Market Hang

AI infrastructure valuation, political market impacts, and AI adoption risks discussed in post-midterm analysis

55 min Listen to the original Episode page
AI InfrastructureMidterm ElectionsMarket ValuationsHyperscalersPolitical RiskPassive InvestingAI ApplicationsRegulatory Moratoriums

Summary

Speakers analyze the market's reaction to midterm elections, noting that while political uncertainty suppresses returns pre-election, clarity often drives post-election rallies, though they critique the statistical significance of such historical patterns. The discussion highlights a bifurcation in the AI sector, where hyperscalers with strong balance sheets outperform industrial suppliers like Caterpillar, whose valuations have normalized from 35x to 20x earnings. Speakers argue that state-level moratoriums on data centers, such as those in Texas, may delay projects but serve as negotiation tools for local revenue rather than permanent bans. They also contrast the current AI boom, funded by cash-rich 'Magnificent 7' companies, with the 2008 subprime crisis, while warning of hidden concentration risks in passive investment vehicles. Finally, the episode covers practical AI applications and risks, including issues with AI-generated legal advice, bias in therapeutic tools, and the shift from chatbots to autonomous agents.

Episode sections

  1. 0:03 – 1:50

    Market Reaction to Midterm Elections

    Speakers discuss the immediate market reaction to Democratic wins in midterms, noting initial futures drops followed by green markets. They argue that while politics cause short-term volatility, long-term corporate fundamentals driven by secular trends are more important.

  2. 1:50 – 5:25

    Historical Performance and Uncertainty

    Emily cites Bloomberg Intelligence research showing flat returns in the three months leading up to midterms due to uncertainty, followed by an average 7.5% gain in the three months after due to relief. One speaker critiques the statistical significance of post-election rally data, noting it reflects general market upward bias rather than election-specific alpha.

  3. 5:25 – 10:27

    AI Sector Rotation and Valuations

    The discussion shifts to a rotation away from AI-related stocks, specifically data center players and semiconductors, due to expected moratoriums in Texas and Ohio. Speakers note a bifurcation where hyperscalers with strong balance sheets (Microsoft, Meta) outperform 'neo-cloud' data center builders, while valuations for industrial AI suppliers like Caterpillar (CAT) normalize from 35x to 20x earnings.

  4. 10:27 – 15:57

    Impact of State-Level Moratoriums

    Speakers analyze Governor Greg Abbott's moratorium in Texas, suggesting it may delay but not stop data center construction. They argue that state-level politics offer leverage for revenue concessions rather than permanent bans, and that slowing CapEx growth could improve free cash flow for hyperscalers like Amazon and Alphabet.

  5. 15:57 – 19:00

    Limits of Political Spending and Consumer Sentiment

    The conversation addresses Elon Musk's $500 million political spend in Texas, which failed to prevent Democratic wins. Speakers conclude that excessive campaign spending has diminishing returns and that consumer affordability issues, rather than political donations, drove the election outcome.

  6. 19:00 – 23:26

    AMD Earnings and Market Sentiment

    AMD reported doubled data center sales but missed on gross margins, causing its stock to underperform despite beating expectations. Speakers attribute this to a broader 'shoot first, ask questions' sell-off in AI/data center names, arguing that market sentiment is currently driven by political fear rather than fundamental earnings quality.

  7. 23:26 – 24:29

    Bubble Dynamics and Market Structure

    The speakers conclude that frequent 'bubble' calls are a necessary part of market mechanics, providing sellers for bulls to buy from. They argue that for the market to continue rising, new buyers must be convinced to enter, making bearish narratives beneficial for long-term bull markets.

  8. 24:29 – 26:06

    AI Adoption and Market Sentiment

    Speaker 1 expresses long-term bullishness on AI, citing adoption studies from the Federal Reserve Bank of St. Louis, but acknowledges short-term market chop due to financing concerns and lack of immediate earnings focus. They anticipate continued selling pressure until a major clearing event occurs.

  9. 26:06 – 29:17

    Valuation Methodologies and Historical Context

    The discussion shifts to valuation, with Speaker 1 rejecting PE multiples in favor of reverse DCFs to determine what is 'priced in.' Speaker 2 compares the current AI cycle to past technological revolutions like electrification and the dot-com boom, noting that productivity gains often lag initial adoption.

  10. 29:17 – 33:26

    Financing Risks and Concentration

    Speaker 2 argues that the AI infrastructure build-out differs from the 2008 subprime crisis because it is financed by the cash-rich 'Magnificent 7' companies rather than fragmented, over-leveraged individuals. However, they highlight concentration risk in passive vehicles where many stocks are now labeled as 'AI stocks'.

  11. 33:26 – 39:22

    Diversification and Market Efficiency

    The speakers discuss a potential shift away from exclusive S&P 500 concentration toward small caps, mid-caps, and international assets. They attribute this to market micro-efficiency, where hedge funds trade specific stock baskets rather than broad indices, and the closing gap between institutional and retail investor behavior.

  12. 39:22 – 48:03

    AI Security and Healthcare Applications

    Speaker 1 notes CrowdStrike's development of kill switches for rogue AI agents, framing it as a necessary response to the technology it helped create. The conversation moves to healthcare, where AI is seen as transformative for drug discovery but faces challenges in commercialization and the need for strict quality control to avoid errors like those seen in the Theranos scandal.

  13. 48:03 – 49:12

    AI in Therapy and Professional Advice

    Speakers discuss using AI for therapy, viewing it as a tool rather than a replacement for human experts. Concerns are raised about AI reinforcing false narratives and potential safety issues like assisted suicides.

  14. 49:12 – 54:54

    AI in Legal Contexts

    The transcript notes lawyers are frustrated by clients using ChatGPT for legal advice, leading to incorrect case assessments. This dynamic can increase legal costs as lawyers must debunk AI-generated arguments.

Assets discussed

AssetKindStanceContext
AppleAAPLequityneutralI don't know what the math would be, but Apple's up six.
NVIDIANVDAequityneutralWell, yeah, but NVIDIA, Amazon, it's like quality AI, not debt-related AI, is way up today.
AmazonAMZNequitymixedWell, yeah, but NVIDIA, Amazon, it's like quality AI, not debt-related AI, is way up today.
MicrosoftMSFTequityneutralThere is a bifurcation between the hyperscalers who have other businesses like your Microsoft's, your Meta's doing well today, and then the data center players, the neo clouds, kind of not as much.
CATERPILLAR INCCATequityneutralSo the cat commons, G, like the ones that are powering the AI data center.
Meta PlatformsMETAequityneutralAnd like kind of meta's there too, a little bit as well.
S&P 500indexneutralI mean, we've been in a structural period of time where the only place to be was the S P 500.

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

Claims worth verifying

In the three months leading up to midterm elections, S&P returns are flat or slightly positive due to high uncertainty, while in the three months after, they average 7.5% due to relief.

Unidentified speaker ·2:06

How to verify: Check Bloomberg Intelligence research cited by Emily, or calculate S&P 500 returns for the 3-month periods before and after all midterm elections since 1950 using Yahoo Finance or Bloomberg data.

The VIX dropped 13.26% on the day of the election results.

Unidentified speaker ·2:48

How to verify: Check historical VIX data for the date of the 2026 midterm election results (November 4, 2026) on financial data providers like Bloomberg or CBOE.

Caterpillar (CAT) is trading at approximately 20 times earnings, down from 35 times at the beginning of the year, and is down 45% from its 52-week high.

Unidentified speaker ·8:23

How to verify: Check Caterpillar's current P/E ratio and stock price history on Yahoo Finance or Bloomberg for the period around November 2026.

Elon Musk spent $500 million on political campaigns, including $224 million in Texas, but failed to prevent Democratic wins.

Unidentified speaker ·16:07

How to verify: Review FEC filings and campaign finance reports for Elon Musk's political spending in the 2026 election cycle.

AMD doubled its data center sales in its recent earnings report but missed on gross margins.

Unidentified speaker ·19:03

How to verify: Check AMD's latest quarterly earnings report (Q3 2026) for data center revenue growth and gross margin figures.

In 2000, the Galleon Group maintained two separate portfolios: one for a Bush win and one for a Gore win.

Unidentified speaker ·6:16

How to verify: This is a personal anecdote by a speaker; verify via biographies of Galleon Group traders or memoirs from that era, though primary verification is difficult as it is a personal claim.

Hyperscalers like Amazon and Alphabet can achieve positive free cash flow by decelerating their CapEx growth rate, even without cutting absolute spending.

Unidentified speaker ·13:06

How to verify: Analyze Amazon and Alphabet's historical cash flow statements to model the impact of decelerating CapEx growth on Free Cash Flow.

Governor Greg Abbott of Texas implemented a moratorium on data center construction.

Unidentified speaker ·10:33

How to verify: Check Texas state government press releases or news reports regarding Governor Abbott's executive orders or legislative actions in late 2026.

The 'Mag 7' stocks (Microsoft, Meta, Amazon, etc.) have cash and flexibility that allow them to pivot and innovate, such as NVIDIA moving from gaming/crypto to AI.

Unidentified speaker ·15:17

How to verify: Compare the cash reserves and R&D pivot histories of these companies in their annual reports (10-K) over the last decade.

Harris's campaign outspent Trump in the 2024 election but lost.

Unidentified speaker ·16:27

How to verify: Check FEC campaign finance reports for the 2024 US Presidential Election to compare total expenditures of the Harris and Trump campaigns.

The Federal Reserve Bank of St. Louis is publishing adoption studies for AI technology.

Speaker 1 ·24:45

How to verify: Search the Federal Reserve Bank of St. Louis website for recent publications or data visualizations on AI adoption.

AMD is not convincing for short-term performance.

Speaker 1 ·25:28

How to verify: Review AMD's recent stock price action and analyst consensus ratings for short-term price targets.

Reverse DCFs are used to determine what is priced in, rather than PE multiples.

Speaker 1 ·26:12

How to verify: Review investment research reports or public statements from Speaker 1 regarding their valuation methodology.

The Navier-Stokes equation, a 100+ year old mathematical problem, was recently solved using AI.

Speaker 2 ·28:17

How to verify: Search for news articles or academic papers from 2024-2026 regarding AI solutions to the Navier-Stokes existence and smoothness problem.

Core Weave and Nebbius have cash flows from Microsoft supporting their debt.

Speaker 2 ·29:40

How to verify: Review SEC filings or press releases from Core Weave and Nebbius detailing financing agreements with Microsoft.

CrowdStrike stock is up three digits in the past year.

Speaker 1 ·39:58

How to verify: Check CrowdStrike's (CRWD) stock performance data from one year prior to the podcast date (Nov 2025) to November 2026.

CrowdStrike is buying startups for about half a billion dollars to build a kill switch for AI agents.

Speaker 1 ·39:32

How to verify: Review CrowdStrike's SEC filings (Form 8-K or 10-Q) for acquisitions and press releases regarding AI security initiatives.

Elizabeth Holmes of Theranos demonstrated fake blood results.

Speaker 2 ·45:05

How to verify: Review news archives or documentaries regarding the Theranos scandal and Elizabeth Holmes' demonstrations.

There were headlines of suicides assisted by chatbots.

Unidentified speaker ·48:44

How to verify: Search news archives for reports of 'suicide assisted by chatbot' or 'AI assisted suicide' around the time of the podcast release or prior.

Lawyers are getting annoyed that people going through divorces use ChatGPT for legal advice.

