The Real Eisman Playbook

Midterm political shifts and AI ROI concerns reshape data center and tech investment landscape

53 min Listen to the original Episode page
Data CentersAI InfrastructureHyperscalersNeocloudsGoogleOracleIntelApple

Summary

Speakers Steve Eisman, Dan Ives, and Gil Luria discuss how a Democratic congressional shift may slow new data center construction, benefiting existing infrastructure providers like Dell and Cisco over neoclouds like CoreWeave. They analyze Google's slowing search growth as a sign of AI-driven ad revenue erosion and express caution on Oracle's debt-heavy AI bets and Intel's structural execution issues. The conversation highlights Anthropic and OpenAI's commoditization risks, the rise of orchestration layers like Palantir's for data privacy, and Apple's potential as a low-capex AI monetization play. Finally, they contrast CoreWeave's high-risk financial structure with Nebius's profitability and suggest AI infrastructure stocks may be undervalued in a responsible growth scenario.

Episode sections

  1. 0:04 – 2:16

    Intro and AI Narrative Shift

    Steve Eisman introduces guests Dan Ives and Gil Luria, noting the political shift with Democrats taking Congress. They discuss how the AI narrative has evolved from pure hype to a focus on ROI and capital expenditure concerns, with the political environment potentially slowing data center build-outs.

  2. 2:16 – 8:00

    Hyperscalers vs. Neoclouds and Infrastructure

    Dan Ives argues that hyperscalers benefit regardless of AI spending levels, while Gil Luria notes that AI results have been strong but non-AI sectors are weak. The political 'blue wave' is seen as a headwind for new data center construction, benefiting existing infrastructure providers like Dell and Cisco over neoclouds like CoreWeave.

  3. 8:00 – 13:40

    Google Search Growth Slowdown

    Eisman highlights Google's slowing search revenue growth (12% vs 19% previously). Luria and Ives interpret this as a fear-driven reaction to AI competition, noting Google admitted AI searches generate less ad revenue and that chatbot traffic is displacing human searches, potentially eroding ad dollars.

  4. 13:40 – 17:40

    Oracle's High-Wire Act

    The guests express caution on Oracle, describing it as a 'high-wire act' due to massive debt/equity raises for data centers while its core software business suffers. Ives compares Oracle's bet on OpenAI to a baseball team betting on one player, while Luria notes Oracle's core database business funds the AI build-out but faces execution risks.

  5. 17:40 – 22:00

    Intel's Structural Issues

    Both guests view Intel as a 'broken company' with poor execution and low gross margins (30s) compared to AMD (50s). Despite a capable CEO and government support, the fundamental inability to produce chips at competitive yields makes it a difficult investment compared to competitors.

  6. 22:00 – 27:04

    LLM Moats and Enterprise Strategy

    Eisman presents a bear case for Anthropic and OpenAI, citing low switching costs and price wars. Ives agrees models will commoditize but argues the moat is enterprise entrenchment. Luria adds that while the team is a moat, the market is narrowing due to open-weight models and orchestration layers reducing reliance on frontier models.

  7. 27:04 – 28:35

    AI Model Privacy and the Rise of Orchestration Layers

    Speaker_2 argues that using proprietary models like Anthropic's risks leaking corporate alpha, driving demand for orchestration layers like Palantir's to protect data. This shift allows companies to switch between models without exposing proprietary information, benefiting Palantir, Microsoft, and Amazon.

  8. 28:35 – 30:49

    Anthropic's Market Share vs. Total AI Growth

    Speaker_2 posits that while Anthropic and OpenAI's slice of the AI pie will shrink as companies adopt open-weight models, the total pie is growing rapidly. The investment thesis depends on whether the total market grows faster than their individual market share shrinks.

  9. 30:49 – 36:03

    Software Transition and the 'SASPocalypse'

    The discussion addresses the 'SASPocalypse' thesis, noting that while AI threatens traditional software, companies like ServiceNow and Adobe are in a painful transition. Speaker_1 argues these firms have a window to monetize AI, but institutional impatience exacerbates stock volatility during this period.

  10. 36:03 – 42:12

    NeoClouds and CoreWeave's Structural Risks

    NeoClouds like CoreWeave are described as competing with hyperscalers but facing higher capital costs. Speaker_2 characterizes CoreWeave as an NVIDIA SPV with negative carry, where equity holders own nothing due to low returns and high debt, while insiders have sold significant stock.

  11. 42:12 – 44:43

    Apple as a Safe Haven in AI

    Apple is viewed as a 'safe trade' or utility-like stock due to its low CapEx and massive install base. Speaker_1 believes Apple will monetize AI through services and developer ecosystems, benefiting from the 30% App Store cut on AI subscriptions without bearing the heavy infrastructure costs of competitors.

  12. 44:43 – 47:47

    Agentic AI and Consumer Surplus

    Agentic AI is expected to create consumer surplus by finding the best prices for services like insurance and banking, bypassing brand loyalty. Speaker_2 views this as positive capitalism that accelerates the decline of inefficient companies while rewarding those that provide genuine value.

  13. 47:47 – 49:15

    Institutional Outlook and Chip Demand

    Speaker_1 predicts chip demand equilibrium until late 2028/early 2029, with hyperscaler CapEx peaking in 2027. He anticipates a significant run in tech stocks over the next 6-9 months as winners are separated from losers, with valuations remaining attractive relative to the broader market.

  14. 49:15 – 52:54

    Goldilocks Scenario for AI Infrastructure

    Speaker_2 describes a 'Goldilocks' scenario where slower, debt-free growth (20-25% annually) extends the AI cycle healthily. He argues that if AI adoption continues responsibly, the sector is hugely undervalued, with chip and cloud companies trading at significant discounts to market multiples.

Assets discussed

AssetKindStanceContext
MicrosoftMSFTequitymixedIt's actually the non-AI stuff that looks weaker. Like PC sales are down. That hurts Microsoft.
AlphabetGOOGLequitymixedIt hurts Intel. Search is down. That hurts Google. Retail's down. that hurts Amazon, it hurts Shopify.
AmazonAMZNequitymixedRetail's down. that hurts Amazon, it hurts Shopify.
Shopify Inc.SHOPequitybearishRetail's down. that hurts Amazon, it hurts Shopify.
NVIDIANVDAequityneutralOkay. So let me press the point. Forget about NVIDIA just for a second.
AppleAAPLequitymixedNeo clouds, power, hyperscale. hyperscalers, Apple, that's almost saying like the Kingmaker in terms of where they sit.
BitcoinBTCcryptoneutralanalogy is important on on the crypto. Bitcoin miners were a great business while Bitcoin was going up fast, and then they were a terrible. business.
Snowflake Inc.SNOWequityneutralThe use cases that are playing out from Palantir. to Snowflake to other software companies, my my view is that
Broadcom Inc.AVGOequityneutralAnd if we can do that, this entire sector is hugely undervalued because you have Microsoft, Amazon, Google trading basically at a market multiple. You have all these really good chip companies, really good companies, NVIDIA, Broadcom,
ADVANCED MICRO DEVICES INCAMDequityneutralAMD, Micron trading at a half, two. to a quarter of a market multiple. And in this scenario that we're growing data centers 20-25% for the next three to five years instead of 70% next year, those companies are worth a lot more than they are now. So to me, that's Goldilocks. Bring it on, regulate,
MICRON TECHNOLOGY INCMUequityneutralAMD, Micron trading at a half, two. to a quarter of a market multiple. And in this scenario that we're growing data centers 20-25% for the next three to five years instead of 70% next year, those companies are worth a lot more than they are now. So to me, that's Goldilocks. Bring it on, regulate,

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

Claims worth verifying

Democrats have taken both houses of Congress.

Steve Eisman ·0:51

How to verify: Check official results from the U.S. House of Representatives and Senate for the 2026 midterm elections. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Dan Ives' new company is called Yorkville.

Unidentified speaker (speaker_1) ·0:20

How to verify: Search for 'Dan Ives Yorkville Investment Partners' in business news or SEC filings. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Google search revenue growth was 12%, down from 19% in Q1.

Steve Eisman ·8:13

How to verify: Check Alphabet's Q3 2026 earnings report for search revenue growth rates. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

More than half of web traffic comes from chatbots now.

Unidentified speaker (speaker_2) ·11:15

How to verify: Cite specific web traffic data sources (e.g., SimilarWeb, Statista) or Google's internal data if referenced. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Oracle is trying to be a hyperscaler and made a huge bet on OpenAI.

Unidentified speaker (UNKNOWN) ·14:53

How to verify: Check Oracle's press releases and SEC filings regarding the OpenAI partnership and data center investments. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Intel's gross margins are in the 30s.

Unidentified speaker (UNKNOWN) ·19:16

How to verify: Check Intel's Q3 2026 earnings report for gross margin percentage. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

AMD's gross margins are in the 50s.

Unidentified speaker (speaker_2) ·19:41

How to verify: Check AMD's Q3 2026 earnings report for gross margin percentage. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Companies using proprietary models like Anthropic's are handing over proprietary information and alpha, as providers will eventually compete with customers.

Unidentified speaker (speaker_2) ·27:06

How to verify: Check Anthropic's privacy policy and terms of service regarding data usage and model training; review SEC filings for any competitive overlaps with enterprise customers. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Palantir is 'killing in the market' due to demand for orchestration layers that protect proprietary data.

Unidentified speaker (UNKNOWN) ·28:17

How to verify: Review Palantir's recent quarterly earnings reports for revenue growth and customer acquisition metrics; check stock performance data. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Anthropic and OpenAI's slice of the AI pie will shrink as companies adopt open-weight models.

Unidentified speaker (speaker_2) ·28:35

How to verify: Monitor industry reports on AI model adoption rates (proprietary vs. open-source) from firms like Gartner or IDC; track Anthropic/OpenAI revenue vs. total AI spend. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

CoreWeave is an NVIDIA special purpose vehicle (SPV) created to diversify NVIDIA's customer base.

