NVIDIA succession speculation and AI wrapper moat dynamics discussed on Lumida Wealth
Summary
Speaker_0 analyzes Meta's late entry into prediction markets and the vulnerability of AI wrappers like Viktor to platform integration. The discussion highlights how major AI providers avoid regulated sectors due to licensing, creating moats for incumbents. A significant portion covers NVIDIA's leadership turmoil, arguing the CFO was a poor CEO choice compared to operational leaders like Ian or ARM's CEO. The speaker contrasts traditional reporting CFOs with hands-on revenue leaders at companies like Anthropic.
Key takeaways
- 1 Meta launched a prediction market app as a fast-follow strategy, leveraging its distribution network to copy successful apps like Polymarket.
Unidentified speaker (speaker_0) · 0:00
- 2 claw's Claude's not gonna going go to after regular. regulated categories. They can't. They need a license, and technology companies avoid regulated categories. They will do enablement for regulated categories.
Unidentified speaker (speaker_0) · 3:13
Episode sections
- 0:00 – 1:03
Meta's prediction market app
Speaker_0 describes Meta launching a prediction market app to compete with Polymarket and Kalshi, characterizing it as a late, fast-follow move similar to Libra and the Metaverse. The speaker notes Meta's advantage lies in direct distribution and reverse-engineering successful apps on its platform.
- 1:03 – 2:24
Claude takes on Viktor in Slack
The speaker discusses Anthropic's Claude integrating with Slack, displacing AI wrappers like Viktor which launched in February with $65-70 million in YTD revenue. Viktor was the first to embed AI into Slack and has Stuart Butterfield as an investor.
- 2:24 – 3:32
AI wrappers and the moat question
Speaker_0 argues that AI wrappers like Viktor and Higgsfield lack a real moat, relying instead on phenomenal go-to-market and user-generated content marketing. He suggests companies in regulated spaces are better protected because major AI providers like Claude avoid direct entry into regulated categories due to licensing requirements.
- 3:32 – 4:15
The NVIDIA drama
The speaker expresses disbelief at NVIDIA's leadership turmoil, dismissing rumors that the CFO was a spy or that infidelity was the sole cause for termination. He argues the board's decision to install the CFO instead of the head of the data center business was a poor strategic choice.
- 4:15 – 5:26
Why the CFO was the wrong pick for CEO
Speaker_0 contends the logical successor would have been Ian, the head of the data center business, citing precedents at Microsoft and Amazon where CEOs came from the largest, fastest-growing profit centers. He contrasts this with the CFO's role in financial forecasting rather than revenue growth and customer interaction.
- 5:26 – 7:13
Anthropic's hands-on CFO
The speaker highlights Anthropic's CFO as a model of a hands-on, revenue-growth leader who participates in daily pipeline stand-ups, unlike traditional reporting CFOs. He contrasts this with NVIDIA's Colette Kraemer and Google's Ruth Porat, noting that strategic CFOs focus on capital allocation rather than pulling the trigger on investment decisions.
- 7:13 – 8:17
Could Arm's CEO run NVIDIA?
Speaker_0 suggests ARM's CEO would be a sensible successor to NVIDIA due to similar IP licensing business models and overlapping supplier/customer bases. He notes that while the fit makes sense, the board's lack of operator experience may hinder effective succession planning.
- 8:17 – 23:13
Stories about Jensen and succession
The speaker discusses stories about Jensen Huang's micromanagement and employee support, arguing that his hands-on style built the company. He posits that the closest successor to a CEO is the general manager of the fastest-growing business, not a CFO, and questions why Ian was passed over.
- 23:13 – 23:44
NVIDIA Leadership and Policy Pressure
The speaker suggests that the turmoil surrounding NVIDIA's CEO Jensen Huang may have inadvertently reduced political pressure on Democrats regarding data center policy. This shift allows the conversation to move away from immediate policy concerns toward speculation about future data center developments.