Unidentified speaker ·49:12

How to verify: Search for Wall Street Journal articles or legal industry reports discussing 'ChatGPT divorce advice' and lawyer reactions.

AI devices adapt to make users stick around by complimenting them.

Unidentified speaker ·50:10

How to verify: Review technical documentation or research papers on Large Language Model (LLM) alignment and reinforcement learning from human feedback (RLHF) regarding user engagement metrics.

AI can reduce a weeks-long project analyzing SpaceX FCC filings to an hour.

Unidentified speaker ·53:02

How to verify: This is a subjective claim by the speaker; verify by checking the complexity of SpaceX's public FCC filings and estimating manual review time versus AI processing time.

AI images used to have six fingers but now are harder to distinguish.

Unidentified speaker ·53:33

How to verify: Compare early AI image generation samples (e.g., 2021-2022) with recent samples (2024-2026) from models like DALL-E or Midjourney to assess anatomical accuracy.

The speaker uses a program for Google Trends to download terms over the past five years and clusters them using Claude.

Unidentified speaker ·51:21

How to verify: Check the speaker's public portfolio or company disclosures for mentions of using Google Trends data and Claude for consumer narrative analysis.

Do your own research

  • Verify Bloomberg Intelligence research on midterm election market performance and the statistical significance of post-election rallies cited by Emily.
  • Check Caterpillar (CAT) current P/E ratio and historical valuation ranges to assess the claim of normalization from 35x to 20x earnings.
  • Review Texas Governor Greg Abbott's executive orders regarding data center moratoriums and any subsequent legislative negotiations or revenue concessions.
  • Analyze AMD's latest earnings report, specifically comparing data center sales growth against gross margin performance and subsequent stock price movement.
  • Examine Federal Reserve Bank of St. Louis studies on AI adoption rates and their implications for short-term vs. long-term market sentiment.
  • Assess concentration risk in major passive indices (e.g., S&P 500) by identifying how many holdings are correlated with AI narratives despite lacking direct exposure.
  • Review CrowdStrike's recent disclosures on AI security features, including kill switches for rogue AI agents, and their impact on cybersecurity spending.
  • Verify the accuracy of AI-generated legal advice by comparing ChatGPT outputs with established legal precedents to understand potential inefficiencies in client-lawyer dynamics.
Full transcriptexpand
0:03
Unidentified speaker: So in over 15,000 conference rooms this morning, Wall Streeters gathered for a morning meeting.
0:10
Unidentified speaker: You had your PMs and analysts sitting around the table, the junior analysts and the admin standing on the outside, and all 15,000 conversations were the exact same.
0:21
Unidentified speaker: It was okay, Dems won, they swept, all right, whatever.
0:25
Unidentified speaker: How are we gonna make money?
0:27
Unidentified speaker: Okay, so eny me miny Drew.
0:31
Unidentified speaker: Last time you were here, you said that you can't make money trading politics.
0:38
Unidentified speaker: But last night we saw futures getting killed.
0:40
Unidentified speaker: We wake up this morning, they're green.
0:43
Unidentified speaker: Were you right?
0:45
Unidentified speaker: I hope so.
0:46
Unidentified speaker: Look, there's a lot of let me just kind of step back for a second.
0:50
Unidentified speaker: Markets make a lot of smart people, smarter than than I, look very stupid at many points in time.
0:57
Unidentified speaker: But it's something I truly believe, and something I've seen a few times, because I've been one of those analysts and strategists in those meetings before joining Roundhill in my past life.
1:08
Unidentified speaker: And the more we obsess about politics in the short run, we end up kind of missing the long-term, I would say like things that actually drive corporate fundamentals and make money.
1:19
Unidentified speaker: So you you worry about it, but at the end of the day, if you take a really good business person that is tied to a secular trend, politics doesn't really upend the direction of that trend.
1:31
Unidentified speaker: They might slightly change the slope.
1:33
Unidentified speaker: So to me, like, look, I personally have never made any money trading politics.
1:38
Unidentified speaker: Right.
1:38
Unidentified speaker: And if I think of a color, like people always say, like purple's the neutral color, a little bit of red, a little bit of blue.
1:45
Unidentified speaker: I think the really only neutral color out there is green.
1:49
Unidentified speaker: Good, I like that.
1:50
Unidentified speaker: You uh Emily, I think you had a piece this morning.
1:52
Unidentified speaker: Um, and I I said it on Monday.
1:54
Unidentified speaker: I said, you know, the market can price you know policy, but it can't price uncertainty.
2:00
Unidentified speaker: And you had something very similar.
2:02
Unidentified speaker: Yeah, I um I wrote about this great research Bloomberg Intelligence had.
2:06
Unidentified speaker: They looked at the three months leading, looked at SP returns in the three months leading up to a midterm election and three months after.
2:13
Unidentified speaker: And it was pretty consistent where in the three months leading up to the election, when uncertainty is really high and no one knows what's gonna happen, even if they pretend that they do, returns are kind of meh.
2:22
Unidentified speaker: They're like a little positive, but basically flat.
2:26
Unidentified speaker: Um and the three months after, they're more like 7.5% on average because there's just there's relief in knowing, right?
2:33
Unidentified speaker: There's and we're seeing that I think at a fairly extreme level today.
2:38
Unidentified speaker: We're seeing, you know, NASDAQ is up around 2%, something like that, SP is up, VIX is way, way down.
2:44
Unidentified speaker: Um something last time I looked, I can just turn my head.
2:48
Unidentified speaker: But um yeah, I don't see there it is, VIX down 13.2.6 percent.
2:53
Unidentified speaker: I mean, I think we're seeing a lot of really that's two points on the VIX, though.
2:58
Unidentified speaker: We don't we don't really quote it as well.
3:00
Unidentified speaker: It's like a zoop down and to the right.
3:02
Unidentified speaker: I mean, it's it's real.
3:03
Unidentified speaker: And I think some of it is just now we know, and some of it is just the um how extreme the win was.
3:10
Unidentified speaker: I don't think anyone was really expecting it to be such a sure win for the Democrats, right?
3:16
Unidentified speaker: I think the the biggest surprise for me was the amount of people conceding.
3:20
Unidentified speaker: I Right.
3:21
Unidentified speaker: Right.
3:21
Unidentified speaker: That was my biggest fear, you know.
3:23
Unidentified speaker: But was anyone surprised?
3:25
Unidentified speaker: That was the turning point last night.
3:27
Unidentified speaker: I just don't think it was close enough to do the whole I'm gonna litigate for 10,000 votes.
3:34
Unidentified speaker: Right.
3:34
Unidentified speaker: Yeah, it was so decisive.
3:36
Unidentified speaker: I think the turning point, I think, would have to look at the charts, but um, when the president talked to the New York Times and said, basically, I don't have the quote in front of me, but like, okay, fine.
3:47
Unidentified speaker: Like, I was on live last night with international investors and international media as this was playing out.
3:55
Unidentified speaker: So it was funny.
3:56
Unidentified speaker: Like, quote, at the beginning of the night was I think this is oversold.
3:59
Unidentified speaker: I think like people are gonna wake up eventually and realize like we're not gonna trade politics anymore.
4:04
Unidentified speaker: And then the real turning point, the clarity point, came, and it's like back to you know, a quote, ages well in what, 12 hours at this point now.
4:12
Unidentified speaker: But like it, it's just clear clarity helps.
4:16
Unidentified speaker: But it's funny that that statistic you talked about, and I don't know about you guys, but I've manipulated a lot of stats on Wall Street in my prior 10 years.
4:26
Unidentified speaker: There's nothing I hate more than 7 out of 7 or 10 out of 10 or whatever small insignificant sample size.
4:33
Unidentified speaker: Like, yes, a lot of people have pointed like out this table or chart that you always make money after midterm elections.
4:40
Unidentified speaker: It's not statistically significant.
4:42
Unidentified speaker: It is not.
4:42
Unidentified speaker: Why have you made money?
4:44
Unidentified speaker: Because when you've measured anything in the US market for the past 50 years, it's gone up.
4:49
Unidentified speaker: Well, guess what?
4:50
Unidentified speaker: Most of the time you lose money in Japan because they had 30 years of you know, just lost performance.
4:56
Unidentified speaker: So everything we do, I think, is is a little bit skewed when we do.
5:00
Unidentified speaker: Did you keep your eye on betting markets?
5:02
Unidentified speaker: A little bit.
5:03
Unidentified speaker: I mean the prediction markets were suggested.
5:05
Unidentified speaker: That's what I meant.
5:06
Unidentified speaker: It was gonna be a blue wave.
5:06
Unidentified speaker: I mean, I think it was evident it was gonna happen.
5:08
Unidentified speaker: I think how overwhelmingly it happened was pretty much a shock and all, especially because two years ago there was a mandate for the Republicans to kind of take the lead, and now the mandate is swift uh shift tremendously.
5:20
Unidentified speaker: Like the pendulum just swung back.
5:22
Unidentified speaker: And I think about you know about politics in the short term and impacting markets.
5:25
Unidentified speaker: I think of Kane's great quote, right?
5:28
Unidentified speaker: In the short term, it's a the market's a voting machine.
5:31
Unidentified speaker: In the long run, it's a weighing machine.
5:33
Unidentified speaker: And but I have a third thing aspect to that, it's it's always a behavioral machine.
5:37
Unidentified speaker: There's always human beings make decisions, right?
5:39
Unidentified speaker: So in the near term, yeah, people are trying to price.
5:41
Unidentified speaker: And if I looked at my portfolio this morning, I saw everyone selling anything related to day data centers, AI, because the moratoriums that are expected coming out of Texas and potentially out of Ohio.
5:51
Unidentified speaker: And then anything related to potentially consumer because you know, Democrats are in, they're gonna be sending packages to the consumer.
5:56
Unidentified speaker: Anything related to rates with rates going down or rising.
5:59
Unidentified speaker: So you do have that initial reaction, especially with uncertainty.
6:03
Unidentified speaker: It's it's it's shoot first, ask questions.
6:05
Unidentified speaker: Although um, I when I thought about it this morning, I was thinking back to uh in 2000, I was at a hedge fund called the Galleon Group, and we literally had two different portfolios.
6:16
Unidentified speaker: We had a Bush portfolio, I was trading healthcare at the time, and a gore, and it was literally two different portfolios, and you don't really see that this time around.
6:27
Unidentified speaker: Like normally you'll see I'll get emails saying, you know, you should buy these stocks if you know if these people win.
6:34
Unidentified speaker: And I don't know, I I didn't see it as much.
6:36
Unidentified speaker: I think also 25 years ago things there was a there is so I agree with with actually both people, and I markets go up.
6:42
Unidentified speaker: But not me and you.
6:44
Unidentified speaker: Eight out of ten years markets go up, so it's like baseball.
6:46
Unidentified speaker: It's wins above replacement, right?
6:48
Unidentified speaker: It you know, if the buys to go higher.
6:50
Unidentified speaker: You got to do better than eight out of ten for it to be significant.
6:53
Unidentified speaker: But I do think that just knowing is a thing.
6:56
Unidentified speaker: Yeah.
6:57
Unidentified speaker: I think you're right.
6:58
Unidentified speaker: Yeah.
6:58
Unidentified speaker: And but we are seeing a rotation, right?
7:00
Unidentified speaker: Like you're healthcare consumer discretionary, good, uh tech uh uh semicap equipment, terrible.
7:07
Unidentified speaker: Right.
7:07
Unidentified speaker: So we're just seeing this backlash.
7:09
Unidentified speaker: Like the the NASDAQ's up.
7:11
Unidentified speaker: I don't know what the math would be, but Apple's up six.
7:13
Unidentified speaker: It's like the anti-AI trade.
7:15
Unidentified speaker: Right.
7:15
Unidentified speaker: Well, yeah, but NVIDIA, Amazon, it's like quality AI, not debt-related AI, is way up today.
7:21
Unidentified speaker: Yeah, that's so interesting.
7:22
Unidentified speaker: There is a bifurcation between the hyperscalers who have other businesses like your Microsoft's, your Meta's doing well today, and then the data center players, the neo clouds, kind of not as much.
7:35
Unidentified speaker: Like your Nebbius's, your you know, the companies you haven't heard of are kind of like, um, there's a lot of people.
7:39
Unidentified speaker: I think a lot of people have heard of them at this point.
7:42
Unidentified speaker: I mean, maybe.
7:44
Unidentified speaker: And people watching.
7:45
Unidentified speaker: Right.
7:45
Unidentified speaker: Watching this.
7:46
Unidentified speaker: But do you remember in 2016 when when Trump took office for the first time, the markets went gangbusters there, remember?
7:52
Unidentified speaker: Because of the tax incentives are just gonna roll out, and it stayed that way because that came through.
7:57
Unidentified speaker: The question comes now, are these moratoriums gonna stick?
8:00
Unidentified speaker: Is there gonna be pushback?
8:02
Unidentified speaker: And you know, ultimately the market has a fantastic way of figuring these things out and going where the money structurally is gonna flow.