Steve Eisman ·37:52

How to verify: Review CoreWeave's SEC filings (S-1 or 10-K) for investor presentations and NVIDIA's stake or strategic partnerships; check NVIDIA earnings calls for mentions of NeoCloud strategy. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

CoreWeave generates low single-digit returns and borrows at 9% (blended), creating negative carry.

Unidentified speaker (speaker_2) ·40:09

How to verify: Analyze CoreWeave's financial statements (10-K/10-Q) for return on invested capital (ROIC) and weighted average cost of debt (WACD). For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

CoreWeave insiders have sold hundreds of millions of dollars in stock, the most seen in 20 years.

Steve Eisman ·41:05

How to verify: Search SEC Form 4 filings for CoreWeave (CRWV) insider transactions to quantify sales volume and frequency. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Nebius has barely any debt and is managed by experienced operators focused on profitability.

Unidentified speaker (UNKNOWN) ·41:32

How to verify: Review Nebius's financial disclosures or public reports on debt levels and management background. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Apple has 2.5 billion iOS devices, providing a massive install base for AI monetization.

Unidentified speaker (UNKNOWN) ·44:32

How to verify: Check Apple's latest quarterly earnings report or investor presentation for active device install base numbers. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

NVIDIA trades at 11-12 times earnings, while average tech companies trade at 19-20 times.

Unidentified speaker (speaker_1) ·48:51

How to verify: Calculate current P/E ratios for NVIDIA and a broad tech index (e.g., XLK) using financial data providers like Bloomberg or Yahoo Finance. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Hyperscaler CapEx is likely peaking in 2027, with free cash flow rebounding in 2028.