Assets discussed
| Asset | Kind | Stance | Context |
|---|---|---|---|
| Meta PlatformsMETA | equity | neutral | Meta's rolling out an app for prediction markets. They want to compete with Polymarket and Calci. I think this is more about late to the party and trying to play fast catch up. |
| AmazonAMZN | equity | neutral | It's it's not unlike Amazon looking at who's growing |
| NVIDIANVDA | equity | neutral | what else? On the Nvidia drama, |
| Invesco QQQ TrustQQQ | etf | neutral | QQQ is up like what six points in Sunday? |
“Stance” is how the speakers talked about the asset, not a recommendation.
Claims worth verifying
Meta is rolling out an app for prediction markets to compete with Polymarket and Kalshi.
Unidentified speaker (speaker_0) ·0:00
How to verify: Check Meta's official press releases or app store listings for a prediction market application released around November 2026. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Mark Zuckerberg was late to the party with Libra and the Metaverse.
Unidentified speaker (speaker_0) ·0:17
How to verify: Review historical timelines of Meta's Libra/Diem project launch and Metaverse announcements relative to competitors. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Polymarket and Kalshi avail themselves of marketing on Meta's app.
Unidentified speaker (speaker_0) ·0:34
How to verify: Analyze Meta's advertising revenue reports or public partnerships with Polymarket and Kalshi. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Viktor launched in February of this year with year-to-date revenue of about $65 to $70 million.
Unidentified speaker (UNKNOWN) ·1:18
How to verify: Check Viktor's public revenue disclosures, press releases, or funding announcements for the relevant year. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Stuart Butterfield, founder of Slack, is an investor in Viktor.
Unidentified speaker (speaker_0) ·1:29
How to verify: Review SEC Form D filings or press releases listing Viktor's investors. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Primary sourceAnthropic's CFO participates in daily 8 AM Pacific time stand-ups reviewing pipeline by category.
Unidentified speaker (UNKNOWN) ·5:36
How to verify: Verify through Anthropic's public case studies, interviews with the CFO, or employee testimonials. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
NVIDIA's CFO is Colette Kraemer.
Unidentified speaker (speaker_0) ·6:25
How to verify: Check NVIDIA's investor relations page or recent SEC filings for executive titles. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Primary sourceNVIDIA's ROIC is somewhere between 15% to 30%.
Unidentified speaker (speaker_0) ·14:25
How to verify: Calculate Return on Invested Capital using NVIDIA's income statement and balance sheet data from SEC filings. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Primary sourceBroadcom was forced to extend less leverage because of how aggressive NVIDIA was.
Unidentified speaker (UNKNOWN) ·14:43
How to verify: Compare Broadcom's and NVIDIA's leverage ratios and credit support agreement disclosures in their respective SEC filings. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Primary sourceDonald Trump received a $25 million donation from Jensen Huang in October of last year.
Unidentified speaker (speaker_0) ·21:12
How to verify: Search FEC (Federal Election Commission) campaign finance reports for donations from Jensen Huang to Donald Trump's campaign committee. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
The turmoil surrounding Jensen Huang may have taken policy pressure off Democrats regarding data center issues.
Unidentified speaker (UNKNOWN) ·23:13
How to verify: Review political commentary and news coverage from late 2026 to early 2026 for mentions of reduced Democratic scrutiny on data center policies coinciding with NVIDIA leadership news. For recent events, verify 2026-08-14 through 2026-11-14. Use another period only when the speaker explicitly names it.
Do your own research
- Verify Meta's prediction market app features and user adoption metrics against Polymarket and Kalshi.
- Review Viktor's revenue reports and Slack integration details to assess competitive positioning.
- Check NVIDIA's latest 10-K and 10-Q filings for ROIC calculations and capital allocation strategies.
- Examine ARM's business model and customer overlap with NVIDIA to evaluate succession fit.
- Analyze Anthropic's CFO background and organizational structure to understand their operational style.
- Research NVIDIA's board composition and succession planning history to assess governance quality.