8:10
Unidentified speaker: And soon I I we don't know what's gonna happen.
8:12
Unidentified speaker: There's something there's a lot of uncertainty that, and as a result, these these companies were selling at premium lofty valuations.
8:17
Unidentified speaker: I think they're just coming back down to a more normalized level.
8:20
Unidentified speaker: I was just looking at my beloved industrials, right?
8:23
Unidentified speaker: At the beginning of the year, they're at like 35 times.
8:26
Unidentified speaker: So the cat commons, G, like the ones that are powering the AI data center.
8:30
Unidentified speaker: They're like 35 times earnings.
8:32
Unidentified speaker: Cat's down 45% from its uh 52-week high.
8:35
Unidentified speaker: Now it's back to 20, now the group is back to like 20 times earnings, which is historically reasonable.
8:41
Unidentified speaker: Uh so so like the the the other part of this equation is where devaluations go, but on the moratorium side, like you know, it's funny, we were afraid that things were going too fast and the capital markets couldn't finance it all, and that free cash flow at Alphabet was negative.
8:56
Unidentified speaker: So now we're gonna slow down, we're gonna release pressure on, and we still we still sell everything.
9:01
Unidentified speaker: I would personally bet that we still have AI data center development slightly slower.
9:09
Unidentified speaker: And then now we have to work out should cat have been a thousand?
9:13
Unidentified speaker: Should now it be five fifty?
9:15
Unidentified speaker: You know, you know what it's not trading.
9:16
Unidentified speaker: I would probably know where it should be in between those.
9:19
Unidentified speaker: You know what's not trading at those multiples?
9:20
Unidentified speaker: Small cap and mid-cap players in industrial space.
9:23
Unidentified speaker: I mean, a lot of these guys are single digits and they're gonna be kind of the you know the picks and shovels building all this.
9:30
Unidentified speaker: Right.
9:30
Unidentified speaker: You know, so again, yeah, some of them don't have pricing power the way some of the bigger guys do.
9:34
Unidentified speaker: They don't, but when you're talking about 1.2 trillion of CapEx spend that's still anticipated to go online next year, I mean that's gonna flow down to these smaller guys, and it doesn't take the needle much to move their earnings and cash flow.
9:44
Unidentified speaker: So for me, I get excited because I don't think anything structurally has changed.
9:47
Unidentified speaker: Like the moratoriums may come and go.
9:49
Unidentified speaker: I mean, there's a lot of issues that may come about it.
9:51
Unidentified speaker: But the reality is, and I think you made that point, the cat's out of the bag.
9:55
Unidentified speaker: We're the train's moving, right?
9:56
Unidentified speaker: These data centers are getting built.
9:57
Unidentified speaker: So the only thing that changes that the business of the slope, not the direction, right?
10:03
Unidentified speaker: The only thing that changes is that if there's a better mousetrap, if they figure out a way where they don't need these data centers, then I start getting worried about it.
10:09
Unidentified speaker: If if if AI creates a better way, like, oh, we're more efficient in creating these LLMs, then I start reevaluating.
10:16
Unidentified speaker: They're already thinking about it now, because at the end of the day, like computes in short supply, no matter how you cut it, they're already talking about photonics and optics, making a data center better and faster so you can process more throughput.
10:27
Unidentified speaker: And then let's just like super high-level step back, just one quick second on the moratoriums.
10:33
Unidentified speaker: Greg Abbott put one in.
10:35
Unidentified speaker: I know.
10:40
Unidentified speaker: This is a lot of reporting that.
10:47
Unidentified speaker: It was gonna come off and now and now it's now it's gonna stick longer.
10:51
Unidentified speaker: But what's really funny at the local level, and you'll definitely know this better than I do because of like your reporting and what you do, when you get a Democrat or a Republican in a state that's kind of the opposite color, they end up end up being a little bit more centrist.
11:07
Unidentified speaker: And like a Democrat running a state can't deficit spend the way a federal government Democrat or Republican could spend.
11:15
Unidentified speaker: And guess what?
11:16
Unidentified speaker: If you need some tax revenue, yeah, maybe the moratorium is not about never building these things, but where?
11:23
Unidentified speaker: And there's a little bit of like pushback from both sides, but at the end of the day, the state and local level is where this battle plays out more than the federal level.
11:32
Unidentified speaker: And I think those are the people that are a little bit more sensitive to having tack remote.
11:37
Unidentified speaker: I think that's right.
11:38
Unidentified speaker: I think it's more, it's it's not like um bans forever.
11:42
Unidentified speaker: It's more like, let's see what we can get out of these data center developers.
11:45
Unidentified speaker: A lot of the development is already set for 2027, so now it's like, what can we, what can we make happen in 2028?
11:52
Unidentified speaker: What concessions can we get from these uh data center developers?
11:56
Unidentified speaker: There's just more leverage now for these governors to sort of rest some deals and concessions that they're and before you had to give away ta like free taxes for 10 years and no property tax against somebody to build something.
12:07
Unidentified speaker: Now you can actually get revenue.
12:11
Unidentified speaker: When uh I'm a Canadian back when Oh, there's two of us.
12:15
Unidentified speaker: I remember when John Cretchen made like a campaign promise to look at the GST sales tax, and then he got elected, and then he was like, I looked at it.
12:23
Unidentified speaker: I like it.
12:25
Unidentified speaker: I think that was the avid approach on moratoria.
12:28
Unidentified speaker: Um you guys all brought something up that um my awesome producer Romana uh showed me a tweet this morning that I want to read to you guys, and it kind of touches based on all of this.
12:37
Unidentified speaker: It says, uh it's by Michael Antonelli.
12:39
Unidentified speaker: It says, Can we talk why this might be good for hyperscalers?
12:43
Unidentified speaker: If you slow down data centers, you are essentially pulling up the drawbage over the moat for these companies.
12:49
Unidentified speaker: They already have compute, they win even more now, which it's it is right in the markets trading that today.
12:56
Unidentified speaker: But then the other piece to this, because you talked about negative free cash flow, and you actually don't need to cut capex for them to get to positive free cash flow.
13:06
Unidentified speaker: You need them to grow it at a slower pace.
13:09
Unidentified speaker: And I think like a stock like Amazon, people are talking a lot about that now, and that's probably why it's it's moving really aggressively today.
13:16
Unidentified speaker: Like that's a type of stock where, like all of them, like all of them, if they cut free cap, like cut capex tomorrow, they have positive free cash flow.
13:24
Unidentified speaker: Does that whack the stock?
13:25
Unidentified speaker: We can debate that, but these guys are going to get there if they decelerate the rate at which they grow CapEx.
13:33
Unidentified speaker: It's also natural things in place that prevent it from going gangbusters to begin with.
13:38
Unidentified speaker: If you look at Oracle, the Project Jupiter thing and Bloom Energy, the thing that was holding that up was a permit, right?
13:43
Unidentified speaker: I mean, like you get we forget that we want this infrastructure today, but it's gonna take time to rewire the entire grid, especially because a lot of it is antiquated.
13:51
Unidentified speaker: And you're talking about this is massive energy consumption here, and we haven't really gone to the alternative energy future that we envision.
13:57
Unidentified speaker: And so in the near term, you're gonna need NAT gas, you're gonna need some old fossil fuels to kind of drive it all.
14:02
Unidentified speaker: You need better power lines and better power lines.
14:08
Unidentified speaker: We like with the hyperscale, this is like a rogue question with an Amazon.
14:11
Unidentified speaker: I've never when they have these giant free cash flow engines attached to an AI business.
14:16
Unidentified speaker: I never know if it's like the AI backlash.
14:18
Unidentified speaker: Like, are we is Amazon doing well because the consumer is gonna feel relief, or is Amazon doing better because free cash flow is coming back?
14:25
Unidentified speaker: You know what I mean?
14:26
Unidentified speaker: It's like Google, uh Google search was being camitalized by its own cloud AI business.
14:31
Unidentified speaker: Now that slows down, so I I find it tough to figure out the Mag 7 when you have this will be good for other than Elon because he he was a loser based on Texas politics.
14:41
Unidentified speaker: Oh yeah, we should talk about Elon.
14:42
Unidentified speaker: But I was just gonna say this is good in terms of borrowing in the bond market.
14:47
Unidentified speaker: There was already like a little bit too much supply of of bonds from these for hyperscalers of borrowing a lot, and investors were saying, like, this is enough.
14:56
Unidentified speaker: Like there was a little less interest yields are sort of going up.
14:59
Unidentified speaker: I think that slowing down is pretty much is is a good for investors and even for these companies who now pay a little less, perhaps, for cost of capital.
15:09
Unidentified speaker: The the thing they all have in common and why that moniker Meg 7 has stuck around so long is they have cash and flexibility.
15:17
Unidentified speaker: So cat like cash is king, and if you have an innovative culture with a whole bunch of cash, you can surprisingly be nimble for a big business and they weren't before.
15:26
Unidentified speaker: Like NVIDIA, that's an AI stock, right?
15:29
Unidentified speaker: But wasn't that a gaming stock before?
15:32
Unidentified speaker: And wait, and wasn't that a Bitcoin miner?
15:34
Unidentified speaker: Yeah.
15:35
Unidentified speaker: So like talk talk about like pivots.
15:37
Unidentified speaker: Like I understand they're still designing chips and doing different things, but when you have cash and you have an innovative culture, you can do a lot of things and pivot and move.
15:48
Unidentified speaker: And where value where investors think of value is accruing is is going to be different over time.
15:54
Unidentified speaker: And if you can get there, it's it's it's good for your stock price.
15:56
Unidentified speaker: Right.
15:57
Unidentified speaker: I don't know.
15:57
Unidentified speaker: Um, you mentioned Elon.
15:59
Unidentified speaker: Uh a few cycles ago, we learned that celebrities and influencers really can't move the needle with with politics, right?
16:07
Unidentified speaker: Last night, you know, Elon spent half a billion dollars, 224 million in Texas.
16:14
Unidentified speaker: He moved his companies to Texas, and he got a big L.
16:18
Unidentified speaker: Like, can donors move the needle on politics still?
16:22
Unidentified speaker: It's a good question.
16:23
Unidentified speaker: I mean, I feel like we learned this a little bit in 2024.
16:27
Unidentified speaker: Um, Harris's campaign, I believe, outspent Trump and obviously lost, and now we're seeing it again on the flip side.
16:34
Unidentified speaker: And it is kind of interesting.
16:35
Unidentified speaker: Like, is there like a max out on the money that big donors are spending to the point where, I mean, I don't know.
16:41
Unidentified speaker: I I live lived in a uh hotly contested district, and everyone I talked to was just like, when will the spam stop?
16:49
Unidentified speaker: When will the mail stop?
16:50
Unidentified speaker: Like at a certain point, you're just throwing bad money after good, good money after bad.
16:54
Unidentified speaker: I don't know how that actually makes me think of the raging Cajun.
16:56
Unidentified speaker: Remember Jimmy Carvel back in the 90s?
16:58
Unidentified speaker: Yeah, of course.
16:59
Unidentified speaker: Remember what is what is what he said?
17:00
Unidentified speaker: Yeah, I didn't he have a viral tweet yesterday.
17:03
Unidentified speaker: I don't know if he did, but what was the thing back in the 90s?
17:05
Unidentified speaker: Did anyone know what he'd say?
17:06
Unidentified speaker: The economy stupid.
17:06
Unidentified speaker: It's the economy stupid.
17:08
Unidentified speaker: It's the consumer stupid.
17:09
Unidentified speaker: Yeah.
17:09
Unidentified speaker: And I think that's what it is.
17:10
Unidentified speaker: I think affordability is was on the ballot.
17:13
Unidentified speaker: And I think people, there's a big part of the uh the constituents out there that have not benefited with this AI boom and the market's going gangbankers because they don't have assets and they've been left behind.
17:23
Unidentified speaker: And I think as a result of that, doesn't matter how much campaign money you're throwing out of it, if the consumer is unhappy, they're gonna let you know.
17:29
Unidentified speaker: And they they they let us know in this election.
17:32
Unidentified speaker: And I think you know, the Road Act that got passed earlier this year is trying to address some of the housing affordability.
17:36
Unidentified speaker: But, you know, if rates don't come down, if we don't figure out a way to have cheaper housing out there, and if we don't somehow kind of cool down inflationary pressures that are coming from this AI cap spend, the energy uh bill that you know everyone's feeling, you know, we're gonna have a problem structurally in the economy if we can't address the consumer affordability issue going forward.
17:55