Unidentified speaker (speaker_1) ·48:31

How to verify: Review capital expenditure guidance and forecasts from major hyperscalers (Microsoft, Amazon, Google, Meta) in their latest earnings calls and investor presentations. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify Google's Q3/Q4 search revenue growth rates and ad load metrics against SEC filings (10-Q/10-K) to assess AI cannibalization claims.
  • Check Oracle's latest 10-K for total debt, equity raises, and segment margins to evaluate the 'high-wire act' risk profile.
  • Review Intel's quarterly gross margins and wafer yield reports to compare against AMD's competitive positioning.
  • Analyze CoreWeave's capital structure, debt covenants, and insider trading activity (Form 4 filings) to assess the 'negative carry' and equity value claims.
  • Confirm Palantir's enterprise adoption metrics and data privacy features to validate the 'orchestration layer' demand thesis.
  • Examine Apple's services revenue growth and App Store commission structures to test the 'low CapEx AI monetization' hypothesis.
  • Research Nebius's debt levels and profitability metrics to contrast with CoreWeave's financial structure.
  • Check hyperscaler CapEx guidance for 2027-2029 to validate the 'peak spending' and 'slower build-out' timelines.
Full transcriptexpand
0:04
Unidentified speaker (UNKNOWN): Hi,
0:05
Steve Eisman: this is Steve Eisman. Welcome to another episode of the Real Eisman playbook. I'm going to make this intro short because there's just so much to discuss. We have back two recurring guests, Gil Lurie of Davidson and Dan Ives, whose new company is called Yorkville
0:20
Unidentified speaker (speaker_1): Ives.
0:20
Steve Eisman: Yorkville. And before we um get into a discussion, I'm gonna going to brag for just one second.
0:27
Steve Eisman: because I made one of the greatest predictions of my entire life with you guys when you guys were here at the end of July. At the very end of the interview, when I was saying thank you, I then said, and we'll have you back in three to six months, because between now and then the world will have turned over at least 15 times.
0:48
Steve Eisman: And you know what? That's an underestimate.
0:50
Unidentified speaker (speaker_1): Underestimation.
0:51
Steve Eisman: So just so everybody knows, um, this today is November 4th. It's the day after the election. The Democrats have taken both houses of Congress. We'll get to that.
1:04
Steve Eisman: Since what I want to do is from your per both of your perspectives,
1:10
Steve Eisman: the AI narrative has migrated. I mean, and even people who like such as mostly yourselves think it's very positive. I think everybody would agree that narrative is a lot more complicated now. So let's just do something very dispassionate, as opposed to taking a position on it, which which we'll come to. Dan, since last we met
1:30
Steve Eisman: in the summer, how has the AI narrative, in your view, changed?
1:35
Unidentified speaker (speaker_1): I think it's changed in on one hand. the use cases and the actual, I think not just the visualization, it's become more of a reality for companies that they're actually going down the use case path. In terms of ROI, that's become much more front-centered. We've seen a Palantir and so many other, I think, tech companies.
1:56
Unidentified speaker (speaker_1): On the negative, I think it's the view that the level of capital raises and the amount of money being spent is something that essentially more or less has drowned the market. I think, and I think the market screamed the last few weeks on the tech side. Nomas, enough's enough. And
2:16
Unidentified speaker (speaker_1): we're at this sort of tug of war point where
2:19
Unidentified speaker (UNKNOWN): the
2:20
Unidentified speaker (speaker_1): AI revolution's accelerating. And We see that in Korea a few weeks ago in terms of chips, Taiwan. We see it on the use case side in terms of enterprise.
2:27
Unidentified speaker (speaker_1): But then the spending, I think investors have said we're done, we have to slow this down, not realizing that it was going to take a blue tsunami to actually slow it down, which is why I think tech stocks are
2:38
Unidentified speaker (UNKNOWN): up.
2:40
Steve Eisman: Explain that last sentence, because I don't quite understand what you just said.
2:43
Unidentified speaker (speaker_1): Because I think, and I'm so interested Interest here, what Gil says, the last few weeks, no matter what tech companies have done. stocks sell off because you have ATMs, capital raises, bond offerings, converts, they're going to Malaysia to raise money, wherever it may be. And I think essentially what's happened is, and I've seen that from an
3:02
Unidentified speaker (speaker_1): institutional perspective, investors like, look, I feel like at these levels of capital raises in ATMs, these stocks almost can't go higher. Finally, now, politically, because of what we're seeing in Texas, because of the the blue wave, now they're gonna have to slow down. CapEx 27 is probably a peak year for
3:22
Unidentified speaker (speaker_1): CapEx. It was thought you probably wouldn't see that until 2029. And for the hyperscalers, the strong get stronger because they're the ones that are ultimately neo clouds on the other side of this. But that's why I think hyperscalers are so much higher today.
3:37
Steve Eisman: Gil?
3:39
Unidentified speaker (speaker_2): Every single AI company that's reported, and every company that's reported about their AI business over the last three months has reported better than expected results, increased their outlook, and increased their visibility. It's actually the non-AI stuff that looks weaker. Like PC sales are down. That hurts Microsoft.
3:59
Unidentified speaker (speaker_2): It hurts Intel. Search is down. That hurts Google. Retail's down. that hurts Amazon, it hurts Shopify. The AI results from all these companies, from the labs, from the hyperskills, from the chip companies, have all been better than expected. So that's the good thing that's happened. The not good thing that's happened is that it became increasingly
4:18
Unidentified speaker (speaker_2): political over the last few months. Leading into these elections, it became a slam dunk issue.
4:25
Unidentified speaker (UNKNOWN): It
4:25
Unidentified speaker (speaker_2): became the slam dunk issue. It was a big part. That
4:28
Unidentified speaker (UNKNOWN): is
4:29
Steve Eisman: bad.
4:29
Unidentified speaker (speaker_2): Data center bad.
4:32
Unidentified speaker (speaker_2): It was a big part of the blue wave. Even though it was a bipartisan issue, there were Republicans running against data centers, but it was part of the Blue Wave. And now we're gonna have to deal with this reality that it's gonna be harder to build a data center. It's not that we're gonna stop, but it's gonna be harder to build a data center. There's gonna be a lot of downstream
4:52
Unidentified speaker (speaker_2): impacts from that. But that's an outcome of the elections, is it's gonna be harder and more expensive
4:58
Unidentified speaker (UNKNOWN): to build
4:58
Unidentified speaker (speaker_2): data centers in the US. Yes.
5:00
Unidentified speaker (UNKNOWN): Especially
5:00
Unidentified speaker (speaker_2): in Texas.
5:01
Steve Eisman: And slower.
5:02
Unidentified speaker (speaker_2): Yes.
5:03
Steve Eisman: Okay. So let me press the point. Forget about NVIDIA just for a second.
5:08
Steve Eisman: Take like equipment companies like Arista, Cisco.
5:13
Unidentified speaker (speaker_1): Networking, yeah.
5:14
Unidentified speaker (UNKNOWN): Um
5:15
Steve Eisman: Dell.
5:16
Unidentified speaker (speaker_1): Yep.
5:16
Steve Eisman: Those stocks are up today.
5:19
Steve Eisman: But the
5:21
Unidentified speaker (UNKNOWN): G
5:22
Steve Eisman: Vanova Novas of the world, the quantas of the world are all down. How do you explain it? Does that make sense to you?
5:29
Unidentified speaker (speaker_1): Because Dell, Cisco, they're not building data centers. They're the infrastructure, the install base, they're ones now, the existing data centers. they're gonna have more opportunity than ever. But with incrementally, let's call it 10 to 15% of data centers now not being built that called the 800 to
5:48
Unidentified speaker (speaker_1): 1000, and let's say 15% of those don't happen. That's negative for G Vernover. It's negative for
5:55
Steve Eisman: negative for Arista, it means less equipment sales.
5:58
Unidentified speaker (speaker_1): But I would say right now is that the worry is that there was almost gonna going to be an overbuilding. out, an overcapacity. And I think that's not going to be the issue. If you look at the next few years, that's why Dell, they, I would say probably the next 18, 24 months, they already see what their sort of prospects are. If you go
6:18
Unidentified speaker (speaker_1): further out, that's one thing. But I think that's why you're seeing like a delineation duiniation that's happening in this market. Neo clouds, power, hyperscale. hyperscalers, Apple, that's almost saying like the Kingmaker in terms of where they sit.
6:31
Unidentified speaker (speaker_2): So if AI is not taking the oxygen out of the room, that's good for the rest of tech, for non-AI tech,
6:38
Unidentified speaker (UNKNOWN): which
6:38
Unidentified speaker (speaker_2): includes networking equipment, et cetera. It's good for the non-AI tech because AI has been taking the oxygen out of the room. Enterprises are spending so much on AI, they don't have money for anything else. If they're going to spend less on AI, that's more money for other things. Dan made a really important point, though, that the hyperscalers win anyway.
6:57
Unidentified speaker (UNKNOWN): The
6:57
Unidentified speaker (speaker_2): hyperscalers win if there's more AI, the hyperscalers win if there's less AI, they win anyway. And you're seeing
7:02
Unidentified speaker (UNKNOWN): that. And
7:03
Steve Eisman: why why do you think that's the case?
7:05
Unidentified speaker (speaker_2): Because if there's more AI, the revenue continues to accelerate. If there's less AI, their cash flow accelerates.
7:10
Steve Eisman: Okay.
7:11
Unidentified speaker (speaker_1): And
7:11
Unidentified speaker (UNKNOWN): so they win
7:12
Unidentified speaker (speaker_2): either way.
7:13
Unidentified speaker (speaker_1): And I'd also add, part of the issue with the Oracle, Oracle's like, we want to be a hyperscaler. SpaceX, we want to be a hyperscaler in Texas. Microsoft, Alphabet, Amazon, they already are hyperscaler. And if they get voted down in like they could go to Alabama or anywhere else, part of the issue,
7:33
Unidentified speaker (speaker_1): that's the problem with an Oracle and a SpaceX that you're seeing play out in the market real time.
7:38
Unidentified speaker (UNKNOWN): And
7:38
Unidentified speaker (speaker_2): the hyper hyperscalers have the installed base of data centers. Right. That just became more valuable because we can't build other data centers.
7:46
Unidentified speaker (speaker_1): That's exactly it's like a collect it's like a collectible. Like why does the value of collectibles continue to go up?
7:52
Steve Eisman: Got it.
7:52
Unidentified speaker (speaker_1): That's exactly it.
7:56
Steve Eisman: I got like 3,000 things I want to talk about. Let's do it. Let's go to Google. So the interpretation, rightly or wrongly. of Google's numbers, and I'm not talking about the cloud part of the business. I'm just talking about the search part of the business, was that Google search revenue growth was 12%.
8:13
Steve Eisman: It was 19% in the first quarter. And the fear around Google for the last two years has been that Google has this incredible business called search, which has insurmountable modes. I
8:25
Unidentified speaker (UNKNOWN): mean,
8:25
Steve Eisman: I've never seen a business. No one's ever seen a business like Google search. To such a degree, it's a paradigm. You can't even think outside, you couldn't even think outside of the concept of Google search. But now you have alternatives. And those alternatives, we could argue about how high those modes are, but those modes are certainly lower than Google search. So you're
8:45
Steve Eisman: exchanging a business where you're basically a monopoly for something where you're competing. It's very capital. intensive. And
8:53
Unidentified speaker (UNKNOWN): the
8:53
Steve Eisman: evidence is now in
8:55
Steve Eisman: Google's Google this is the thesis. Google search growth is slowing because it's moving over to AI. And that's bad. That's that was without question the interpretation of the quarter. I'd be curious as to how you guys actually see it. And and again, I'm not talking about the cloud business, it
9:13
Unidentified speaker (UNKNOWN): was
9:13
Steve Eisman: great, the AI stuff
9:14
Unidentified speaker (UNKNOWN): was great.
9:15
Steve Eisman: great. But
9:17
Unidentified speaker (UNKNOWN): the
9:17
Steve Eisman: crown jewel of this company, without question, has always been search. Gil, take it away.
9:22
Unidentified speaker (speaker_2): This is the quarter we feared. To your point, we've been afraid of this quarter for the last couple of years. Ever since we started using Chat GPT more than Google,
9:30
Unidentified speaker (UNKNOWN): we
9:30
Unidentified speaker (speaker_2): were afraid this quarter would come. But they reported quarter after quarter after quarter of good results. And so investors became complacent. But then all of a sudden, out of the blue, abruptly, to your point, they grew 90%, 70%, 12%.
9:44
Steve Eisman: Right.
9:44
Unidentified speaker (speaker_2): Out of the blue, this happens without them warning us, without them guiding to lower growth. It happens. So it's inconclusive, because by the way, 12% growth is still fine. But it indicates that there may be Maybe something broken. Let's not forget, if this breaks, it's gonna break in slow motion.
10:01
Steve Eisman: No question.
10:02
Unidentified speaker (speaker_2): Because it's ad dollars that we're talking about. It's not eyeballs.
10:06
Steve Eisman: Right. Right?
10:07