- Compare CEO succession patterns at Microsoft and Amazon to validate the speaker's operational leadership thesis.
Full transcriptexpand
- 0:00
- Unidentified speaker (speaker_0): Meta's rolling out an app for prediction markets. They want to compete with Polymarket and Calci. I think this is more about late to the party and trying to play fast catch up.
- 0:10
- Unidentified speaker (UNKNOWN): Uh
- 0:10
- Unidentified speaker (speaker_0): Mark's done this a few times. Mark is a an impressive executive. Uh but with uh
- 0:17
- Unidentified speaker (speaker_0): digital currency, he was late to the party with Libra and then DM
- 0:23
- Unidentified speaker (UNKNOWN): uh
- 0:23
- Unidentified speaker (speaker_0): and then also Metaverse, who's he was late to the party. around that.
- 0:28
- Unidentified speaker (UNKNOWN): Uh,
- 0:28
- Unidentified speaker (speaker_0): I think on polymarket, Cal sheet competition knockoff, I don't
- 0:33
- Unidentified speaker (UNKNOWN): think
- 0:34
- Unidentified speaker (speaker_0): that is uh is well timed. Okay. Those businesses grow in large part due to the marketing that they avail themselves on Meta's app, by the
- 0:44
- Unidentified speaker (UNKNOWN): way. So
- 0:45
- Unidentified speaker (speaker_0): he has direct distribution, he's figured that out. It's it's not unlike Amazon looking at who's growing
- 0:52
- Unidentified speaker (UNKNOWN): on
- 0:52
- Unidentified speaker (speaker_0): their platform and then reverse engineering from
- 0:55
- Unidentified speaker (UNKNOWN): that.
- 0:55
- Unidentified speaker (speaker_0): You know.
- 0:56
- Unidentified speaker (UNKNOWN): You know,
- 0:57
- Unidentified speaker (speaker_0): Meta can see the growth of of those businesses. Okay.
- 1:02
- Unidentified speaker (speaker_0): Uh
- 1:03
- Unidentified speaker (UNKNOWN): what
- 1:03
- Unidentified speaker (speaker_0): else? The uh in in related news, Claude is also just taking out more of these companies, is building on their platform. Okay. Uh here's one for you. Have you guys heard of Victor? Victor launched in February of this year,
- 1:18
- Unidentified speaker (UNKNOWN): and
- 1:18
- Unidentified speaker (speaker_0): the year-to-date revenue is about 65 to 70 million.
- 1:22
- Unidentified speaker (UNKNOWN): I
- 1:23
- Unidentified speaker (speaker_0): might be underestimating by a bit, but it's around
- 1:25
- Unidentified speaker (UNKNOWN): that.
- 1:26
- Unidentified speaker (speaker_0): So Victor was
- 1:29
- Unidentified speaker (speaker_0): the first to embed AI into Slack. In fact, I think Stuart Butterfield, founder of Slack, is one of their investors.
- 1:37
- Unidentified speaker (speaker_0): And they they've positioned the product as an AI worker
- 1:41
- Unidentified speaker (UNKNOWN): that
- 1:41
- Unidentified speaker (speaker_0): you can talk to in Slack and
- 1:43
- Unidentified speaker (UNKNOWN): it
- 1:43
- Unidentified speaker (speaker_0): interacts with different applications. If you use it, let me know. Curious to hear what you
- 1:48
- Unidentified speaker (UNKNOWN): think.
- 1:48
- Unidentified speaker (speaker_0): Their marketing's phenomenal. It's UGC-based marketing. UGC-based marketing is exploding the age of AI. User-generated user generated content marketing. Okay.
- 1:59
- Unidentified speaker (UNKNOWN): It's
- 2:00
- Unidentified speaker (speaker_0): the same tactic that Higgsfield is using. uh and others. And they're also doing marketing that's niche down to use case. I was checking out their site today.