Unidentified speaker: So I think that offsets any political money you're throwing at campaigns.
17:58
Unidentified speaker: Right, right.
17:59
Unidentified speaker: And I and to some extent the problem is it's now table stakes.
18:02
Unidentified speaker: Like, yes, Kent Paxton got 150 or spent 150 or 200 million, Telurico spent tens of millions, like there's so much money.
18:10
Unidentified speaker: It's like I wish we could put it to productive use instead of pouring it into campaigns.
18:15
Unidentified speaker: But for fun, I tried to do the math, and Elon's half a billion is like probably us donating about 200 bucks.
18:24
Unidentified speaker: He'll be okay.
18:25
Unidentified speaker: But we're but people are just desensitized to the well of an ego thing.
18:30
Unidentified speaker: Yeah, you're just not getting any value for the money you're spending after a certain point.
18:33
Unidentified speaker: Like how many flyers is gonna change someone's mind?
18:36
Unidentified speaker: Probably no flyers.
18:41
Unidentified speaker: Yeah.
18:41
Unidentified speaker: Like it the kind of running joke.
18:43
Unidentified speaker: Yeah, it's just it it feels like burning money a little bit, but a lot of traditional ad spend, even like if we get off politics, has been been tough.
18:50
Unidentified speaker: It's been tough.
18:51
Unidentified speaker: Right.
18:51
Unidentified speaker: But wait, so I I have a really important question.
18:54
Unidentified speaker: Am I not getting my five thousand dollar check?
18:57
Unidentified speaker: No, not yet.
18:58
Unidentified speaker: No, he didn't he didn't win.
19:00
Unidentified speaker: Um, so let's talk about AMD.
19:03
Unidentified speaker: They reported last night, doubled data center sales, stocks was jumping this morning.
19:08
Unidentified speaker: I don't see it up there on the board.
19:10
Unidentified speaker: Not up as much now.
19:11
Unidentified speaker: Okay, it's not five percent.
19:13
Unidentified speaker: So um, you know, you know, when you go to a concert, there's always sort of like a warm-up band right before the main event.
19:21
Unidentified speaker: And AMD kind of is it giving warm-up band vibes?
19:26
Unidentified speaker: I think Jerry Seinfeld's opener, I think he just flopped intentionally before the next guy came on, uh that old Seinfeld episode.
19:34
Unidentified speaker: Right.
19:35
Unidentified speaker: Um and guess what?
19:36
Unidentified speaker: Everyone else flopped before AMD.
19:39
Unidentified speaker: And uh they're not really the warm-up act, the Mag 7 was.
19:43
Unidentified speaker: And I'm gonna be honest with like this.
19:45
Unidentified speaker: You're saying Mag 7 was the warm-up act you invented.
19:48
Unidentified speaker: Yep.
19:48
Unidentified speaker: I would say All right, that okay, that's fair.
19:50
Unidentified speaker: And I would say like AMD Wait, so yeah, what is yeah, give me the analogy.
19:54
Unidentified speaker: Look, I like the warm-up act flopped.
19:57
Unidentified speaker: So people, mind you, hot take, I actually don't think that report was as good as everyone thinks it was.
20:03
Unidentified speaker: Um I think like earnings have not been good to this point.
20:08
Unidentified speaker: And it's funny when I look at it, everyone knows like the running joke, everyone beats Wall Street expectations.
20:14
Unidentified speaker: We kind of get that.
20:15
Unidentified speaker: We all know that.
20:16
Unidentified speaker: But the street is really good at something.
20:18
Unidentified speaker: And again, I'm I'm coming from that world, and what matters and what the street is really good at is profitability.
20:25
Unidentified speaker: They are so spot on on gross margins.
20:28
Unidentified speaker: It's like I mean, within fractions.
20:31
Unidentified speaker: If you miss on gross margins, that I feel like right now is like a focal point on I would say durable demand.
20:39
Unidentified speaker: If do I have pricing power, do I have like really good end demand that I can keep my gross margins up?
20:45
Unidentified speaker: They missed on gross margins.
20:48
Unidentified speaker: Everyone always beats on sales a little bit, they beat on earnings more.
20:52
Unidentified speaker: There's you can kind of move stuff from when it gets booked in one quarter versus the uh the you know the next.
20:58
Unidentified speaker: Accountants are surprisingly creative.
21:00
Unidentified speaker: Was that an inflationary issue?
21:02
Unidentified speaker: Was that a what was that?
21:03
Unidentified speaker: Do you think?
21:04
Unidentified speaker: Well, they they thought they did really well on cost, but at the end of the day, and we'll go back, even like the memory trade, same thing.
21:11
Unidentified speaker: If you keep pushing price and you get nervous about pushing price to a point where you're gonna kill demand, you're gonna see that kind of in your gross margins come down or be below expectations.
21:22
Unidentified speaker: So I think like AMD last night, again, watching this stuff after hours and not really sleeping, so I'm a little caffeinated today.
21:30
Unidentified speaker: Um, it was up like nine percent.
21:33
Unidentified speaker: Right, and this is also looking at a rearview mirror, right?
21:36
Unidentified speaker: I mean, yeah.
21:37
Unidentified speaker: But that gross margin number stood out to me because they missed on it.
21:41
Unidentified speaker: I I was gonna agree with you on one and disagree with you on the other.
21:43
Unidentified speaker: So, on gross margin, if you look at Zebra Technologies yesterday, which is a really small company, but they had men are uh they make uh little scanners.
21:51
Unidentified speaker: Robotics?
21:52
Unidentified speaker: No, they make little scanners.
21:53
Unidentified speaker: Like when you see people in Home Depot and they're like, Where is where is the nut?
21:56
Unidentified speaker: It's like a zebra.
21:58
Unidentified speaker: Anyway, they missed my memory chips, gross margins were pressured, and and the and the outlook for 27 was a little wonky on margins.
22:05
Unidentified speaker: So people sold the stock, even though they said their business was okay.
22:08
Unidentified speaker: So I agree.
22:09
Unidentified speaker: I actually think earnings have been okay.
22:11
Unidentified speaker: Um I just don't think anybody cares about third-quarter earnings.
22:14
Unidentified speaker: I don't think anybody cares what AMD just said, man.
22:17
Unidentified speaker: Because we're in a data center.
22:18
Unidentified speaker: Oh my god, so many companies report this morning and that they did really well.
22:22
Unidentified speaker: They beat, they raise, and normally they'd be up 10, 15 percent in the small cap world, and they were down because they have anything related to data centers, anything was related to AI.
22:30
Unidentified speaker: If they were down that supply chain somehow, it didn't matter.
22:33
Unidentified speaker: And I guess down big or well, down, you know, marginally to what I expected them to be up, right?
22:39
Unidentified speaker: And what they reported.
22:40
Unidentified speaker: And I don't think it was them missing a whisper number.
22:42
Unidentified speaker: I think it's shoot first, ask questions later.
22:45
Unidentified speaker: Same thing would happen to software, right?
22:46
Unidentified speaker: I mean, like they were just blanketed earlier in the year.
22:50
Unidentified speaker: They they were getting crushed because AI, you know, claw, you're gonna be able to replicate any software program out there.
22:56
Unidentified speaker: And the reality is you start recognizing, okay, that's not the case.
22:59
Unidentified speaker: You still need some quality control.
23:01
Unidentified speaker: You some of these software companies are entrenched.
23:03
Unidentified speaker: And I think eventually And human beings don't change that fast.
23:05
Unidentified speaker: And they don't change that fast.
23:07
Unidentified speaker: So I think eventually when you get back to it and you start looking, all right, when we start digesting and getting used to this dynamic here, we'll get back to what the fundamentals matter and people looking at gross margins and pricing and who can pass that.
23:18
Unidentified speaker: I guess it's just like what's happening with the election and the data center, the data center election and the blowback on AI that we're kind of seeing in the market today.
23:26
Unidentified speaker: It's like this is just the latest sort of blip in the AI trade.
23:29
Unidentified speaker: I feel like we get one every few weeks now where everyone's like, this is it.
23:33
Unidentified speaker: It's over.
23:33
Unidentified speaker: Right.
23:34
Unidentified speaker: The AI trade is done, and everyone's like, is it?
23:37
Unidentified speaker: And then it it picks back up.
23:39
Unidentified speaker: Doesn't it seem like that's what I'm watching too?
23:41
Unidentified speaker: Doesn't it seem like the bubble is people calling for the bubble?
23:45
Unidentified speaker: Yes, yes.
23:46
Unidentified speaker: Bubbles on bubbles.
23:47
Unidentified speaker: There's a bubble on people saying it's a bubble.
23:48
Unidentified speaker: I think that's it, but I agree with you.
23:50
Unidentified speaker: That's the beauty, that's the beauty of it because I say this quite a bit, like, but to be a good bull, you have to understand the bear.
23:57
Unidentified speaker: And the fact that the bears are popping up, that means you might be able to find incremental buyers of stocks.
24:02
Unidentified speaker: The super bulls can't buy the dip forever because eventually you just don't have any more capital to do it.
24:08
Unidentified speaker: You have to convince somebody new to come in and buy a stock.
24:11
Unidentified speaker: Yeah, but you can't buy the dip when you have no cash left out.
24:14
Unidentified speaker: That's what I'm saying.
24:15
Unidentified speaker: So the idea of people talking about bubbles and bears coming out and talking about negatives is a good thing for markets because you need people on the other side of the tree.
24:25
Unidentified speaker: I gotta say, I'm super confused.
24:27
Unidentified speaker: I can't tell if you're bullish or bearish.
24:29
Unidentified speaker: So it's funny, this is this is very interesting.
24:32
Unidentified speaker: This is why people always get confused around me.
24:36
Unidentified speaker: I am long-term bullish.
24:38
Unidentified speaker: Absolutely.
24:38
Unidentified speaker: I use this technology, I see the difference from what big companies where they actually are in an adoption cycle of technology.
24:45
Unidentified speaker: Even the Federal Reserve of St.
24:47
Unidentified speaker: Louis is now putting out like adoption studies, and you can see how early this is.
24:50
Unidentified speaker: Right.
24:51
Unidentified speaker: I see the use case, I see the adoption, I see the scale.
24:55
Unidentified speaker: I think it's really interesting long term.
24:57
Unidentified speaker: Short term, why you are dead right no one cares about earnings, the current quarter earnings, unless you miss.
25:06
Unidentified speaker: Unless you miss.
25:08
Unidentified speaker: Because it translates into how can I trust on you to deliver for the long term.
25:13
Unidentified speaker: And the bare thesis right now that's in the market is how do we finance this longer term?
25:19
Unidentified speaker: And that is very hard near term to prove wrong.
25:23
Unidentified speaker: Right.
25:24
Unidentified speaker: So tactically, I think we're in for a lot of chop.
25:28
Unidentified speaker: And I don't think like AMD and a lot of these names convince me short term that it's all clear.
25:34
Unidentified speaker: So I I think people will continue to sell first, ask questions later until we get the next big clearing event, anthropic.
25:42
Unidentified speaker: And this is you're much smarter than I am, but so probably not.
25:46
Unidentified speaker: 970 SAT score.
25:49
Unidentified speaker: Didn't write it, Canadian.
25:52
Unidentified speaker: But could you make any connection, let's say subprime is to circular financing?
25:59
Unidentified speaker: Like, is that where the crack's gonna be?
26:02
Unidentified speaker: Or I think the the crack is gonna be pricing and more growth than people can deliver.
26:06
Unidentified speaker: And I don't look at PE to value stocks.
26:10
Unidentified speaker: I don't, not even in indexes.
26:12
Unidentified speaker: I do reverse DCFs.
26:14
Unidentified speaker: I got I got beat up for 10 years by an absolutely great boss that used to always ask me the question, what's priced in?
26:23
Unidentified speaker: Right.
26:23
Unidentified speaker: And then he goes, I don't care about the whisper number or what someone told the trader.
26:26
Unidentified speaker: It's funny because he kind of used to be one.
26:29
Unidentified speaker: Um and he goes, What's priced in long term?
26:32
Unidentified speaker: Because that's where we live.
26:33
Unidentified speaker: Right.
26:33
Unidentified speaker: Try to figure that out.
26:34
Unidentified speaker: So we ended up building a whole bunch of models.
26:37
Unidentified speaker: So this is this is my my pitch to you.
26:39
Unidentified speaker: This entire AI trade, I don't think we're even remotely scratching the surface of what this means for for the world and and and and and every company involved.
26:51
Unidentified speaker: And so I don't think it's possible to price in the upside.
26:56
Unidentified speaker: I think we haven't overpriced it.
26:58
Unidentified speaker: I think you're totally right.
26:59
Unidentified speaker: I think um I've been thinking a lot about uh previous, I'm not the only one, thinking a lot about you know, technology, previous technological revolutions, right?
27:09
Unidentified speaker: Um, the computer revolution, you were talking about the dot-com boom earlier.
27:13
Unidentified speaker: These things are long-term, huge, massive changes.
27:17
Unidentified speaker: And like people like us sitting here talking about the day-to-day moves in the market are never gonna capture just the breadth of what is about to happen, right?
27:27