Unidentified speaker (speaker_2): It's not consumers don't pay to search Google. Right. So if we're searching less Google, that doesn't matter until the advertisers say, you know what, instead of spending 90% of my search dollars on Google, I now have another choice.
10:22
Steve Eisman: Right.
10:22
Unidentified speaker (speaker_2): And by the way, OpenAI has ramped up their um advertising very requirements. cautiously, very slowly. It's maybe a billion-dollar business right now, which is a very
10:31
Unidentified speaker (UNKNOWN): small part
10:32
Unidentified speaker (speaker_2): of the market. But
10:34
Unidentified speaker (UNKNOWN): at
10:34
Unidentified speaker (speaker_2): the point where they start advertising as much as Google, when their ad load catches up to Google, all of a sudden all the advertisers will switch 10, 20% of their budget
10:46
Unidentified speaker (UNKNOWN): to
10:46
Unidentified speaker (speaker_2): Chat GPT, to OpenAI. And so that's not going to happen overnight. That's slow. motion, but it feels like this happened abruptly. So, first of all, for the first time, Google admitted that they don't generate as much ad dollars on an AI search than they do on a Google search. They'd been denying it up
11:06
Unidentified speaker (speaker_2): until now.
11:06
Steve Eisman: Really?
11:07
Unidentified speaker (speaker_2): And this quarter, they admitted that that's the case. And that's again, that's one piece. The other piece is there's a lot less searches.
11:15
Unidentified speaker (speaker_2): More than half of web traffic comes from chatbots now, right? It's not humans searching anymore, it's chatbots. Are advertising advertisers paying when a chatbot clicks through the site? These are the things you have to worry about. At
11:31
Unidentified speaker (UNKNOWN): what
11:31
Unidentified speaker (speaker_2): point do advertisers go to Google and say, I don't want to pay to have chatbots? They're not actually looking at my site, they're looking at your site. They're not, I don't want to pay cost per thousand based on a chat bot. I'm going to pay a lot less. This is still stuff that could still happen.
11:48
Steve Eisman: So you,
11:48
Unidentified speaker (UNKNOWN): if
11:49
Steve Eisman: I'm going to interpret what you just said, you walked away from that quarter more negative on Google than before.
11:54
Unidentified speaker (speaker_2): Absolutely.
11:55
Steve Eisman: Okay. Yeah.
11:57
Unidentified speaker (speaker_1): No way that you could walk away from that quarter feeling incrementally positive. Now I think that the question will be to Gil's point:
12:05
Unidentified speaker (speaker_1): is this just an airpocket period as painful as it is to when Google figures out the monetization engine on Gemini? They ultimately, when you think about the frontier models, what's happened to OpenAI as well as with Cloud, Claude, do they start to sort of ring fence things where they
12:25
Unidentified speaker (speaker_1): feel like, okay, like we never
12:26
Unidentified speaker (UNKNOWN): have to be able to do that?
12:26
Unidentified speaker (speaker_1): now figured out the way that we're gonna monetize search
12:31
Unidentified speaker (UNKNOWN): and
12:31
Unidentified speaker (speaker_1): we're gonna monetize our ecosystem that Gemini needs to basically narrow the gap with the likes of an OpenAI and anthropic. That's essentially what needs to happen because it's no different than another way what's happened with Palantir and Sovereign AI. And Frontier models are here.
12:48
Unidentified speaker (speaker_1): The so many other companies are here. They're like, should we put a wall up that then this becomes not just our data, but it's our ecosystem to monetize? That's what essentially Alphabet's going to. My view is that the market's baking in a very negative scenario for the advertising. If they just show
13:08
Unidentified speaker (speaker_1): even 30, 40% success rates, I think it's a stock that could go significantly higher. That's sort of my view along with the cloud business. But no that no doubt, this was the quarter. This
13:19
Steve Eisman: was
13:19
Unidentified speaker (UNKNOWN): the
13:19
Steve Eisman: quarter that people were worried about.
13:21
Unidentified speaker (UNKNOWN): Because
13:21
Unidentified speaker (speaker_1): a lot of times it was like bears are called 10 the last two downturns, whatever. Then finally, like this was the one, they come out of hibernation mode and they're like
13:30
Unidentified speaker (UNKNOWN): told
13:31
Unidentified speaker (speaker_1): you. And you can deny. So now they have to prove it otherwise.
13:36
Steve Eisman: Okay. Let's move on because like I said, I got we got so much talk. to about. Imagine I'm an Oracle shareholder. Okay. You
13:43
Unidentified speaker (speaker_2): don't sleep.
13:44
Unidentified speaker (UNKNOWN): I am upset.
13:45
Steve Eisman: I'm upset because I have lost so much money in in in this stock over the last year or so. And I come to you guys and I say, Guys,
13:59
Steve Eisman: help me out. A, explain to me what the hell just happened. to my stock. And B, should I still own it? I'll take you for
14:08
Unidentified speaker (UNKNOWN): I mean
14:08
Unidentified speaker (speaker_1): that's the look, that's the one that like we're most cautious on because my view is like, I know so many people at Oracle over the years. The gold standard. I mean, if if I look at like software or tech, I could go through every company and like a handful of people start at one point, you know, at Oracle. So many. of those people have
14:28
Unidentified speaker (speaker_1): either been cut or left because of all the layoffs that they've done. They've cut into the muscles, the boon, in order for this data center build-out to happen. And that that to me is the issue is like the capital raises never stop, the debt raises never stopped, stop, just a secondary
14:47
Unidentified speaker (speaker_1): this week. And yet the core business.
14:51
Unidentified speaker (speaker_1): is suffering.
14:53
Unidentified speaker (UNKNOWN): The
14:53
Unidentified speaker (speaker_1): hyperscalers are able to raise money, and and there's cuts around, you know, across the board, but they're able to continue to sort of build out their business. Oracle is trying to be a hyperscaler. The OpenAI revenue, I get it. It's just in this market right now, they just I think they just made the wrong bet. I mean, that's sort of my view.
15:13
Unidentified speaker (UNKNOWN): What
15:13
Unidentified speaker (speaker_3): do you mean they made the wrong bet?
15:15
Unidentified speaker (speaker_1): They bet the whole it's no different than like in baseball when a company spends so much of their salary, cap, or whatever.
15:23
Unidentified speaker (UNKNOWN): On one guy.
15:24
Unidentified speaker (speaker_1): On one guy. And then
15:27
Unidentified speaker (speaker_1): after four years, and that guy's played not like a Stanton or whatever, but like he's played like 40 games in four years. You're like, we should have just gone the way of Tampa and Kansas City. and others that have built the farm system. Oracle is a good example. Like they bet hugely on OpenAI with the 300 billion,
15:47
Unidentified speaker (speaker_1): and the businesses suffer significantly, and investors are basically like, How could you own the stock? Knowing every week there's just gonna be another debt raise, another equity rill.
16:01
Unidentified speaker (UNKNOWN): It's a high wire act.
16:02
Steve Eisman: same question.
16:02
Unidentified speaker (UNKNOWN): Yeah, it's
16:03
Unidentified speaker (speaker_2): a high wire act. Here's the thing. We're going to agree that hyperscaler is very good business.
16:08
Unidentified speaker (UNKNOWN): Right.
16:09
Unidentified speaker (speaker_2): Neo cloud's probably bad business.
16:11
Steve Eisman: We'll get to that.
16:12
Unidentified speaker (speaker_2): Guess what's right in the middle of that? Oracle. Oracle. Okay. Right? Because they do have a core business that's very cash generative, right? Oracle databases are not going away. It's a 50% margin business. It's paying the interest on the loans that we're taking to build data. centers. Right. Which is a little bit of what the hyperscalers are doing. It's just that the hyperscalers have such
16:32
Unidentified speaker (speaker_2): better balance sheets. And Oracle already had a challenge balance sheet because of Cerner and other acquisitions they made. So it's a high wire act. If they execute perfectly, the stock is a multiple of where it is today. But it's a high-wire high wire act because if they don't execute well, if OpenAI doesn't materialize, there's a lot of downside.
16:52
Unidentified speaker (UNKNOWN): And
16:52
Unidentified speaker (speaker_1): also you don't come off those conference calls. Like those calls don't instill
16:57
Unidentified speaker (UNKNOWN): confidence.
16:58
Unidentified speaker (speaker_1): Why?
16:59
Unidentified speaker (UNKNOWN): No.
16:59
Unidentified speaker (speaker_1): Why? Why? Because it just feels like there's pilots in the plane. You're not sure.
17:05
Steve Eisman: Just some of you to understand. We're talking about the quarterly conference calls when they give you an update. What why does that not give in him confidence?
17:10
Unidentified speaker (speaker_1): Because you don't understand the direction of the business. Some of the stuff is sort of like long and like word saladish. You feel like is Ted Stryker flying the plane?
17:21
Unidentified speaker (UNKNOWN): The
17:22
Unidentified speaker (speaker_3): references to a movie called Airplane. But
17:23
Unidentified speaker (speaker_1): go a go ahead. And I think that I think that's part of it. It's part of like the intel.
17:28
Unidentified speaker (speaker_1): Oh
17:29
Unidentified speaker (speaker_3): boy.
17:29
Unidentified speaker (speaker_1): You came off there be like, I gotta go to the bar or something like that. Because it you felt like it wasn't just like a soft quarter. You felt like they were just deer in ahead. a headlight. And I think the
17:40
Steve Eisman: let me let me ask you, let's let's switch that to Intel. So Intel had a terrible quarter, the stock got crushed. Just explain what happened. Why was the calling?
17:46
Unidentified speaker (UNKNOWN): But basically
17:47
Unidentified speaker (speaker_1): for three months, given memory prices and given what we see in the PC market,
17:52
Unidentified speaker (speaker_1): and everyone's like, We had we got under control. We feel good. If you're an institutional investor, you're like, okay, they you know the steady hand type thing. Then all of a sudden that was a disaster quarter. and you felt like there's not necessarily no end in sight, and you don't have the confidence of management to sort of navigate that. So I think that like
18:12
Unidentified speaker (speaker_1): when you like look, Gil and I that have done this, you know, for so many years, there's times where companies, they miss a quarter, something happens, but you feel like the confidence is there where it's a one one-quarter quarter issue or a two-quarter two quarters. issue. You don't throw a baby out with the bathwater. Then there's others you're like, could this just
18:32
Unidentified speaker (speaker_1): be a sort of catch a falling knife type stock?
18:35
Steve Eisman: What did you think of the call of Intel?
18:37
Unidentified speaker (speaker_2): It's broken company.
18:40
Unidentified speaker (UNKNOWN): But
18:40
Steve Eisman: it's always been a broken company.
18:42
Unidentified speaker (speaker_2): Yes. Yes, that is the point. But they brought in a very capable CEO.
18:48
Steve Eisman: Right. Li
18:49
Unidentified speaker (speaker_2): Libutan, Bhoutan, very capable. guy, comes in, does the best deals you could absolutely do. U.S. government, NVIDIA, Apple, Elon lines up all the best deals, secures the next node of 14A, gets everybody all excited.
19:07
Unidentified speaker (UNKNOWN): But
19:08
Unidentified speaker (speaker_2): those of us that have been following Intel for a while know that it's a broken company.
19:12
Unidentified speaker (UNKNOWN): It's
19:12
Unidentified speaker (speaker_2): broken.
19:13
Unidentified speaker (UNKNOWN): It's They run like a motor
19:14
Unidentified speaker (speaker_1): vehicle.
19:15
Unidentified speaker (UNKNOWN): Mean
19:15
Steve Eisman: Department of Motor Vehicle.
19:16
Unidentified speaker (UNKNOWN): Yeah.
19:16
Unidentified speaker (speaker_1): I mean it's uh yeah, get a license and motor vehicle. That's that's the vibe at Intel.
19:21
Unidentified speaker (speaker_2): They their gross margins are in the 30s. They compete with companies whose gross margins are either in the 50s or the 70s. Right. That is bad execution at every level of the company. And you can bring in a great CO CEO that can do great deals, but if if you can't make chips at a high enough yield. to get at least a 50%
19:41
Unidentified speaker (speaker_2): margin, you're a broken company. And what really puts the exclamation point on that is AMD reporting. AMD is in the exact same market.
19:51
Unidentified speaker (speaker_3): Right.
19:51
Unidentified speaker (UNKNOWN): And
19:51
Unidentified speaker (speaker_2): they're in the 50s. AMD is in the PC market. They didn't bat an eyelid. Intel collapsed because of the collapse in the PC market. And AMD. barely, barely nudged based on that. And so you just know its execution at every level. It's very hard to fix a
20:11
Unidentified speaker (speaker_2): company that's broken this badly.
20:13
Unidentified speaker (speaker_1): Yeah, Intel feels like it's like going like a White Castle. AMD is like a real nice steak restaurant. And I think it comes out again and again on the calls relative to, you know, and also how Lisa Sue's Seuss position.
20:24
Unidentified speaker (UNKNOWN): She's a
20:24
Unidentified speaker (speaker_3): great C. CEO.
20:25
Unidentified speaker (speaker_1): I mean, she is we're talking about
20:26
Unidentified speaker (UNKNOWN): the C.
20:27
Unidentified speaker (speaker_1): one of the best CEOs, not just in tech, but I'd say really ever as a public company, you know, across the board.
20:35
Steve Eisman: Let's talk about where you guys
20:37
Unidentified speaker (UNKNOWN): stand
