasdaq's down 30%. People don't think it can happen. It happened in 2022, people, and the technology that was, you know, really at the um I don't know, that was fueling the excitement in the stock market.
1:07:15
Dan Nathan: metaverse was a joke, right? And then what we did see is such uh digested phase. We believe this is magic, right? But the way in which it gets monetized is really not being appreciated from all the things that you just said
1:07:27
Unidentified speaker (UNKNOWN): just yet.
1:07:28
Unidentified speaker (speaker_2): But the the reinvention of search, or let's not even call it the reinvention, the advancement of search, which the entire internet economy is based on, which means the global economy is based on it. is not a two or three year digestion phase. That is a happening right now portion. The knock-on effects for all the businesses that need to show up in search results in order to get found is not a
1:07:48
Unidentified speaker (speaker_2): two or three years down. It's happening right now. And so I think thinking about what is resilient to a search, let's call it a search disruption rather than an And AI disruption, and what things are human, like how is human taste going to get into the machine? Like how if there's a great coffee shop around the corner, how does the machine know unless we tell it? Right. Where is that information coming from? I I'd
1:08:08
Unidentified speaker (speaker_2): invest in those. I'd be investing in local, right? Because local at scale, there's, you know, there's all politics is local. Like I think all marketing becomes local. Those are the things that I think you can feel good about. And by the way, have both. Have the AI upside with with some assets that are not disrupted by AI.
1:08:25
Dan Nathan: Yeah. And just like, you know, um you coined attention economy. Um I didn't coin this.
1:08:30
Unidentified speaker (UNKNOWN): Quite coin it.
1:08:31
Dan Nathan: This is like I think about everything you just said, and I think intelligent assistance. I turned AI around in many different ways. And I think that's something that I think is probably j everything you just said gonna be underappreciated. Joe Marcese, one of my favorite humans in the world. This is human.ai. If you want a copy of it, go to
1:08:51
Dan Nathan: your favorite podcast store. Check out Market Matrix. That is our daily email that's a summary of business topics, market topics, that sort of thing. Go follow it, rate it, send a screenshot to Amanda at riskreversal.com. And we're going to send you uh a link where you can download one of Joe's fabulous new books. And shout out to Jonathan Bing, who is a legend
1:09:11
Dan Nathan: in his own right. So um great, great book. Um so Joe, thank you for
1:09:16
Guy Adami: thank you for having me. Appreciate
1:09:17
Unidentified speaker (UNKNOWN): it.
1:09:21
Unidentified speaker (UNKNOWN): This
1:09:21
Guy Adami: podcast is for informational purposes only. All opinions expressed by me, Dan Nathan, Guy Dani, and any other participants are solely our opinions and should not be relied upon for specific investment decisions.
1:09:32
Guy Adami: This episode of the Risk Reversal Podcast is presented by Fidelity Investments and their all all-new new advanced trading platform. Fidelity Trader Plus. If you want to stay close to the trade from any device, Fidelity Trader Plus offers seamless access to your preferred tools and charts across your desktop, web, and mobile devices with free
1:09:51
Guy Adami: third-party analyst ratings and custom alerts. Learn more about the next generation of advanced trading from Fidelity and try it for yourself at Fidelity.com/slash TraderPlus. Fidelity Investments and Risk Reversal are not affiliated. Views, opinions, products, services, and strategies discussed are not endorsed or promoted by
1:10:11
Guy Adami: Fidelity Investments, Fidelity Brokerage Services LLC, Member NYSE, SIPC.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from RiskReversal Pod

RiskReversal Pod

There’s a New Kid in Town: Nvidia, Anthropic & the AI Unwind

Episode discussion: Hosts discuss Nvidia valuation, Anthropic IPO, and broad AI trade unwind amid leadership changes.

Speakers’ claims: Guy Adami and Dan Nathan analyze Nvidia's recent price action and valuation, noting the stock traded near support levels despite high growth expectations. They highlight the leadership change with Jensen Huang's departure and cite a Wall Street Journal report on internal concerns and circular financing involving neo-clouds. The hosts connect the AI sector's weakness to rising Treasury yields and Bitcoin's drop, while discussing Anthropic's massive IPO. They draw parallels to Cisco's 2007 collapse and note a broad unwind in semiconductors, leaving only large-cap tech stocks as safe havens.

NvidiaAI TradeAnthropic IPO
36 min6 assets20 claims to verify
RiskReversal Pod

Pricing Nvidia's "Key Man Risk" with Gene Munster

Episode discussion: Gene Munster argues NVIDIA's drop prices in excessive exit risk, predicts sustained hyperscaler CapEx and Apple's hardware super cycle

Speakers’ claims: Gene Munster joins the RiskReversal Pod to discuss market reactions to NVIDIA and broader AI infrastructure trends. He argues that NVIDIA's stock decline prices in a 25% probability of CEO Jensen Huang leaving, which he estimates is higher than the actual 10-15% risk. Munster forecasts that hyperscaler capital expenditure will remain robust through 2028, driven by inference demand and a prisoner's dilemma among cloud providers. He also analyzes Google's valuation challenges from personalized AI agents and maintains a bullish long-term thesis on Apple, citing a hardware super cycle and shifting revenue mix. Finally, he highlights the critical importance of Anthropic's upcoming IPO for sector confidence.

NVIDIAJensen HuangHyperscaler CapEx
30 min3 assets10 claims to verify
All insights