Unidentified speaker: There's a famous Nobel Prize-winning economist who said something about the computer revolution, like, I see it everywhere except in the productivity numbers.
27:34
Unidentified speaker: Like it just takes time to like really infiltrate with a new technology.
27:39
Unidentified speaker: Electrification took like a half a century to show up in the factories, computers took a long time to show up, you know, in companies and AI.
27:47
Unidentified speaker: Companies don't know what they're doing with AI, let's be honest.
27:50
Unidentified speaker: Like, we're all using it, but are we doing good things with it?
27:53
Unidentified speaker: Are we really transforming?
27:55
Unidentified speaker: I don't know.
27:57
Unidentified speaker: I would agree with you, and if and I wrote about this earlier in the year, if you go back in time with you know the 1960s with the supercomputer revolution, or go back during the industrial revolution and the 1800s, each period afterwards you saw an increase in productivity, wage growth, and GDP growth, right?
28:10
Unidentified speaker: All these inflection points.
28:11
Unidentified speaker: And now this is different in a sense because we're living through it.
28:15
Unidentified speaker: And I brought up the Navier-Stokes equation.
28:17
Unidentified speaker: You guys, have you heard about this?
28:18
Unidentified speaker: The mathematical equation, Navier-Stokes, that was just like a hundred-plus-year-old mathematical equation that was recently solved with the use of AI.
28:24
Unidentified speaker: Oh, yeah.
28:25
Unidentified speaker: And I think that's what's so amazing.
28:26
Unidentified speaker: It's like it's this black box of magic, right?
28:29
Unidentified speaker: You know, we we physically have adopted it, we're we're consuming it, and we have no idea what's doing in the background, and we have no idea what the answers are giving us are correct, right?
28:39
Unidentified speaker: Even though it's it's it's accurate to a certain extent, there's an illusion of certainty there.
28:43
Unidentified speaker: And I equate it to the show Cheers.
28:45
Unidentified speaker: You guys remember Cheers in the 80s?
28:47
Unidentified speaker: Of course.
28:47
Unidentified speaker: Remember Cliff Clavin?
28:48
Unidentified speaker: Yes.
28:49
Unidentified speaker: The resident bar guy, you go in there and you have knowledge and everything.
28:51
Unidentified speaker: Oh man, I'm too young.
28:53
Unidentified speaker: He was a mailman, yeah.
28:54
Unidentified speaker: And he you have no idea if he was full of it or if he was accurate.
28:57
Unidentified speaker: And I feel like AI, in a certain extent, the answers it gives me is like a Cliff Claven at times.
29:00
Unidentified speaker: Like at times, like, oh, it's it has it.
29:02
Unidentified speaker: And other times I'm like, this can't be right.
29:03
Unidentified speaker: Now I'm worried he's gonna call me Cliff for the rest of the hour.
29:06
Unidentified speaker: He was famously on Jeopardy for an episode.
29:08
Unidentified speaker: People should YouTube that.
29:09
Unidentified speaker: Cliff Claven on Jeopardy.
29:10
Unidentified speaker: Um, but to take a step back, right?
29:12
Unidentified speaker: Like AI and uh.com, there's a few differences, right?
29:17
Unidentified speaker: Dot com, they were laying fiber for future demand.
29:21
Unidentified speaker: Right now we are short AI compute.
29:23
Unidentified speaker: The companies financing this are the richest companies in human history.
29:27
Unidentified speaker: So again, if they say, all right, we don't need AI data centers, I'll go back to buying 50 billion a stock a quarter.
29:33
Unidentified speaker: So that's a little different.
29:35
Unidentified speaker: And then, like with the with the subprime, uh, you know, the like if you look at uh we'll get Core Weave in a week and Nebbius, these companies, uh, like the cash flow is coming to support that debt, i.e., like the ninja loans from 2008, the cash flows coming from Microsoft to support that debt are fundamentally different.
29:54
Unidentified speaker: It's just it history repeats, but rhyme.
29:57
Unidentified speaker: It's a different animal.
30:00
Unidentified speaker: Subclim is very fragmented, right?
30:01
Unidentified speaker: I mean, we're talking about the individual who was over-levered on their home is much different.
30:06
Unidentified speaker: It's tougher to control because you have so many loans out there.
30:08
Unidentified speaker: This is concentrating in one specific industry.
30:11
Unidentified speaker: And to your point, these are companies that are financing it, have tremendous amount of cash flow.
30:16
Unidentified speaker: You know, they can pull back CapEx, right, and they're and they're looking pretty good.
30:21
Unidentified speaker: The other concentration risk is people have no idea how long they are this trade, right?
30:29
Unidentified speaker: Through the passive vehicles, right?
30:30
Unidentified speaker: Right.
30:30
Unidentified speaker: Not even through the phone.
30:31
Unidentified speaker: Oh, well, I mean, just look at the small cap industrials today.
30:34
Unidentified speaker: I mean, it's everywhere.
30:35
Unidentified speaker: Yeah, everything's an AI stock.
30:37
Unidentified speaker: Well, I mean, a lot of the GDP growth in this past year has contributed to AI.
30:41
Unidentified speaker: I could imagine people waking up one day being like, wait, what?
30:44
Unidentified speaker: Like, I lost 40% of my unity.
30:46
Unidentified speaker: You know what I mean?
30:47
Unidentified speaker: Like, if there is some massive sell-off, they're not prepared to look at their accounts.
30:51
Unidentified speaker: This is what we're digging into a ton.
30:53
Unidentified speaker: And it's funny because Roundhill people associate us with memory and tech and AI.
31:00
Unidentified speaker: And you could see this in some of the new products we've brought to market.
31:03
Unidentified speaker: We're talking about heavy assets, low obsolescence.
31:06
Unidentified speaker: How do you stay in the AI trade through all the whipsaws and the roller coaster and not get seasick?
31:11
Unidentified speaker: You have to actually own a barbell.
31:13
Unidentified speaker: You need to own something that's different.
31:15
Unidentified speaker: And when I was on the show last time, I pushed back on somebody saying, I think utilities are a great thing to own.
31:20
Unidentified speaker: I'm like, you just own low beta AI.
31:22
Unidentified speaker: It's the same thing.
31:23
Unidentified speaker: What about healthcare though?
31:24
Unidentified speaker: We just had the midterms Democrats won.
31:27
Unidentified speaker: They've already, they've talked about, you know, get rid of getting rid of Medicaid cuts, I don't know what else they're gonna do.
31:33
Unidentified speaker: Those were rising sort of in anticipation of this election.
31:37
Unidentified speaker: Seems like that's that's not halo per se, but it's it's a different type of risk.
31:41
Unidentified speaker: And it's funny because everything we're getting around to is basically are we still in a structural bull market?
31:47
Unidentified speaker: And we think so, yes.
31:49
Unidentified speaker: We want risk, but we want to get different types of risk.
31:52
Unidentified speaker: That's just harder to find.
31:53
Unidentified speaker: Right.
31:54
Unidentified speaker: So we want true cyclical risk that has low, not negative correlation to AI risk.
32:00
Unidentified speaker: And then you can, and actually, great point, like healthcare kind of plays into this as well.
32:04
Unidentified speaker: It's like growth differentiated almost like political risk in a sense, but a different spending bucket.
32:09
Unidentified speaker: Right.
32:09
Unidentified speaker: So yesterday we um we had a debate on uh Apple.
32:13
Unidentified speaker: Do you where do you put that stock?
32:15
Unidentified speaker: It's somewhere in the middle.
32:17
Unidentified speaker: Uh it it I would say like it is in the AI value chain at a different point.
32:23
Unidentified speaker: Uh at some point you're going to see people are hiding out there thinking it's not.
32:27
Unidentified speaker: Uh they're they're in it in a different way.
32:29
Unidentified speaker: It's gonna be like that end user kind of consumer side of it.
32:32
Unidentified speaker: And like kind of meta's there too, a little bit as well.
32:35
Unidentified speaker: Right.
32:36
Unidentified speaker: And and honestly, this is why like the mag 7, when you break it down, they're probably the most differentiated they've been in the past 10, well, yeah, their entire history, which is basically the last time.
32:45
Unidentified speaker: Are we still saying Mag 7?
32:47
Unidentified speaker: Like that's still happening.
32:48
Unidentified speaker: I mean, you're saying it, so literally.
32:49
Unidentified speaker: Yeah, and the SP 493.
32:51
Unidentified speaker: We still say it.
32:51
Unidentified speaker: I think we say it.
32:52
Unidentified speaker: Okay.
32:53
Unidentified speaker: And never call it Mag 8, because that's not a movie.
32:55
Unidentified speaker: The Magnificent 7 is a wonderful movie.
32:57
Unidentified speaker: We should not disrespect.
32:58
Unidentified speaker: The remake or the original?
33:00
Unidentified speaker: Both.
33:00
Unidentified speaker: I don't mind I don't mind either, and Seven Samurai is very good.
33:03
Unidentified speaker: Oh, it's it is evolving.
33:05
Unidentified speaker: That's that's true.
33:06
Unidentified speaker: There's other like private companies.
33:08
Unidentified speaker: There's no better name yet.
33:09
Unidentified speaker: We're just waiting.
33:10
Unidentified speaker: Hey, wait for you to come up with one.
33:12
Unidentified speaker: Look, we'll just we'll let the market dictate that one and I'll listen.
33:15
Unidentified speaker: But I think, you know, uh whether or not we're in a bull bear market, I think you can have the main market go sideways for quite some time and other asset class start working now.
33:26
Unidentified speaker: Right?
33:26
Unidentified speaker: I mean, we've been in a structural period of time where the only place to be was the S P 500.
33:31
Unidentified speaker: And I think over the past year we're waking up, oh, wait, there are other asset classes we can start diversifying into that are not overvalued, that are not extended, and I can get a pretty good return.
33:40
Unidentified speaker: And i.e.
33:41
Unidentified speaker: that might be small cap, that might be smid, that might be international, whatever that might might come to fruition.
33:46
Unidentified speaker: But I think you can still have tough times in a certain a specific area of the market and do very well in other areas of the market.
33:52
Unidentified speaker: How do you how do you feel about because you just gave me like an acid flashback?
33:56
Unidentified speaker: I remember in the early 2000s, I was at um It's an asset flashback.
34:01
Unidentified speaker: Just to say that versus acid.
34:03
Unidentified speaker: You gotta make it.
34:06
Unidentified speaker: I thought somebody said actually.
34:08
Unidentified speaker: I was like, wait.
34:09
Unidentified speaker: No, no, no.
34:10
Unidentified speaker: I said acid.
34:11
Unidentified speaker: Oh, all right.
34:13
Unidentified speaker: Um but I remember like early 2000s, uh there was this like everyone waking up to you could invest outside of the US.
34:23
Unidentified speaker: You I when I graduated college, there were three things, three themes, narratives.
34:27
Unidentified speaker: There were the bricks.
34:28
Unidentified speaker: Bricks.
34:29
Unidentified speaker: Right?
34:29
Unidentified speaker: Remember the bricks?
34:30
Unidentified speaker: Also a good name.
34:30
Unidentified speaker: Right?
34:31
Unidentified speaker: Peak oil, so commodity super cycle.
34:33
Unidentified speaker: We're back.
34:34
Unidentified speaker: And then on top of that, uh you didn't want to own technology because it was obsolete.
34:37
Unidentified speaker: Now I graduated in 02 and it was right after the dot-com bubble.
34:40
Unidentified speaker: And and after um Y2K.
34:43
Unidentified speaker: And after Y2K.
34:45
Unidentified speaker: And then 25 years later, what was it?
34:47
Unidentified speaker: Commodities are dead, right?
34:48
Unidentified speaker: International was no place to be up until a year or two ago.
34:51
Unidentified speaker: And the only companies you want to own were technology companies.
34:54
Unidentified speaker: And it's amazing how these narratives shift, and it takes time.
34:57
Unidentified speaker: I think we're in that, I think we've been in this process of higher rates for longer, higher inflation for longer than people anticipated, this massive amount of capital that was infused in the market coming out of COVID that that are in the concentrated within these Mag 7 brilliant companies, now is getting diffused out to other parts of the market.
35:14
Unidentified speaker: And I think we look back 10 years from now, we're gonna say, why was I a little bit more diversified?
35:18
Unidentified speaker: Why didn't I have a little different type of portfolio than instead of this massive concentration?
35:22
Unidentified speaker: Yeah, I mean, I I've this is my 13th show I've hosted, and I think this is the first time we've even mentioned something outside of the US.
35:31
Unidentified speaker: Like it's it's not really on on people's radar.
35:34
Unidentified speaker: It should be.
35:35
Unidentified speaker: It absolutely should.
35:36
Unidentified speaker: It used to be that way, right?
35:38
Unidentified speaker: Yeah, but a bunch of that's correlated too.
35:40
Unidentified speaker: If you bought, if you bought EM, you're getting a lot of the AI trade too.
35:45
Unidentified speaker: Yeah.
35:45
Unidentified speaker: Well, you can't, I mean AI different places.
35:48
Unidentified speaker: But again, AI is a different animal.
35:49
Unidentified speaker: I think my point is that once that narrative changes, it's it's it's no longer about fundamentals, it's about the story people are telling each other.
35:55
Unidentified speaker: I think once that story gets shifted, then other asset classes can start participating as well.