20:37
Unidentified speaker (speaker_3): on the debate of
20:38
Steve Eisman: MOTS in AI. Hi, Steve Eisman here. To create
20:43
Unidentified speaker (UNKNOWN): my podcast,
20:44
Steve Eisman: I
20:44
Unidentified speaker (UNKNOWN): do research on
20:45
Steve Eisman: markets, finance, and anything
20:47
Unidentified speaker (UNKNOWN): else
20:47
Unidentified speaker (speaker_3): relevant
20:47
Unidentified speaker (UNKNOWN): every
20:48
Steve Eisman: day. In doing
20:49
Unidentified speaker (UNKNOWN): that research, I use AI
20:50
Steve Eisman: all the
20:50
Unidentified speaker (UNKNOWN): time.
20:51
Unidentified speaker (speaker_3): to retrieve data and to help formulate my ideas. And I'm not alone. AI is changing how every business runs, and the leaders who pull ahead will be the ones who know where it matters and how to use it. And that's exactly what Masterclass Executive is built for. If you've ever wanted to learn business from the people actually shaping it, not just
21:11
Unidentified speaker (speaker_3): studying it, you need to hear about Masterclass Executive. Masterclass Executive is the first AI native business education program built with the University of Chicago Booth School of Business in collaboration with OpenAI. You can learn directly from Fortune 500 CEOs, operators,
21:30
Unidentified speaker (speaker_3): AI builders, and Nobel Prize winning academics, people like Ray Dalio, Mark Cuban, and Paul Krugman. It's designed to fit around a full-time full time job. with self paced coursework and weekly live sessions. If you're serious about leading in the AI era, this is a program I'd point you to. I look through the program and the instructor list,
21:50
Unidentified speaker (speaker_3): and you should really check it out. Seats in each masterclass executive cohort are limited, and applying is free. Apply today at masterclass.comslash Iisman. eisman. That's masterclass.comslash Iisman. eisman, masterclass.com slash eisman.
22:05
Steve Eisman: So I'll give you the negative. It's a very simple simplistic negative case, and you guys tell me what you think. The bare case about LLMs and agentic AI is it's incredible technology, gonna change the world. But it's very, very easy to switch from one model to the next model to the next model. And with respect to anthropic and open
22:25
Steve Eisman: AI,
22:26
Unidentified speaker (UNKNOWN): who
22:27
Steve Eisman: charge a lot more,
22:29
Steve Eisman: you don't need
22:30
Unidentified speaker (UNKNOWN): a
22:30
Steve Eisman: Mercedes-Benz to go to the grocery store. You could take a bike. So you can use these Chinese models. So therefore, a lot of market share is going to go away. There's going to be a price war. There's word that uh Anthropic cut prices recently by 50%.
22:45
Steve Eisman: You can have a technology that's wonderful, but the business itself. is not so great. And that's that's the negative case about Anthropic and OpenAI. And I would add that um if you look at Anthropic's um third quarter, the the revenue growth rate slowed
23:05
Steve Eisman: pretty dramatically, probably because token maxing is ended. And I'm very happy that I know enough to be able to say the word token maxing.
23:13
Unidentified speaker (speaker_1): Actually, when when you say it, you actually say it like you're almost it's like a technical.
23:17
Unidentified speaker (UNKNOWN): It's
23:17
Steve Eisman: almost like I understand what I'm
23:19
Unidentified speaker (UNKNOWN): saying.
23:19
Unidentified speaker (speaker_1): And yeah, and you almost say it in like such a confident way.
23:22
Steve Eisman: So that's the bear case. Um
23:26
Unidentified speaker (UNKNOWN): tell
23:26
Steve Eisman: me what you think about that, and and tell me what you think about that.
23:29
Unidentified speaker (speaker_1): I mean, I don't disagree. Like look, my view models are gonna commoditize To over time. You're gonna have more and more, whether it's open source, vertical, countrywide. in terms of you'll have hundreds of models over the coming years. I think Anthropic and OpenAI, they recognize that, which is why they're so focused on the enterprise, which is why every
23:49
Unidentified speaker (speaker_1): enterprise salesperson I know like has tried to get picked off by one of them because they understand like their moot is the initial on the model side and trying to max out their essential essentially install base. But it's about enterprise implanting themselves and ultimately,
24:08
Unidentified speaker (UNKNOWN): you know,
24:09
Unidentified speaker (speaker_1): and this kind of goes to the debate with Sovereign AI, what's happened with Palantir and NVIDIA and others. They're using it as like to open the door through their models, but their monetization opportunity is going to
24:21
Unidentified speaker (UNKNOWN): be
24:22
Unidentified speaker (speaker_1): multitudes of that on the enterprise. Like when you think about Anthropic, if they could do five 500 billion, 700 billion in revenue, get to a trillion over the next you know eight to ten years, it's not going to be charging people like me and you per month. It's going to be
24:36
Steve Eisman: getting in companies.
24:37
Unidentified speaker (UNKNOWN): Oh,
24:37
Unidentified speaker (speaker_1): yeah. And that's and that is their like that is.
24:39
Unidentified speaker (UNKNOWN): So you so
24:40
Steve Eisman: you agree with the concept of the mode is it's not great.
24:43
Unidentified speaker (speaker_1): I I agree on, but they're using it as Is a front door to get in, then entrench themselves. Okay,
24:50
Steve Eisman: what do you think about this whole thing? debate?
24:51
Unidentified speaker (speaker_2): I have a lot to say. So this is why I'm happy this is long form, because I think this is at the core of everything that's going on.
24:58
Unidentified speaker (UNKNOWN): I
24:58
Steve Eisman: totally agree with that.
24:59
Unidentified speaker (speaker_2): So let me start with the good news. Okay. They do have a moat. Their moat is the team. They have there's only a hundred people in the world. I'm I'm not an exaggeration, a hundred people in the world that are smart enough to move AI forward.
25:14
Unidentified speaker (UNKNOWN): Okay.
25:15
Unidentified speaker (speaker_2): They are heavily concentrated at anthropic, OpenAI, and Google DeepMind. Heavily. The balance is in China.
25:24
Unidentified speaker (UNKNOWN): Of
25:24
Unidentified speaker (speaker_2): those hundred people, I
25:26
Steve Eisman: notice you didn't mention Grok.
25:28
Unidentified speaker (speaker_2): Elon is very smart too.
25:30
Steve Eisman: Okay.
25:30
Unidentified speaker (speaker_2): He's got a good team, so does Zach, but very few in terms of really being able to compete head to head with anthropic and open. OpenAI.
25:39
Steve Eisman: Okay.
25:39
Unidentified speaker (speaker_2): So here's where we're at. That means that those two will continue to build the best models. And Anthropic made a really smart move, as Dan is pointing out, to pivot to the enterprise. And by the second quarter, they had an overwhelming share of the enterprise market in terms of tokens, right?
25:59
Unidentified speaker (speaker_2): And then a couple of things happened. One is OpenAI. caught up, and in the third quarter, they gained share. And the other thing is CTOs were spending so much on tokens
26:11
Unidentified speaker (UNKNOWN): that
26:12
Unidentified speaker (speaker_2): they realized we can't keep doing this. We have to have open weight alternatives. And they started diversifying their AI spend towards that. And another thing is happening, which is the emergence of diffusion. layer, control plane, orchestration layer, harness. I'll I'll introduce those
26:31
Unidentified speaker (speaker_2): concepts in a second. But the point is that if Anthropic started with a market that was this big, right now their market is narrowing because a big part of it's going to open weight. Because to summarize my email, I don't need the frontier model.
26:46
Unidentified speaker (UNKNOWN): There's like NVIDIA
26:47
Unidentified speaker (speaker_1): hugging acquisition.
26:48
Unidentified speaker (speaker_2): And their market is narrowing because companies are starting to put something on top of them.
26:53
Steve Eisman: What do you mean, something on top of them?
26:56
Unidentified speaker (speaker_2): So if you use anthropic
26:58
Unidentified speaker (UNKNOWN): This is
26:59
Unidentified speaker (speaker_1): very important. What he's talking about, this is like very important, like what he's about to say.
27:04
Unidentified speaker (speaker_2): That's
27:04
Steve Eisman: why I'm listening.
27:06
Unidentified speaker (speaker_2): When you as a company sign a deal with anthropic and start using their models in your business, not only are you handing them all of your proprietary information, all of your alpha, you're paying them to take that from you. Because they can tell you till they're blue in the face that they're keeping your privacy. They are not. They
27:26
Unidentified speaker (speaker_2): are not. In order to justify their growth and their valuation, they will need to compete with most of their customers at some point. And so they are now happily taking everybody's information to train their model to know how to compete with. those companies. And so smart companies are going to Palantir, but to some extent to Microsoft and Amazon and saying,
27:46
Unidentified speaker (speaker_2): hey, hey, hey, I need a buffer. Alex Carp refers to this as I need to have protected sex here, right? I need you to come in and have a control plane, an orchestration layer, and a harness. Those are three different types of applications. that sit on top of the model. So my
28:06
Unidentified speaker (speaker_2): business can leverage any model. I can switch back and forth between anthropic and open, I can switch to open weight, and they don't get my proprietary information. This is why Palantir
28:17
Unidentified speaker (UNKNOWN): is
28:18
Unidentified speaker (speaker_2): killing in the market right now. By the way, big driver for Microsoft as well. Starting to be a big driver for Amazon. So
28:23
Steve Eisman: Palantir is kind of like a cop in a way.
28:26
Unidentified speaker (speaker_1): Yeah, the sovereign and the boot camps, and Gil and I have been there together. I mean, right? Like you see it from that's their whole positioning.
28:34
Steve Eisman: Okay.
28:35
Unidentified speaker (speaker_2): So the pie is going to grow very fast. OpenAI is an anthropic slice of the pie is going to shrink. The whole question is going to be: is the pie going to grow faster than the slice is going to shrink? Because they will not be able to provide all of AI, which is what at some point in the second
28:55
Unidentified speaker (speaker_2): quarter, they were providing all the AI.
28:57
Unidentified speaker (UNKNOWN): Right.
28:58
Unidentified speaker (speaker_2): Going forward, they're going to provide the frontier models for mission critical decisions. So am I doing security software? Am I doing trade monitoring? Am I doing predictive maintenance? Things that I'm willing to pay for the best model because it's so important to my business that I'm willing to pay. a premium.
29:15
Steve Eisman: I'm not gonna do that for just plain research. No,
29:17
Unidentified speaker (speaker_2): I'm not gonna do that for eighty to ninety percent of what I use AI for. I'm gonna use an open weight model.
29:23
Unidentified speaker (UNKNOWN): By the way,
29:23
Unidentified speaker (speaker_2): I'm gonna
29:24
Unidentified speaker (UNKNOWN): be able to do that. Because NVIDIA
29:25
Unidentified speaker (speaker_1): does the hugging acquisition.
29:26
Unidentified speaker (speaker_2): That's right. And and I'm gonna buy it from either NVIDIA or Microsoft or Amazon because they we I still need compute. Open weight models still run
29:36
Steve Eisman: compute
29:36
Unidentified speaker (speaker_2): on GPUs, CPUs and memory. And to
29:38
Unidentified speaker (speaker_1): give Gil's point, as he's talking about the different layers, anthropic OpenAI, especially anthropic, like hey, on the cybersecurity side, we want that too. CrowdStrike's like, no, that's us. Paloutto is like, that's us. Zscale is like, that's us. So that's another example like why cybersecurity stocks have done what they've done, because
29:58
Unidentified speaker (speaker_1): now that's no longer
30:00
Unidentified speaker (UNKNOWN): so
30:00
Steve Eisman: if I could just, you know
30:03
Steve Eisman: the anthropic IPO is coming soon. You know, if I were just to take what you just said, I would not be buying the Anthropic IPO because what you're saying is their business box is getting smaller.
30:18
Steve Eisman: Fair interpretation?
30:20
Unidentified speaker (speaker_2): We don't have critical pieces of information. We don't have what their current growth rate is, what they're what they can do next year, and most importantly. we don't have the price. Until we have
30:29
Unidentified speaker (UNKNOWN): the price, we
30:29
Unidentified speaker (speaker_2): can't make that uh distinction. But uh what I'm laying out is a framework that says you have to be you have to be very confident that the pie is growing faster than their slice is shrinking in
30:41
Unidentified speaker (UNKNOWN): order to
30:42
Unidentified speaker (speaker_2): buy into this.
30:42
Steve Eisman: All right, let's switch gears again. Let's go to the what my favorite word in the entire English language, the sasspocalypse. SAS pocalypse.
30:49
Unidentified speaker (speaker_1): Yep.
30:49
Steve Eisman: Okay. So if we go back in time, from the summer of last year through early summer of this year, the thesis was that
30:59
Steve Eisman: AI was going to destroy all software with no exceptions. And when you guys came in, you were much more nuanced about it. And right around the time you were in here, that was It's kind of the beginning
31:10
Unidentified speaker (UNKNOWN): of the
31:10
Steve Eisman: software.
31:11
Unidentified speaker (UNKNOWN): The
31:11
Steve Eisman: beginning of the software rally. You know, if you look at ServiceNow, it went from 230 to 90 to 160.