- 2:10
- Unidentified speaker (UNKNOWN): A
- 2:11
- Unidentified speaker (speaker_0): friend of mine is an indirect LP and then through a fund. So the fund was super excited about all the traction they had.
- 2:17
- Unidentified speaker (UNKNOWN): And
- 2:17
- Unidentified speaker (speaker_0): then Claude stepped into the market. Claude integrated Claude Plawed with Slack. Kind of an obvious
- 2:23
- Unidentified speaker (UNKNOWN): thing.
- 2:24
- Unidentified speaker (speaker_0): So there's some quick money that's being made by rappers.
- 2:29
- Unidentified speaker (speaker_0): Okay? Wrappers of AI.
- 2:34
- Unidentified speaker (speaker_0): But there's no real moat around that stuff. It's
- 2:38
- Unidentified speaker (UNKNOWN): you
- 2:38
- Unidentified speaker (speaker_0): know, it's really tricky. I
- 2:40
- Unidentified speaker (UNKNOWN): mean
- 2:40
- Unidentified speaker (speaker_0): you could argue that Higgsfield doesn't really have a moat. They just have phenomenal go-to-market, Go to market, okay?
- 2:47
- Unidentified speaker (speaker_0): So uh
- 2:49
- Unidentified speaker (UNKNOWN): I
- 2:49
- Unidentified speaker (speaker_0): think we're gonna see this explosion. of all these AI apps coming. Um
- 2:56
- Unidentified speaker (speaker_0): an explosion of these apps coming,
- 2:58
- Unidentified speaker (UNKNOWN): and
- 2:58
- Unidentified speaker (speaker_0): they're gonna grow very quickly, very fast.
- 3:02
- Unidentified speaker (UNKNOWN): I
- 3:02
- Unidentified speaker (speaker_0): don't know how many of them will actually have staying power though. You really have
- 3:05
- Unidentified speaker (UNKNOWN): to you know
- 3:06
- Unidentified speaker (speaker_0): think through a lot to get this stuff right.
- 3:09
- Unidentified speaker (UNKNOWN): Uh
- 3:09
- Unidentified speaker (speaker_0): I think if you're playing in a regulated space, you're you can gonna do a lot better.
- 3:12
- Unidentified speaker (UNKNOWN): All Alright, right,
- 3:13
- Unidentified speaker (speaker_0): claw's Claude's not gonna going go to after regular. regulated categories. They can't. They need a license, and technology companies avoid regulated categories. They will do enablement for regulated categories.
- 3:25
- Unidentified speaker (UNKNOWN): Okay,
- 3:25
- Unidentified speaker (speaker_0): and that's been Claude's position to date around like financial services.
- 3:32
- Unidentified speaker (UNKNOWN): So
- 3:32
- Unidentified speaker (speaker_0): what else? On the Nvidia drama,
- 3:35
- Unidentified speaker (UNKNOWN): you
- 3:35
- Unidentified speaker (speaker_0): can't make this stuff up. I'm waiting for the simulation to end. I'm waiting for the space. sky to crack open
- 3:40
- Unidentified speaker (UNKNOWN): and
- 3:40
- Unidentified speaker (speaker_0): say, hey, look, we we got you. Can't you can't make this up, right? It's
- 3:44
- Unidentified speaker (UNKNOWN): just really
- 3:45
- Unidentified speaker (speaker_0): amazing. Now like now some of
- 3:47
- Unidentified speaker (UNKNOWN): the uh
- 3:48
- Unidentified speaker (speaker_0): rumors are that maybe she wasn't a spy.
- 3:50
- Unidentified speaker (UNKNOWN): That's
- 3:50
- Unidentified speaker (speaker_0): what David Sachs was saying. Uh I think there's something more to it. I doubt that
- 3:58
- Unidentified speaker (UNKNOWN): uh
- 3:59
- Unidentified speaker (speaker_0): the board would be that incompetent to terminate. with so without some basis.