36:02
Unidentified speaker: And I think we're in the process of that occurring right now.
36:04
Unidentified speaker: Right.
36:04
Unidentified speaker: I think I'd say we're already there.
36:07
Unidentified speaker: And it's funny, what's happening to me is people are finally recognizing that it's not a stock market, that it's a market of stocks.
36:14
Unidentified speaker: Kind of flip that around back to things that have been beaten in my head for 10 years.
36:17
Unidentified speaker: Um the trading tools have evolved to a point, and I would say behavior at even big institutions has trade changed a lot where markets are now very micro efficient.
36:30
Unidentified speaker: Everyone feels like like when you talk to a macro person, a true macro person, they go, the SP doesn't make sense.
36:37
Unidentified speaker: It needs to go down.
36:38
Unidentified speaker: Why are stocks still high?
36:39
Unidentified speaker: Like, why is it?
36:40
Unidentified speaker: And like, well, that's because when that risk happens, like oil pressure happens, inflation pressure happens, everyone traded the consumer stocks because there's a million ways to do that more effectively rather than just trade index.
36:53
Unidentified speaker: And like what I was seeing before I left my previous role, hedge funds weren't looking for 100, 200 stock baskets.
37:02
Unidentified speaker: They wanted 20, 30 stock baskets that had the most leverage to a trade idea.
37:06
Unidentified speaker: Right.
37:07
Unidentified speaker: Not leverage as in borrowing money, but like the most bang for their buck to make a trade.
37:11
Unidentified speaker: And they're like, great, now I want 40 other ones that are different.
37:15
Unidentified speaker: So like I'd say, like the trend of trading risk is become more focused, and that's why correlation's really low.
37:23
Unidentified speaker: And now we're starting to realize that there's a lot more assets out there to help me actually diversify.
37:28
Unidentified speaker: And I think people are trading the pieces within them.
37:30
Unidentified speaker: Right.
37:31
Unidentified speaker: And I you you brought up an interesting fact about the about the gap.
37:34
Unidentified speaker: The other gap that I think is significant is institutional investors and retail investors.
37:40
Unidentified speaker: That, like, in my opinion, has become very, very close.
37:45
Unidentified speaker: You don't agree, or not?
37:47
Unidentified speaker: I completely agree.
37:48
Unidentified speaker: Right.
37:48
Unidentified speaker: Completely agree.
37:49
Unidentified speaker: You know, it you we used to we used to call it like dumb money, right?
37:52
Unidentified speaker: Or like, you know, we would you as as a hedge fund trader, we'd kind of make fun of the retail investor, and it's well, social coordination has enabled that.
38:01
Unidentified speaker: And totally and commissionless free trading has also reduced any speed bumps and people putting it in.
38:07
Unidentified speaker: We were talking about Reg FT too.
38:09
Unidentified speaker: Well, Reg FT, but but you know, go back to Robinhood and the impact that it had just on from a behavioral standpoint.
38:15
Unidentified speaker: Before you had to pay, you know, in the 90s it was like, you know, $18, $25 to get in and out of a stock.
38:20
Unidentified speaker: And then it went down to nine, and now it's zero, and that prevents you.
38:24
Unidentified speaker: It's called loss aversion.
38:25
Unidentified speaker: It's a it's a behavioral phenomenon.
38:27
Unidentified speaker: It was proven out that it prevents you from putting on the brakes when you go into trading.
38:31
Unidentified speaker: Now you just kind of go in and trade whatever you want because there's not that negative feeling of paying $8 out of your pocket.
38:37
Unidentified speaker: Right.
38:37
Unidentified speaker: Combine that with the ability to kind of use your iPhone to have these discussions on Reddit, they become a force to influence the market.
38:44
Unidentified speaker: Now, whether that's right or wrong, we'll find that out over time.
38:47
Unidentified speaker: And you know, as an active investor, I should believe that the markets are inefficient, and I don't.
38:51
Unidentified speaker: I think they're very efficient at pricing in the biases of the underlying individuals.
38:55
Unidentified speaker: And as a result of that, because that efficiency, you're gonna have periods of time where things can be greatly overextended for whatever the reason might be.
39:02
Unidentified speaker: Right.
39:03
Unidentified speaker: And, you know, under realizing that it's humans behind it making these decisions that have these biases, it helps you better understand we have periods that are stretched or periods that are fearful and everyone's selling and capitulating.
39:15
Unidentified speaker: Right.
39:16
Unidentified speaker: And I'm so impressed.
39:17
Unidentified speaker: You mentioned two of my favorite names in one thought Reddit and Robinhood.
39:22
Unidentified speaker: I love them both.
39:23
Unidentified speaker: Yeah.
39:23
Unidentified speaker: Um, so CrowdStrike announced that they are billing uh built a kill switch for AI agents that go rogue.
39:32
Unidentified speaker: They're buying a couple startups, about a half billion dollars per the June uh the journal.
39:37
Unidentified speaker: Uh my first thought was shouldn't have this been done on day one?
39:42
Unidentified speaker: Yes.
39:43
Unidentified speaker: And so yeah, so yeah, tell me.
39:45
Unidentified speaker: Like, I mean, uh there's so much talk, you know, initially about AI getting rid of jobs, you know, j killing companies, whatever.
39:52
Unidentified speaker: But CrowdStrike, I think their stock is up like three digits for the past, you know, for the past year.
39:58
Unidentified speaker: Cybersecurity is very hot right now.
40:01
Unidentified speaker: These the guys who work at these companies and gals are, you know, they're up all day and night fighting, fending off, you know, rogue AI attacks.
40:10
Unidentified speaker: I mean, yes, they should have done this before, they're doing it now.
40:15
Unidentified speaker: These companies are seeing an immense amount of business.
40:19
Unidentified speaker: It's almost like anthropic, like um they basically they created this technology and then they created the need to fight the technology, and it's just like this war uh, you know, between the two.
40:34
Unidentified speaker: It's kind of like mafia.
40:35
Unidentified speaker: Let me ask you this does does the kill switch make us safer or braver?
40:41
Unidentified speaker: Interesting.
40:42
Unidentified speaker: I'm I'm gonna say safer.
40:44
Unidentified speaker: We we need to be able to stop whatever's coming down the pike, right?
40:47
Unidentified speaker: I mean, we had the attack a couple of weeks ago, and everyone was like, this is the end of the internet.
40:51
Unidentified speaker: And then I woke up the next day, they were like, it's kind of it was more like Y2K and everything's fine, actually.
40:56
Unidentified speaker: Do you do you guys remember that?
40:57
Unidentified speaker: Oh yeah.
40:58
Unidentified speaker: Yeah, absolutely.
40:59
Unidentified speaker: I I I I mean, I'm all for having some way to you know make sure these things that get out of hand.
41:05
Unidentified speaker: I'm more concerned about when these AI is actually in and robots that are moving around.
41:10
Unidentified speaker: I mean, let's hopefully there's a kill switch for that, right?
41:13
Unidentified speaker: When you think about it.
41:14
Unidentified speaker: Right.
41:14
Unidentified speaker: Because as a software can take down a grid, absolutely, it's terrible, and they should be thinking about it, but we should be talking and having these discussions because whether it makes us braver, safer, whatever it is, it's imperative to kind of get our arms around.
41:26
Unidentified speaker: Again, what I view is this black box.
41:29
Unidentified speaker: I mean, we really don't understand what is going on when it comes to AGI.
41:35
Unidentified speaker: We think we do, but we don't.
41:37
Unidentified speaker: Right.
41:37
Unidentified speaker: I mean, isn't that the plot of every like dark sci-fi, you know, movie?
41:43
Unidentified speaker: There's all there's the always the utopian version where, like, you know, we got Star Wars where the droids are helpful, and then we have Terminator where they're not helpful.
41:52
Unidentified speaker: And we just becomes helpful.
41:55
Unidentified speaker: He becomes helpful.
41:56
Unidentified speaker: Like we can see a positive.
41:57
Unidentified speaker: We just have to wait for the sequel.
41:59
Unidentified speaker: Exactly.
42:00
Unidentified speaker: The sequel to humanity, hopefully not.
42:02
Unidentified speaker: Um so you brought up healthcare before, uh, which I think is is worthy of uh discussion.
42:08
Unidentified speaker: You know, people have come out and said that AI is gonna, you know, cure all these diseases and and everything, but drug makers are saying, hold on, you know, not quite yet.
42:18
Unidentified speaker: We haven't seen enough data.
42:21
Unidentified speaker: Um I I I don't know.
42:22
Unidentified speaker: I I find it, I think the industry is ripe for transformation.
42:27
Unidentified speaker: And I could I think it could be the most transformative uh industry.
42:32
Unidentified speaker: Uh and and the other thing that we we should talk about is you mentioned AI is taking jobs, right?
42:38
Unidentified speaker: Well, what if we start living 10 years, 20 years longer?
42:43
Unidentified speaker: Like what is what is that gonna do to our society?
42:46
Unidentified speaker: Yeah, the demographics are actually in a really good place for AI to take a few jobs.
42:50
Unidentified speaker: I'm not saying I want AI to take jobs, and then they might it might take the wrong jobs, white-collar jobs, and we need more healthcare workers to take care of all the older people that have been that are living longer because AI somehow something something, healthcare, blah, blah, blah, everything worked out great.
43:06
Unidentified speaker: Um, but we could use some AI to help with healthcare, with elder care, as the demographics of around the world and in the US really, really change.
43:18
Unidentified speaker: So it could turn out to be the right technology at the right time.
43:21
Unidentified speaker: Right.
43:21
Unidentified speaker: I think it's great for biotech development.
43:25
Unidentified speaker: It might not eventually be great for all the biotech stocks.
43:30
Unidentified speaker: It's funny in a sense where it levels the playing field.
43:33
Unidentified speaker: Because if you're very big and you have more researchers, you can develop more drugs.
43:38
Unidentified speaker: Right.
43:39
Unidentified speaker: Um, smaller companies can now do this more efficiently.
43:42
Unidentified speaker: The problem is, and it's kind of like the running joke with biotech stocks, they're absolutely great stocks until they have to become real companies.
43:49
Unidentified speaker: So they're very great development companies, they're not really good selling companies.
43:53
Unidentified speaker: So AI helps with one thing, developing drugs.
43:56
Unidentified speaker: Is it going to have a massive impact on the world if no one could sell them, distribute them, figure out making them the physical side of the state?
44:05
Unidentified speaker: Would your expertise in small caps be able to like is is that with the biotech space?
44:10
Unidentified speaker: Yeah.
44:11
Unidentified speaker: Like biotech is a tough space to invest in just because it's a very binary event.
44:16
Unidentified speaker: Uh it's not based on any fundamentals.
44:18
Unidentified speaker: It's kind of figuring out whether they go through trials and get FDA approval.
44:21
Unidentified speaker: And the hit rates, you know, the success rate are pretty low in that space.
44:24
Unidentified speaker: But I will say success.
44:26
Unidentified speaker: It what it was probability based or weighted DCFs of how is how everyone valued a random that goes back to your statistics.
44:34
Unidentified speaker: It doesn't matter if you had a 51%, it only matters if the drug worked or if it didn't work.
44:39
Unidentified speaker: You were either up a hundred or down to zero.
44:42
Unidentified speaker: You know, the when you went and pitched to your boss, it was a 75% chance of success.
44:46
Unidentified speaker: He doesn't like the hearts.
44:47
Unidentified speaker: The probability of success is, yeah, extremely.
44:49
Unidentified speaker: But I will say, getting back to healthcare and the benefits of AI, like absolutely, making research more efficient, smaller players, being able to develop drugs.
44:57
Unidentified speaker: Um, but there's supposed to be quality control over it, especially when it comes to healthcare.
45:01
Unidentified speaker: And I say this because if you remember uh Elizabeth Holmes with Theranos of their board making machine?
45:05
Unidentified speaker: We remember.
45:05
Unidentified speaker: Yeah, we all remember, right?
45:07
Unidentified speaker: I mean, you registrated a lot of you caught it demonstrating.
45:12
Unidentified speaker: But look at my eyes, what I lied.
45:14
Unidentified speaker: But no, but it it would the crazy thing that's I have no reason to deceive you.
45:17
Unidentified speaker: It's it's wild.
45:18
Unidentified speaker: But you know, the problem with that one was she was trying to do a fake it till you make it, which was successful for many in uh innovators of the time.
45:25
Unidentified speaker: But the problem was it was healthcare.
45:26
Unidentified speaker: You can't fake it till you make it in healthcare.
45:28
Unidentified speaker: Yeah, you can't deliver wrong blood results to people that you know live and die by that.
45:31
Unidentified speaker: And that's the one thing with AI when it comes down to like, yes, it's an amazing tool, it's gonna help, it's gonna uh uh advance us amazingly over the years to come.
45:40