31:18
Unidentified speaker (speaker_1): Salesforce.
31:19
Steve Eisman: And then ServiceNow reported. And it was and people took that, they took it the stock apart. And so where are we on this? Where do you think we are on the SAS pocalypse?
31:33
Unidentified speaker (speaker_1): Like, Are there issues for software companies? No doubt. You hear from salespeople all the time. it's a different way to sell. They don't have their products at a point where we could talk about Palantir, even a Snowflake. They are so advanced relative to where they are, to just a traditional software company. If you think about, you know, what
31:53
Unidentified speaker (speaker_1): Salesforce ServiceNow, so many of these software companies are doing, Adobe among others, they have like a window right now
32:01
Unidentified speaker (UNKNOWN): to
32:01
Unidentified speaker (speaker_1): basically get this. right and monetize.
32:04
Steve Eisman: Palantir.
32:05
Unidentified speaker (speaker_1): Palantir's already done it. But for ServiceNow, Salesforce, Adobe, you go across, you know, UiPath. I mean, there's so many other software companies. The problem is that you like going back, McDermott's one of the best CEOs out there. You can't just snap the fingers and it happens. This to me on
32:24
Unidentified speaker (speaker_1): ServiceNow was I view it as much more of a painful transition period where they will get on the other side, they'll be able to monetize their install base, and the stocks basically reflecting none of that. So it's a good example where, but going back to Gil, when he talked about Alphabet,
32:44
Unidentified speaker (speaker_1): this was the quarter for ServiceNow service now that you feared. It was the quarter for Alphabet outfit that you feared, right? And I think that's why there was such a violent reaction.
32:53
Steve Eisman: They were back to back. They were like one one one day and the next day.
32:57
Unidentified speaker (speaker_1): Nightmare on Elm Street
32:59
Unidentified speaker (UNKNOWN): last
32:59
Unidentified speaker (speaker_1): few weeks as a tech investor.
33:01
Unidentified speaker (UNKNOWN): What do you think?
33:02
Unidentified speaker (speaker_2): Software, there's a normal distribution of companies in software right now, right? On the the one tail, you have the companies that are truly leveraged to the growth of AI. And those are the companies that have performed best this year, by the way. Not only did do they have accelerating revenue growth, but their multiple expanded back.
33:19
Unidentified speaker (speaker_2): Palantir, Snowflake, Datadog, CrowdStrike, Shopify, JFrog, right? Then you have the other tail, the tail on the other side. Companies that are truly jeopardized by the growth of AI, companies that do workflow, UiPath, Pega, Appian, companies that do Data Center, uh that do Call Center, sorry, NICE, 5.9. five nine. Those are companies.
33:39
Unidentified speaker (speaker_2): that are truly impacted by AI right now.
33:43
Unidentified speaker (speaker_2): The big fad middle is just software companies that AI is actually not impacting them that much one way or another. If if something happened, it's because it happened to the core software business. ServiceNow missed their Their guidance by a little bit, by like 50 bips. Right. But that's enough. Because when investors are on edge, that's
34:03
Unidentified speaker (speaker_2): all you need.
34:03
Steve Eisman: So, what about Adobe?
34:05
Unidentified speaker (UNKNOWN): And
34:05
Steve Eisman: the reason why I ask about Adobe, I read, I first I read this article that was some guy whose name I can't recall basically created the entire Adobe suite for free. And then and then uh I forget where I was. I I was talking to somebody who used it,
34:23
Unidentified speaker (UNKNOWN): and
34:23
Steve Eisman: he said it's good. Yeah.
34:25
Unidentified speaker (UNKNOWN): So
34:25
Steve Eisman: I'm curious what you think about it. Like that was pretty astonishing that one person could recreate the entire Adobe suite. Yeah.
34:32
Unidentified speaker (speaker_2): Well, they could probably create Word and Excel as well, but we're not switching away from Microsoft Office because software companies, that that software as a service, I think people sometimes forget the service part.
34:43
Steve Eisman: Right?
34:44
Unidentified speaker (speaker_2): Adobe doesn't sell us the this code that we use internally. They sell us workflows and and and data governance and um uh and uh uh the ability to generate campaigns and measure them and edit them and share them and collaborate and all these things that we continue to use.
35:03
Unidentified speaker (speaker_2): That we can re I can recreate the code for Adobe. That doesn't mean I can go into an enterprise and sell them, that they're gonna run their whole marketing organization around what I vibe coded.
35:14
Unidentified speaker (speaker_3): I see.
35:15
Unidentified speaker (UNKNOWN): Or
35:15
Unidentified speaker (speaker_2): they've been running on Adobe for 30 years, but hey, I wrote the same code, use my software. That's not happening.
35:22
Unidentified speaker (speaker_1): And I I would just just add one thing, and you understand this world so well.
35:26
Unidentified speaker (UNKNOWN): As
35:26
Unidentified speaker (speaker_1): a hedge fund or uh institutional, even if you're like a long only,
35:31
Unidentified speaker (speaker_1): you can't have the patience to see a lot of these turnarounds. If you have one quarter like a ServiceNow, service now, like an Adobe, like we've seen on some of these others, you basically have to just cut and run on the stock. And I think that's that's added to it is that in this market,
35:50
Unidentified speaker (speaker_1): because of risk and because of the narratives, you can't wait to get onto the other side. And I think that's kind of added to a lot of the pressures that we're seeing on an institutional perspective. Trevor
36:01
Unidentified speaker (UNKNOWN): Burrus
36:01
Steve Eisman: Well, let's turn to neo clouds.
36:03
Unidentified speaker (UNKNOWN): It's
36:03
Steve Eisman: amazing how many new words have been invented in tech in the last two years. So for those viewers who don't know NeoClouds would be companies like Coreweave, IREN, Nebius.
36:14
Steve Eisman: To make a long story short, they basically tell me if this is right, they basically compete with the hyperscalers who are building data centers. They're building their own data centers. They're just not nearly as big. Um, They don't have other business businesses to
36:27
Unidentified speaker (UNKNOWN): support them.
36:28
Steve Eisman: So basically what they do is they they raise debt, they build data centers, they rent it out.
36:33
Unidentified speaker (speaker_1): Take a we work for data centers.
36:35
Steve Eisman: Right. And Coreweed's debt just went to 15%, which I I would think makes it prohibitive to build another data center. So where are these is there any hope for these guys?
36:46
Unidentified speaker (UNKNOWN): I
36:46
Unidentified speaker (speaker_1): mean, like to me, it's like let's say someone like Nebius, Nebbius, I like guess as a qu uh as a neoclad, Neo Cloud, I doubt like that's one because of the debt and because of how their position is probably one that's gonna be like a winner. They'll consolidate more. I think Core Weave gets through this, but let's just be clear. Like, especially also when you go back to like the blue wave and what we're seeing, it's negative for
37:06
Unidentified speaker (speaker_1): NeoClouds. Not all of them are gonna survive. Some will, I think Core Weave and Nebbius Nebbias will ultimately be ones that lead, but you're gonna see you're gonna see a there's a scarcity of capital for that. And for them to to raise the money, it's gonna be incrementally more expensive. But it's a good example of like in this market, neo clouds are
37:26
Unidentified speaker (speaker_1): definitely on the wrong side of this trend.
37:29
Steve Eisman: Gil?
37:30
Unidentified speaker (speaker_2): Let's just start with the fact that in the short term, they're crushing it.
37:34
Unidentified speaker (UNKNOWN): And
37:34
Unidentified speaker (speaker_2): and so if anybody's considering shorting that, keep in mind that they can. sell compute in the spot market at prices so high that they're going to have a windfall over the next couple of quarters. Then let's talk about what these businesses are and what their value is, right? Core weave is a NVIDIA SPV.
37:52
Steve Eisman: You got to say that sentence again because it's so precious. I want you to just say it again.
37:57
Unidentified speaker (speaker_2): Core Weave is an NVIDIA special purpose vehicle. NVIDIA created this whole category. to create competition for their main customers. They didn't want to be beholden to five customers. So they created this category. They seeded it, they funded it, they set them up with
38:14
Steve Eisman: advanced chess.
38:15
Unidentified speaker (speaker_2): Yes, it is. Well, Jensen.
38:18
Steve Eisman: Jensen plays advanced
38:19
Unidentified speaker (UNKNOWN): chess. Jensen TV.
38:20
Unidentified speaker (speaker_1): The godfather of AI. A. Well, he's not just the godfather of AI, he's playing 3D chess. So I
38:25
Steve Eisman: just want to interpret for my viewers. view of what you're you're saying is that prior to these neo clouds, Jensen Wong of NVIDIA is basically selling all his chips to Microsoft and Amazon and Google. And he didn't want to be beholden to just those guys because they're big guys too. So they might have negotiating power over him. So what he basically did, and
38:45
Steve Eisman: you called it a special purpose vehicle, was he set up these neo clouds who would also build data centers. So now there's even more competition for his chips.
38:54
Unidentified speaker (speaker_2): And guess what? Microsoft doesn't mind. Because what happens is Mike, I'm now Microsoft. And to meet demand, I need to put 300 billion into the ground this year. I don't want to do that. I want to keep some of my cash flow. I'm willing to put 200.
39:09
Steve Eisman: Right.
39:09
Unidentified speaker (speaker_2): You know what I'll do? I'll just short-term short term lease the rest from somebody else. because I know that I'll I will have in five years,
39:19
Unidentified speaker (UNKNOWN): I
39:19
Unidentified speaker (speaker_2): will have built enough data centers.
39:21
Steve Eisman: I'll lease it from I'll
39:22
Unidentified speaker (UNKNOWN): lease it from
39:22
Unidentified speaker (speaker_2): them for three to five years. Okay. And
39:24
Unidentified speaker (UNKNOWN): when the
39:25
Unidentified speaker (speaker_2): lease is over, that's fine. I'll have enough capacity and and they can go away.
39:31
Unidentified speaker (UNKNOWN): Right.
39:31
Unidentified speaker (speaker_2): So it's good for Microsoft, it's good for NVIDIA, it's good for all the capital that wanted to chase AI. which was there up until six months ago, right? We're talking, we've talked it, then
39:43
Unidentified speaker (UNKNOWN): talked
39:43
Unidentified speaker (speaker_2): about what's happened to capital since. But up until six months ago, capital was chasing AI.
39:47
Unidentified speaker (speaker_3): Yes.
39:48
Unidentified speaker (UNKNOWN): And
39:48
Unidentified speaker (speaker_2): so it was good for all those people. You know who it's not good for? Core weave shareholders. Because at no point did any contemplate adding value to core weave shareholders. Core weave is an enterprise that generates low single digit returns and borrows at 9%.
40:09
Unidentified speaker (speaker_2): It's a negative
40:10
Unidentified speaker (UNKNOWN): carry.
40:11
Steve Eisman: Now it's 15, but it hasn't borrowed that price yet.
40:14
Unidentified speaker (UNKNOWN): Well
40:14
Unidentified speaker (speaker_2): nine blended, right? Because the secured debt they were able to get down into the into the even into the high I sevens, but to but they do need unsecured debt. And if to your point, if they go to the market now, it's 15 points. Right. So they're the cost of debt's actually going to go up for Corweave. And yet again,
40:34
Unidentified speaker (speaker_2): they're generating low single-digit returns. That is, nobody should be engaged in that enterprise. That means that equity holders own nothing. The debt holders may end up being uh uh made whole, the equity holders own nothing. Trevor
40:49
Unidentified speaker (UNKNOWN): Burrus, Jr.
40:49
Unidentified speaker (speaker_2): By
40:49
Steve Eisman: the
40:50
Unidentified speaker (speaker_2): way.
40:50
Steve Eisman: if I take your interpretation as gospel,
40:55
Unidentified speaker (UNKNOWN): my
40:55
Steve Eisman: conclusion is you know who completely agrees with you? The executives of Corweave. Because they've sold a ton of stock. A ton. I mean more than
41:03
Unidentified speaker (UNKNOWN): that. I'm not talking about like ten
41:05
Steve Eisman: million dollars because
41:05
Unidentified speaker (UNKNOWN): we
41:05
Steve Eisman: had hundreds of millions of dollars worth of stock.
41:08
Unidentified speaker (speaker_2): It's been twenty years since I've seen so much insider selling. Trevor
41:10
Unidentified speaker (UNKNOWN): Burrus, Jr.:
41:11
Steve Eisman: It's astonishing.
41:12
Unidentified speaker (speaker_2): The company the um uh twenty 20 years ago
41:13
Unidentified speaker (UNKNOWN): doesn't make have to be a lot of people. sense. I mean, if if
41:14
Steve Eisman: you're not going to believe in your own company, why should I believe in your company?
41:17
Unidentified speaker (speaker_2): Exactly. Exactly. But I do agree with what Dana's saying, that it that doesn't mean this business can't succeed. Take a company like Nebius.
41:25
Unidentified speaker (UNKNOWN): Yeah, but I think
41:25
Unidentified speaker (speaker_1): Nebbius is a good example of like that's been thrown out with the rest of them, but that's in a much different spot.
41:32
Unidentified speaker (UNKNOWN): Why is
41:32
Steve Eisman: it in a different spot? Trevor
41:33
Unidentified speaker (UNKNOWN): Burrus, Jr.
41:33
Unidentified speaker (speaker_2): They barely have any debt.