- 4:05
- Unidentified speaker (UNKNOWN): And
- 4:05
- Unidentified speaker (speaker_0): I I don't
- 4:06
- Unidentified speaker (UNKNOWN): think
- 4:07
- Unidentified speaker (speaker_0): uh the uh infidelity would have been a sufficient basis.
- 4:13
- Unidentified speaker (UNKNOWN): I
- 4:13
- Unidentified speaker (speaker_0): don't think that's
- 4:14
- Unidentified speaker (UNKNOWN): it.
- 4:15
- Unidentified speaker (speaker_0): So
- 4:15
- Unidentified speaker (UNKNOWN): uh
- 4:16
- Unidentified speaker (speaker_0): I mentioned it would be a really stupid decision for the board to put the CFO in the CEO role. They did that. The logical person
- 4:24
- Unidentified speaker (UNKNOWN): would
- 4:25
- Unidentified speaker (speaker_0): be Ian the guy who runs the data center business, the largest PL, it's the fastest growing business. He knows the customers, he knows the product.
- 4:32
- Unidentified speaker (UNKNOWN): Um,
- 4:32
- Unidentified speaker (speaker_0): you know, there's a reason why the
- 4:34
- Unidentified speaker (UNKNOWN): the
- 4:34
- Unidentified speaker (speaker_0): current CEOs of Microsoft and Amazon previously ran the largest, fastest growing businesses in those companies. That's how things work. It's the
- 4:46
- Unidentified speaker (UNKNOWN): P
- 4:46
- Unidentified speaker (speaker_0): PL and L owner, it's the business leader that's driving revenue that grows and gets promoted and then runs the business. That's how it's supposed to
- 4:57
- Unidentified speaker (UNKNOWN): be.
- 4:58
- Unidentified speaker (speaker_0): The CFO is not touching the company.
- 5:03
- Unidentified speaker (speaker_0): Just 'cause of C in the title doesn't mean that they're driving the direction of the company.
- 5:08
- Unidentified speaker (UNKNOWN): You
- 5:08
- Unidentified speaker (speaker_0): know, they're running spreadsheets, they're looking at financial budgeting forecasts, but they have less texture and feel for that.
- 5:17
- Unidentified speaker (UNKNOWN): The
- 5:17
- Unidentified speaker (speaker_0): general manager knows it because the general managers on the call. They're they have the pipeline also. They own the PL too. They're talking to the salespeople. Okay.
- 5:27
- Unidentified speaker (UNKNOWN): That's
- 5:27
- Unidentified speaker (speaker_0): why the Anthropic CFO is so good. Um I in my visit to Anthropic
- 5:33
- Unidentified speaker (UNKNOWN): it was
- 5:33
- Unidentified speaker (speaker_0): around two months ago, I talked a bit about that.
- 5:36
- Unidentified speaker (UNKNOWN): You know,
- 5:36
- Unidentified speaker (speaker_0): they have a morning stand-up stand
- 5:38
- Unidentified speaker (UNKNOWN): up,
- 5:39
- Unidentified speaker (UNKNOWN): eight
- 5:39
- Unidentified speaker (speaker_0): AM local time, Pacific time, where they're going through their pipeline every day by category, by industry group, uh and and and building like a real time PNL,
- 5:50
- Unidentified speaker (UNKNOWN): you
- 5:50
- Unidentified speaker (speaker_0): know. But that's a very hands-on CFO that uh is at a young startup that's growing quickly. So it's a bit different. Like he's a a revenue growth leader, right? Of a different profile. That's not like Colette at Nvidia. She's not doing
- 6:07
- Unidentified speaker (UNKNOWN): the
- 6:07
- Unidentified speaker (speaker_0): revenue growth stuff. Same comment I'd have for like any CFO, including some incredible CFOs out there, like Ruth Perrat
- 6:14
- Unidentified speaker (UNKNOWN): at uh
- 6:15
- Unidentified speaker (speaker_0): Google. Now she's the president, actually. She was actually more of a strategic CFO, not what's called a reporting CFO, how Jensen kind of described
- 6:24
- Unidentified speaker (UNKNOWN): uh
- 6:25
- Unidentified speaker (speaker_0): Colette.