Unidentified speaker: But we have to have quality control, especially in healthcare, because if I'm relying on it to give me answers to my health and it sends me down a wrong rabbit hole, we're gonna have massive issues.
45:50
Unidentified speaker: Well, that's I mean, that's uh we could we could bring out the flags, okay?
45:54
Unidentified speaker: Oh, flags.
45:54
Unidentified speaker: Would you my flag's down?
45:56
Unidentified speaker: Would you rely on a chat bot for a let's say serious health issue?
46:06
Unidentified speaker: Green flag or red flag?
46:08
Unidentified speaker: Oh, this is this is gonna be weird.
46:10
Unidentified speaker: I'm gonna put both on the phone.
46:11
Unidentified speaker: I'm I actually, you know what, that might be the answer.
46:13
Unidentified speaker: I'm both.
46:14
Unidentified speaker: All right.
46:14
Unidentified speaker: I think I think Howard.
46:16
Unidentified speaker: I no, no, no.
46:17
Unidentified speaker: Just so you know, yellow flags are against the rules and you have been banned from marketing.
46:23
Unidentified speaker: Uh no, I'll tell you why.
46:25
Unidentified speaker: Uh no, I'll tell you why.
46:26
Unidentified speaker: I think you I think you you must absolutely use it to get potentially second opinions, but I don't think you can rely on it wholeheartedly.
46:32
Unidentified speaker: That's fair.
46:32
Unidentified speaker: I honestly I don't think it's a binary event.
46:36
Unidentified speaker: It it's so gray because to say no, don't use it at all, you you're putting yourself at a disadvantage.
46:41
Unidentified speaker: Because like I now I put it, any type of blood results I get, I throw it up there to find out information because it it answers me immediately and it gives me some kernel's information that my doctor's not going to find out.
46:50
Unidentified speaker: But to rely on solely to make decisions of of whether or not someone's gonna surgery, right?
46:55
Unidentified speaker: Or should I do this, like a biopsy for this?
46:58
Unidentified speaker: I I think it's a tough call at this juncture.
47:00
Unidentified speaker: I love it for the the prep for how do I actually understand what you're trying to tell me as a doctor?
47:06
Unidentified speaker: Because I have no discernible skills in that world.
47:09
Unidentified speaker: So it's like I gotta go in for XYZ.
47:13
Unidentified speaker: What like can I have like a quick little tearsheet to learn a little bit about it so I can ask my doctor the right question?
47:19
Unidentified speaker: Super helpful.
47:20
Unidentified speaker: Also, it saves me a couple calls to my mom and mother-in-law who are fairly.
47:24
Unidentified speaker: Well, you know what else is great?
47:25
Unidentified speaker: Have you uh does anyone own a plot?
47:27
Unidentified speaker: It's a recording device for the back of your phone.
47:30
Unidentified speaker: Okay, so I I have a recording device, not it's not on now.
47:34
Unidentified speaker: But um I can go into a TV.
47:41
Unidentified speaker: When I come up with my friends, I'm like, don't worry, it's not on.
47:44
Unidentified speaker: But you can go into a doctor's a little a little higher, a little higher.
47:51
Unidentified speaker: Um but you can go to a doctor's appointment and literally record the whole thing, and then you have the transcript, you can drop drop it into AI, and you can have all of your answers.
48:02
Unidentified speaker: Let me pose a question to you.
48:03
Unidentified speaker: Yeah.
48:03
Unidentified speaker: Would you allow, do you think therapy advice from AI is a red flag or green flag?
48:09
Unidentified speaker: I'll pose it to everybody.
48:10
Unidentified speaker: Yeah, so I mean, I I I have 25 years of therapy.
48:14
Unidentified speaker: I'm so over 17 years.
48:16
Unidentified speaker: Uh I think I probably would answer the same way you did with both flags.
48:21
Unidentified speaker: Like, I think it's a resource, but not, no, I'm not relying on it solely.
48:26
Unidentified speaker: And it's a tool.
48:27
Unidentified speaker: It's a tool.
48:28
Unidentified speaker: It's not it's not the handyman that comes to your house.
48:31
Unidentified speaker: It's like maybe you can like tighten a screw here, tighten a screw there, but then when you have a serious plumbing issue, you gotta call in an expert.
48:38
Unidentified speaker: Otherwise, stuff is gonna hit the fan.
48:40
Unidentified speaker: You know what I mean?
48:41
Unidentified speaker: Yeah, but on the wrong side of that.
48:44
Unidentified speaker: Well, and what I mean, correct me if was weren't there some headlines of some suicides that were assisted by chatbots, right?
48:53
Unidentified speaker: So not you know, scary stuff.
48:56
Unidentified speaker: Look, the AI in general, to me, good people will do good things without any tool you give them.
49:02
Unidentified speaker: Bad people will do bad things or find or someone in a bad place will find bad things to do with them.
49:09
Unidentified speaker: Like that's you you see that all over with any type of thing.
49:12
Unidentified speaker: There was also an article.
49:13
Unidentified speaker: Um, I guess lawyers are getting getting kind of annoyed that um, you know, people who are going through divorces couples therapy, or yeah, yeah, yeah.
49:22
Unidentified speaker: They're people are coming in guns blazing, like ChatGPT told me this, you know.
49:26
Unidentified speaker: So like I I I don't know.
49:29
Unidentified speaker: I was laughing reading that.
49:30
Unidentified speaker: I I did a story similar to that maybe last year, talking to some uh plaintiff's lawyers who said people were coming in and were like, I have a great case.
49:38
Unidentified speaker: And they would have like really long explanations of the case, and the lawyers would be like, You're no, like this isn't right.
49:45
Unidentified speaker: You you were driving drunk.
49:46
Unidentified speaker: Like you can't, it's not gonna work.
49:49
Unidentified speaker: And they're like, but look what you know, ChatGPT says there's this law, you know.
49:52
Unidentified speaker: Right.
49:53
Unidentified speaker: Um, it benefits the lawyers though.
49:56
Unidentified speaker: In the Wall Street Journal piece you're talking about, about divorce.
49:59
Unidentified speaker: Um, there are some lawyers who are like, it's gonna cost the client more because I'm gonna need to explain to them point by point why their, you know, Chat GPT analysis of their case is incorrect.
50:10
Unidentified speaker: Like they're gonna wind up.
50:12
Unidentified speaker: And let's not forget this device, it also adapts to make you stick around by complimenting you.
50:17
Unidentified speaker: Right?
50:18
Unidentified speaker: It's it it's you know, like over time, it's like you know, it's trying to make sure you come back and reinforcing whatever narrative you might have.
50:25
Unidentified speaker: It's not it's not gonna be somebody that's gonna put a hard line in the sand and say, no, you're wrong.
50:30
Unidentified speaker: It's gonna be very gentle, and there's a reason behind it because it wants its incentive is for you to stick around and use it as long as possible.
50:36
Unidentified speaker: So you have to be mindful.
50:37
Unidentified speaker: That can't be good for therapy, right?
50:38
Unidentified speaker: Reinforcing false narratives is what it comes down to.
50:40
Unidentified speaker: And what happens is you just have individuals that are bringing their own biases into these, into these kind of conversations, and instead of dispelling them, which what I think you know potentially sh could happen, having normal conversations with real people, you're just perpetuating this romanticized idea that you have in your head, and and that can't be a good thing either.
50:59
Unidentified speaker: You have to be mindful, again, back to your point, and what I wrote in my book as well.
51:03
Unidentified speaker: This is a tool, it's an amazing tool, but it's just a tool.
51:06
Unidentified speaker: It can never replace the the human aspect.
51:09
Unidentified speaker: Do you do you use an AI agent?
51:12
Unidentified speaker: I I use it for yes, I use it, I use AI daily for news assessment, for um going through uh transcripts very effectively.
51:21
Unidentified speaker: Like I'll take I have a program that my partner does for Google Trends, like for all my consumer companies.
51:27
Unidentified speaker: We'll download um all the terms and over historically the past five years, like how they've been trending on Google uh search, and I'll upload it a claude and it clusters them into kind of what the narratives are and it gives me an idea of what's going on.
51:38
Unidentified speaker: And you know, sometimes it's not really spot on, but other times it makes sense and it and it gives me it gives me a good understanding to see what pe other people are thinking.
51:47
Unidentified speaker: Right.
51:47
Unidentified speaker: I'm loving this because it's basically pulling out it's a tool, and it you have to have different skills to use it and know how to use it.
51:57
Unidentified speaker: You use a hammer and a screwdriver differently, right?
52:00
Unidentified speaker: If you prompt well, you get a better answer.
52:04
Unidentified speaker: You are not necessarily a maker of data, used to be an Excel jockey, thought I was good, turns out Claude's a hell of a lot faster than I am, right?
52:12
Unidentified speaker: To a checker, which is a different skill.
52:15
Unidentified speaker: Right.
52:15
Unidentified speaker: What have you output to me?
52:17
Unidentified speaker: Is it right or wrong?
52:18
Unidentified speaker: Different skill sets.
52:19
Unidentified speaker: Right.
52:19
Unidentified speaker: So I think like AI use, like you said it right at the beginning, like we're just scratching the surface on this stuff, right?
52:24
Unidentified speaker: Yeah.
52:24
Unidentified speaker: My favorite is putting like I think I said that.
52:28
Unidentified speaker: I gave credit that way.
52:28
Unidentified speaker: You ever play AIs off each other?
52:30
Unidentified speaker: You're like, well, like I do, I'll go to OpenIC.
52:32
Unidentified speaker: I try not to anger them.
52:33
Unidentified speaker: I take it as like, well, chat told me this, and Claude, and Claude comes out with a different thing.
52:36
Unidentified speaker: It's like you're just having conversation between two chatbots, and sometimes you get very divergent answers depending upon what you're talking about.
52:43
Unidentified speaker: Al what's your most use for AI?
52:45
Unidentified speaker: Uh, right stories.
52:47
Unidentified speaker: That is a joke.
52:48
Unidentified speaker: That is a joke.
52:50
Unidentified speaker: I know, Alan.
52:50
Unidentified speaker: Uh no, it just it makes it um so there's a gazillion uses.
52:55
Unidentified speaker: It makes going through transcripts easy, it makes going through documents from the FCC.
52:59
Unidentified speaker: I would I said this the other day to somebody.
53:02
Unidentified speaker: Um, you know, SpaceX and its drive to become a mobile wireless company and spectrum purchases and FCC hearings and uh purchase uh agreements and what's been disclosed publicly, it has made uh a weeks-long project an hour.
53:17
Unidentified speaker: Right.
53:17
Unidentified speaker: Oh yeah.
53:18
Unidentified speaker: So that was very helpful.
53:18
Unidentified speaker: But were you talking about actual agents?
53:20
Unidentified speaker: Like I was I feel like we're going from you know chatbots to agents to eventually sort of embodied AI.
53:27
Unidentified speaker: And so That's already happening.
53:28
Unidentified speaker: You know, it's it's all sort of marching forward.
53:30
Unidentified speaker: Like the the chatbots were terrible, now they're better.
53:33
Unidentified speaker: You know, the images used to have six fingers.
53:35
Unidentified speaker: Right.
53:36
Unidentified speaker: Now it takes my 16-year-old to tell me no, that's an AI image, and I still don't understand.
53:40
Unidentified speaker: My my favorite use.
53:41
Unidentified speaker: The agents are sort of okay, but they'll get better.
53:44
Unidentified speaker: My favorite use is I can say, uh, I'm going to this Broadway show.
53:49
Unidentified speaker: Here's the address.
53:50
Unidentified speaker: I'm with someone who's a vegan, I'm with a vegetarian, and I want Italian.
53:54
Unidentified speaker: I don't want to walk more than eight minutes to the theater.
53:58
Unidentified speaker: Go.
53:59
Unidentified speaker: And it gives me like three options.
54:01
Unidentified speaker: It's like, that's a home run.
54:03
Unidentified speaker: I would have thought, and that that cost as much as the math problem to figure out to get the vegan Italian and but I don't know about you guys, but I find myself having more work to do.
54:12
Unidentified speaker: Because there are like projects that I've been wanting to do that would take massive amount of man hours, like weeks upon weeks.
54:18
Unidentified speaker: And now I'm able to condense them down to a day.
54:20
Unidentified speaker: And there's just so many more creative projects I want to do, and I find myself going down rabbit holes and more work and more work.
54:26
Unidentified speaker: So to close, I'd love to hear who's your favorite uh AI use, chat bots, whether it's Claude, ChatGPT, Perplexity, Muse.
54:37
Unidentified speaker: Do you have a favorite?
54:38
Unidentified speaker: I'm a Claude guy.
54:39
Unidentified speaker: Uh Claude.
54:40
Unidentified speaker: Love Claude.
54:40
Unidentified speaker: Yeah.
54:40
Unidentified speaker: Yeah.
54:41
Unidentified speaker: Love it.
54:41
Unidentified speaker: Uh we are allowed to use Gemini, and that is the favorite and one I use.
54:45
Unidentified speaker: Okay.
54:46
Unidentified speaker: Well, thank you all for being here.
54:47
Unidentified speaker: It was an awesome show.
54:49
Unidentified speaker: You guys are phenomenal.
54:50
Unidentified speaker: Hopefully, you all come back.
54:51
Unidentified speaker: We'll do.
54:52
Unidentified speaker: Thank you.
54:53
Unidentified speaker: Thanks, Turney.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3 and interpreted by Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from Market Hang