41:35
Steve Eisman: Okay.
41:35
Unidentified speaker (speaker_2): And they do care about profitability. This is a very experienced operator
41:42
Unidentified speaker (UNKNOWN): who
41:42
Unidentified speaker (speaker_2): is who is making sure that he's buying these data centers at a good price and he's running them out at a good at an evening.
41:48
Unidentified speaker (UNKNOWN): As opposed to former
41:48
Steve Eisman: crypto dudes.
41:52
Unidentified speaker (speaker_1): Nebbius are they're tacticians the way they have built this out.
41:56
Unidentified speaker (UNKNOWN): The
41:56
Unidentified speaker (speaker_2): analogy is important on on the crypto. Bitcoin miners were a great business while Bitcoin was going up fast, and then they were a terrible. business. While the cost of compute is going up fast, these companies will be okay. It's when the price of compute starts even leveling off that you may see issues.
42:12
Steve Eisman: Apple. Let's talk Apple. They got a new phone. I'm kind of interested, but more importantly, still doesn't seem to have much of an AI strategy, which given bit which, in their defense, given the change in the gestalt about capital intensity, doesn't look so so stupid.
42:32
Steve Eisman: Where's Apple just generally right now?
42:35
Unidentified speaker (speaker_1): I mean, my view, well, first of all, Turnus, given his, you know, comes from the hardware DNA and Cupertino.
42:42
Unidentified speaker (UNKNOWN): It's
42:42
Steve Eisman: a new CEO. Yes,
42:43
Unidentified speaker (speaker_1): a new CEO, John Turnus. He's really an innovator when it comes to hardware. I mean, any hardware that's come out of Apple, you know, call it especially over the last decade, he's played a huge role in. I believe, and the duo, you know, the foldable, that really being the first part of what's gonna be a number
43:03
Unidentified speaker (speaker_1): of years of AI enabled devices and really a hardware innovation cycle that they're gonna Going go to through. My view on Apple, and is it's the reason you know the stock is where it is, almost a C. safety name in a tech storm.
43:17
Unidentified speaker (UNKNOWN): It's a
43:17
Steve Eisman: staple.
43:18
Unidentified speaker (speaker_1): Well, it's like a utility, you know, some could say it's a boomer stock or whatever you want to call it. My my whole view on it is just
43:26
Unidentified speaker (speaker_1): AI
43:28
Unidentified speaker (speaker_1): is just starting in terms of how they're going to be able to monetize on the consumer side. 20% 90% of the world will access AI over the coming years to an Apple device. So when when they monetize on the services side through developers, which is a huge part of their ecosystem, that hasn't even been factored into the stock today. And that's why I
43:48
Unidentified speaker (speaker_1): just think, especially on a light CapEx, they're sitting in just such an unbelievable position. I think the stock reflects it.
43:56
Steve Eisman: Yep.
43:57
Unidentified speaker (speaker_2): They're the safe trade. Right.
43:59
Unidentified speaker (UNKNOWN): Right.
43:59
Unidentified speaker (speaker_2): They're the safe trade.
44:00
Steve Eisman: You don't have to worry about the balance sheet. No,
44:01
Unidentified speaker (speaker_2): you can still sleep
44:02
Unidentified speaker (UNKNOWN): cash at out. night.
44:03
Unidentified speaker (speaker_2): You can sleep at night. To your point, they were criticized for not investing in AI, and now they're sitting pretty because they're letting everybody else
44:11
Unidentified speaker (speaker_2): throw hundreds of billions into the ground or tens of billions into developing frontier models. And as Dana is saying, at the end of the day, how are we using the models? On our iPhone, Phone through our Mac, and when I download Chad Chat GPT and do a $20 subscription, guess what? They get paid 30% of that.
44:28
Unidentified speaker (speaker_1): And there's 1.5 billion reasons because that's the amount of iPhones, 2.5
44:32
Unidentified speaker (UNKNOWN): billion.
44:33
Unidentified speaker (speaker_1): 2.5 iOS devices. So it's like with because install base, no different than like a meta with muse or this. That's installbase is the golden goose for so many of these companies.
44:43
Unidentified speaker (UNKNOWN): All right.
44:43
Steve Eisman: Let's switch gears again. Let's talk about agentic AI, which is the new kid on the block.
44:48
Unidentified speaker (UNKNOWN): Um
44:49
Steve Eisman: you know, my my wife Al. Valerie bought theater tickets through agentic AI. It was a pleasure. But I think what people
44:57
Steve Eisman: in the financial services world, for example, are worried about is agentic AI going in going into your bank account and taking out the excess cash and putting into a money market fund. Or going into Schwab and taking out the excess cash and putting put it into a money market fund. Or it in theory would hurt the companies like Bookings or Progressive
45:17
Steve Eisman: or Geico, because you'd say to if you want to buy auto insurance, for example, you'd say to your agentic AI buddy, I just bought a new car, I want to buy auto insurance. I I'll I'll I don't want to buy one from a company with a credit rating below double A, let's just say. Um what do you got?
45:37
Steve Eisman: Agentic AI doesn't. care about Progressive's brand name, doesn't care about Geico's brand name. It just won't it's gonna give you the best price. So
45:45
Steve Eisman: I think this is coming. On the other hand, you know, JP Morgan won't allow any agentic AIs into the bank. So that's just
45:52
Unidentified speaker (UNKNOWN): the
45:52
Steve Eisman: defense. But I think this is sort of inevitable. But I'd be curious as to what you guys think.
45:57
Unidentified speaker (UNKNOWN): No, I think it's
45:58
Unidentified speaker (speaker_1): look, it's it's gonna
45:59
Unidentified speaker (UNKNOWN): you
45:59
Unidentified speaker (speaker_1): talk about AI, it's there will be companies that you don't even know. today that will it will create categories across industries on the eugenic side, whether it's retail, hospitality, whether it's on the financial services.
46:14
Unidentified speaker (UNKNOWN): But
46:14
Unidentified speaker (speaker_1): to my view is like just given the regulatory and given the trust from an AI perspective, we're still a few years from this becoming
46:22
Steve Eisman: normalized.
46:23
Unidentified speaker (UNKNOWN): I
46:23
Unidentified speaker (speaker_1): I think normalized. And to some extent you can could argue in especially the political backdrop, what we see today, you're gonna have a much higher bar relative to like regulatory. And I think that's something that would be like negative, that will slow roll this.
46:41
Unidentified speaker (speaker_2): Reducing friction creates value, it creates surplus. That story you told is gonna going to create surplus for consumers. And guess what? The good company. companies that provide a good product at a good price and add value will do even better. It will accelerate them winning. That's capitalism. It's beautiful. This is what we want.
47:01
Unidentified speaker (speaker_2): And we want this in every industry. And yes, the companies that are slower moving, that don't provide value, they'll Their d their decline will accelerate. And that's good. We want that. And um
47:14
Unidentified speaker (UNKNOWN): the
47:14
Unidentified speaker (speaker_2): consumer surplus that's going to be created, we can't underestimate that too. Because as consumers spend less time and money on things they shouldn't be spending time and money on, they will spend it elsewhere on places where they do get value.
47:29
Steve Eisman: I'm gonna give you both the chances under the last word. I mean, I would think it'd be safe to say the world's a lot more complicated now than what it was in the in the summer. Where b back then I feel like I we were playing with tiddly wings compared to this. So imagine I'm an institutional investor.
47:47
Steve Eisman: Like, where are we? What what what how should I think about things?
47:50
Unidentified speaker (speaker_1): I mean, my view is just come back from Asia and seeing what we see with chip demand in Taiwan and In Korea, 13 to 1. So even if you could argue some of that comes off, I mean, the demand that's is by, we probably don't have equilibrium from a chip perspective till late 2028, early 2029. The hyperscalers, even though
48:10
Unidentified speaker (speaker_1): CapEx will will be altered a bit in the political landscape, you see the spending on the cloud side because the cloud piece is accelerating at a point no one would have thought even six months ago. The use cases that are playing out from Palantir. to Snowflake to other software companies, my my view is that
48:31
Unidentified speaker (speaker_1): now that we probably know, given the sort of guardrails in terms of this boo tsunami, CapEx is probably peak year 2027. Free cash flow is going to start to significantly rebound 2028. twenty eight. The use cases, the second, third derivative is going to play out across cybersecurity software.
48:51
Unidentified speaker (speaker_1): Valuations, the average tech company call it 19, 20 times, NVIDIA 11, 12 times. So my view is like, yeah, memory stocks and some of these others could continue to be under pressure just given what we're seeing in the broader market. But I actually think like tech stocks are going to see a pretty significant run, not just into year end, but over the next six, nine
49:11
Unidentified speaker (speaker_1): months as the sort of winners get separated from losers.
49:14
Unidentified speaker (UNKNOWN): Yep.
49:15
Unidentified speaker (speaker_2): I actually think that we're headed into a Goldilocks scenario for AI. Because I think that yes, growth of of data centers and CapEx next year will be lower than we're expected now. But that actually extends the cycle. I actually think that that means
49:32
Unidentified speaker (UNKNOWN): we can
49:32
Unidentified speaker (speaker_2): gr if we grow twenty 25% five percent next year instead of seventy, 70.
49:35
Unidentified speaker (UNKNOWN): that
49:36
Unidentified speaker (speaker_2): means the year after that we can grow twenty 25, five and the year after that 25, which means we're not going to do lending. We're going to grow it based on the hyperscalers cash flow. That's so much healthier. And the chip companies will continue to increase their outputs. The hyperscales will continue to build out data centers, and the
49:55
Unidentified speaker (speaker_2): surplus will continue to be created. Credit. Uh The key to me to this whole thing is a first principle exercise about the use of AI.
50:03
Unidentified speaker (speaker_2): Am I using AI more and paying for it more? Do the people in my personal life use AI more and are paying for it more? Are the employers that I interact with using AI more, paying for it more? Do I expect that to happen next year? If I answer yes to all those questions, we're going to build more data centers next year and the year after that, and the year
50:23
Unidentified speaker (speaker_2): after that. Maybe we won't have this exponential growth.
50:27
Unidentified speaker (speaker_2): that is debt fueled, but that was going to be a bad thing anyway. Let's do this in a responsible way. And if we can do that, this entire sector is hugely undervalued because you have Microsoft, Amazon, Google trading basically at a market multiple. You have all these really good chip companies, really good companies, NVIDIA, Broadcom,
50:47
Unidentified speaker (speaker_2): AMD, Micron trading at a half, two. to a quarter of a market multiple. And in this scenario that we're growing data centers 20-25% for the next three to five years instead of 70% next year, those companies are worth a lot more than they are now. So to me, that's Goldilocks. Bring it on, regulate,
51:07
Unidentified speaker (speaker_2): take away the capital, make every data center count, because that means we're going to build them in a responsible way to make meet the real demand that is out there. My real estate agent told me that every response she gets now is edited by AI. My son made the uh his college softball
51:27
Unidentified speaker (speaker_2): team a website in a weekend that tracks every pitch. These are things that are happening every day.
51:35
Unidentified speaker (speaker_2): I have a friend who all of her digital interactions now. she checks with Chad GPT first. Should I answer that way? Should I answer that way? How can I phrase this in a friendly way? These are not people that anybody's asking to token max. These are people that are actually getting good value and are willing to pay the $20 a month. And
51:55
Unidentified speaker (speaker_2): the companies they work at will eventually, through a process of diffusion, figure out how to use this. And this will continue to grow. If it ends up growing in a responsible way, without debt, with data centers that communities want, that's a fantastic thing.
52:10
Unidentified speaker (speaker_1): That's like a TED talk right there. I mean, that that was
52:14
Unidentified speaker (speaker_1): sick. Very inspiring. I know.
52:16
Unidentified speaker (UNKNOWN): Thank you.
52:16
Steve Eisman: Guys, thank you very
52:17
Unidentified speaker (UNKNOWN): much. That was great.
52:19
Steve Eisman: We'll have you back again in six months. I'm gonna make the same prediction.
52:21
Unidentified speaker (UNKNOWN): And
52:22
Steve Eisman: the world will have turned over at least 15 times. There
52:24
Unidentified speaker (UNKNOWN): you
52:24
Unidentified speaker (speaker_2): go.
52:25
Steve Eisman: Thanks for coming. Thanks for being here.
52:31
Unidentified speaker (speaker_2): This
52:32
Steve Eisman: podcast is for informational purposes only and does not constitute investment advice. A host and guests may hold positions and stocks discussed. Opinions expressed are their own and not recommendations. Please do your own due diligence and consult a licensed financial advisor before making any investment decisions.
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 The Real Eisman Playbook