- 6:26
- Unidentified speaker (UNKNOWN): But uh
- 6:26
- Unidentified speaker (speaker_0): you know. Ruth would look at capital allocation, what's the return on those bets, uh, the strategic direction, you know, like that's that's that's a different level of uh of uh thinking you know meta CFO doesn't look like
- 6:41
- Unidentified speaker (UNKNOWN): that
- 6:41
- Unidentified speaker (speaker_0): CFO's at the same level you know uh there was uh
- 6:46
- Unidentified speaker (UNKNOWN): an
- 6:46
- Unidentified speaker (speaker_0): interview with uh Alexander Wang
- 6:50
- Unidentified speaker (UNKNOWN): on
- 6:50
- Unidentified speaker (speaker_0): the the Pines podcast, and the CFO pulled him aside and says, you know, if this doesn't work out, it's really bad for us. It's like, okay, that's thank you. We're spending a lot of money. I got it. Okay, great. CFOs can't pull triggers like that. They cannot make big investment decisions like that. You need vision, you need true leadership, uh, and you need good decision making too.
- 7:10
- Unidentified speaker (speaker_0): It's not enough to have the courage to pull the trigger, right?
- 7:13
- Unidentified speaker (UNKNOWN): So
- 7:13
- Unidentified speaker (speaker_0): another rumor is that maybe this. the CEO of ARM will run Nvidia. That's not a bad idea. That's not a bad idea. I haven't spent much time
- 7:21
- Unidentified speaker (UNKNOWN): uh
- 7:21
- Unidentified speaker (speaker_0): studying that CEO, but the business model is very similar. ARM is an IP licensing based type of company.
- 7:28
- Unidentified speaker (UNKNOWN): They
- 7:28
- Unidentified speaker (speaker_0): don't actually make anything. Same thing with NVIDIA. They don't make anything.
- 7:31
- Unidentified speaker (UNKNOWN): It's
- 7:32
- Unidentified speaker (speaker_0): an IP licensing based company. Um
- 7:35
- Unidentified speaker (UNKNOWN): and
- 7:35
- Unidentified speaker (speaker_0): they deal with similar suppliers
- 7:38
- Unidentified speaker (speaker_0): their customers have some overlap too.
- 7:41
- Unidentified speaker (UNKNOWN): So
- 7:42
- Unidentified speaker (speaker_0): the ARM CEO makes
- 7:44
- Unidentified speaker (UNKNOWN): uh
- 7:45
- Unidentified speaker (speaker_0): a lot of sense, just initial reactions.
- 7:48
- Unidentified speaker (UNKNOWN): You
- 7:48
- Unidentified speaker (speaker_0): know, ARM uh IP is in
- 7:52
- Unidentified speaker (UNKNOWN): uh
- 7:52
- Unidentified speaker (speaker_0): all sorts of uh technology, including Microsoft, uh Google,
- 7:58
- Unidentified speaker (UNKNOWN): you
- 7:59
- Unidentified speaker (speaker_0): know, I imagine they're an Android, also Dell.
- 8:03
- Unidentified speaker (speaker_0): So that one that one really seems to make sense. But there's a lot more that you need. That's not enough just to have those sides of it.
- 8:11
- Unidentified speaker (UNKNOWN): You
- 8:11
- Unidentified speaker (speaker_0): know, th I saw
- 8:12
- Unidentified speaker (UNKNOWN): uh
- 8:12
- Unidentified speaker (speaker_0): that these uh employees at Nvidia put together a petition to bring Jensen back. That's great. It's great to see that, you know. There's a lot of stories coming out about Jensen now. Like he was this, you know, uh micromanager. Um
- 8:27
- Unidentified speaker (speaker_0): but then you also hear stories about how he helped one of his employees who had a medical issue cut through all the red tape and get him to the best care at a hospital. Like, hey, that's that counts as something too.