Market Hang

Next Chapter For NVIDIA Leads Back To Jensen?

Episode discussion: NVIDIA Leadership Transition, Anthropic IPO, and AI Investment Themes

Speakers’ claims: Speakers analyze the market reaction to Jensen Huang's resignation from NVIDIA, debating whether the company needs a visionary or an operator, while noting Polymarket odds of his return at 42%. The panel discusses the Anthropic IPO as a stabilizing market event, citing six-times oversubscription and a 30% valuation discount. They argue that AI adoption benefits smaller companies through margin expansion rather than immediate revenue growth, with enterprise adoption lagging expectations. Intel's struggles and government involvement are contrasted with NVIDIA's transition, raising concerns about execution risks. The episode concludes with warnings about overconcentration in AI assets and the commoditization of memory chips.

NVIDIAJensen HuangAnthropic IPO
54 min1 assets20 claims to verify
Market Hang

Why Nvidia’s leadership shift could reshape its competitive edge

Episode discussion: Panel debates Nvidia leadership transition, Anthropic IPO, and AI agent interoperability risks

Speakers’ claims: Speakers analyze Nvidia's governance risks following Jensen Huang's reported resignation, comparing the situation to Apple's transition while highlighting key man risk and board expertise concerns. The discussion covers Anthropic's IPO valuation relative to OpenAI, noting claims of significant revenue differences and competitive dynamics with Chinese models. Panelists evaluate hyperscaler valuations, with some arguing Amazon and Google offer reasonable growth potential driven by installed compute, while others question the sustainability of AI spending in a rising rate environment. Additional topics include SpaceX-Tesla merger speculation, Federal Reserve policy impacts, and barriers to AI agent adoption such as interoperability issues between platforms.

NvidiaAnthropicAI Agents
55 min6 assets24 claims to verify
All insights