The Real Eisman Playbook

Jensen Huang Resigns, Data Centers Face Political Pushback & Housing Hits New Low

Episode discussion: Eisman cites Jensen Huang resignation and housing data to argue for stock picking amid AI valuation shifts

Speakers’ claims: Steve Eisman argues that the resignation of Jensen Huang, combined with CoreWeave's debt issues and Anthropic's valuation concerns, has removed trillions from market valuations. He contends that Democratic midterm wins have introduced fears of data center moratoriums, contributing to a Nasdaq decline. Eisman describes the housing market as 'locked' due to high mortgage rates and rising ancillary costs, with sales hitting a 16-year low. He reviews earnings for Charter, Skydance, CoreWeave, Nebius, and Cisco, highlighting mixed results and cash flow concerns in the AI sector. Finally, he outlines bull and bear cases for the market, suggesting a slowdown in AI CapEx could help hyperscalers return to cash flow.

AI InfrastructureHousing MarketCoreWeave
27 min1 assets10 claims to verify
The Real Eisman Playbook

My Brutally Honest Thoughts About the AI Bubble

Episode discussion: Steve Eisman outlines AI investment risks, citing massive CapEx, off-balance-sheet financing, and private credit refinancing dangers.

Speakers’ claims: Steve Eisman argues that the AI boom is dangerously concentrated in a few hyperscalers whose failure could trigger a US recession. He highlights the extreme capital intensity of AI infrastructure, noting that a single gigawatt data center costs approximately $30 billion. Eisman points to significant financial risks, including Oracle's debt surge to $155 billion and Meta's use of off-balance-sheet vehicles for data center financing. He also warns of refinancing risks for private credit lenders in 2027 as equity in SaaS companies is wiped out by stock declines. The episode covers four market scenarios, with Eisman assigning a 40% probability to an intermediate case involving CapEx slowdowns.

AI InfrastructurePrivate CreditHyperscaler Debt
40 min7 assets10 claims to verify
All insights