- 8:43
- Unidentified speaker (UNKNOWN): I
- 8:43
- Unidentified speaker (speaker_0): don't think the micromanagement Things is actually a negative. I mean he the guy built the four trillion dollar company. It's like, okay, you are earned it. Whatever you did to get there worked. You know, when a CEO builds a team around them, the CEO is filling gaps in their skills.
- 9:03
- Unidentified speaker (UNKNOWN): Okay,
- 9:03
- Unidentified speaker (speaker_0): that's what makes this hard. It's not like you have a number two person
- 9:07
- Unidentified speaker (UNKNOWN): in
- 9:08
- Unidentified speaker (speaker_0): the wings that's identical to the CEO. That just doesn't exist because CEOs are surrounding themselves with people
- 9:15
- Unidentified speaker (UNKNOWN): that
- 9:16
- Unidentified speaker (speaker_0): have specialized skills uh that are counterparts to their own um expertise and skills. Okay? The closest person you can find to CEO is a general manager of the fastest growing business. That's the closest you can find.
- 9:32
- Unidentified speaker (UNKNOWN): So
- 9:32
- Unidentified speaker (speaker_0): that's the usual move that's made when you have to deal with succession. So
- 9:37
- Unidentified speaker (UNKNOWN): I
- 9:37
- Unidentified speaker (speaker_0): thought it was really stupid. that they put the CFO in charge and not Ian. It might mean that there is speculation that Ian might be complicit or y knowledgeable about export control violations. He would know all of the customers and he would have
- 9:54
- Unidentified speaker (UNKNOWN): had
- 9:55
- Unidentified speaker (speaker_0): uh as many, if not twice or three times as more customer interactions than Jensen. So
- 10:02
- Unidentified speaker (UNKNOWN): uh
- 10:03
- Unidentified speaker (speaker_0): he he would have known about you know the Singapore based company and doing the exports
- 10:07
- Unidentified speaker (UNKNOWN): there.
- 10:08
- Unidentified speaker (speaker_0): Uh
- 10:09
- Unidentified speaker (UNKNOWN): so
- 10:09
- Unidentified speaker (speaker_0): that you know that you put that on the bearish side of the ledger.
- 10:13
- Unidentified speaker (UNKNOWN): You know,
- 10:13
- Unidentified speaker (speaker_0): side of the border is is really, really
- 10:16
- Unidentified speaker (speaker_0): bad judgment or there's something there. That's um
- 10:22
- Unidentified speaker (UNKNOWN): i
- 10:22
- Unidentified speaker (speaker_0): it it wouldn't make sense to just hey have an open field. Like you got gotta a default default to a jockey to ride the horse that's most capable and suited at that moment in time. You don't have time to say, let's just do a three month search with uh Hydric and Struggles. You don't have time for that. Even if you're doing that, you gotta
- 10:42
- Unidentified speaker (speaker_0): get a jockey
- 10:44
- Unidentified speaker (UNKNOWN): uh
- 10:44
- Unidentified speaker (speaker_0): because decisions are being made every single day, especially with a guy like Jensen who's hands hands-on on with a the flat organizational structure.
- 10:51
- Unidentified speaker (speaker_0): So, you know, the the people on this board, I mentioned yesterday, they're more VC oriented, Goldman Sachs, Cooley,
- 11:00
- Unidentified speaker (UNKNOWN): uh,
- 11:00
- Unidentified speaker (speaker_0): but none of them are operators, none of them are founders.
- 11:05
- Unidentified speaker (speaker_0): So they're not gonna look at this, I think, the right way. That maybe that's a gap.
- 11:13
- Unidentified speaker (speaker_0): Uh
- 11:15
- Unidentified speaker (speaker_0): so
- 11:16
- Unidentified speaker (speaker_0): some quick thoughts on uh
- 11:18
- Unidentified speaker (UNKNOWN): on
- 11:18
- Unidentified speaker (speaker_0): NVD. Here's some other interesting highlights that I picked up today. So one is uh some research I I saw come out of uh Goldman around how essentially the next
- 11:29
- Unidentified speaker (speaker_0): twelve to eighteen months of revenue to NVIDIA is is essentially booked, so they'll achieve those numbers. Now we don't know what the impact of China is. That's an open question.
- 11:40
- Unidentified speaker (UNKNOWN): Uh
- 11:41
- Unidentified speaker (speaker_0): we're gonna learn that real soon though when the video reports
- 11:45
- Unidentified speaker (speaker_0): they'll have to adjust their guidance and I hope they take a kitchen sink approach to it and just get whatever bad news out of the way.
- 11:52
- Unidentified speaker (UNKNOWN): That
- 11:52
- Unidentified speaker (speaker_0): would be the right way to do
- 11:54
- Unidentified speaker (UNKNOWN): it.
- 11:54
- Unidentified speaker (speaker_0): That's the right way to do it. Yeah? So
- 11:58
- Unidentified speaker (UNKNOWN): uh
- 11:58
- Unidentified speaker (speaker_0): but they have some time is the point. They have some time. Uh to get the right leadership in place before revenue breaks. Okay.
- 12:09
- Unidentified speaker (UNKNOWN): This
- 12:10
- Unidentified speaker (speaker_0): is a fast-paced market though. It's a fast-paced market, so every day counts. It's not they should act with urgency and also make a quality decision. So not an easy thing to do. Some of the things that were coming out in the reporting, it's interesting to see all this backstory stuff come out.
- 12:27
- Unidentified speaker (UNKNOWN): right?
- 12:28
- Unidentified speaker (speaker_0): The CFO was concerned about all these credit support agreements and the financing that Nvidia was uh doing to entrench with customers and build out the open weights ecosystem and the neo cloud business.
- 12:41
- Unidentified speaker (UNKNOWN): So
- 12:41
- Unidentified speaker (speaker_0): that's old Jensen, right? That came from Jensen. Like another CEO, nine out of ten of them, just would not have done that. The CFO would not have done that.
- 12:50
- Unidentified speaker (UNKNOWN): You'd be
- 12:51
- Unidentified speaker (speaker_0): on a different reality path with anybody else
- 12:55
- Unidentified speaker (UNKNOWN): uh
- 12:55
- Unidentified speaker (speaker_0): other than Jensen. There's no question, it'd be very different business, right? Uh
- 13:01
- Unidentified speaker (speaker_0): so
- 13:02
- Unidentified speaker (UNKNOWN): I
- 13:03
- Unidentified speaker (speaker_0): do think you're gonna see these credit support agreements
- 13:06
- Unidentified speaker (UNKNOWN): uh
- 13:07
- Unidentified speaker (speaker_0): wane
- 13:08
- Unidentified speaker (UNKNOWN): uh
- 13:08
- Unidentified speaker (speaker_0): and I think uh Broadcom will pull back also. Broadcom was forced to do these large finances. financings for their customers, also. I mean, look, they're looking at Nvidia. Imagine you're OpenAI.
- 13:20
- Unidentified speaker (UNKNOWN): You
- 13:20
- Unidentified speaker (speaker_0): say, Hey, Nvidia's offering me this. Can you give me financing? Seller financing, that's what it's called, right? Boeing sells an airplane, Boeing will attach the financing for
- 13:29
- Unidentified speaker (UNKNOWN): it.
- 13:30
- Unidentified speaker (speaker_0): Okay?
- 13:31
- Unidentified speaker (UNKNOWN): It's
- 13:31
- Unidentified speaker (speaker_0): called layaway. So you started this stuff a long time ago.
- 13:37
- Unidentified speaker (UNKNOWN): So
- 13:37
- Unidentified speaker (speaker_0): I think the financing for data centers will diminish uh on the margin. Okay. So now
- 13:45
- Unidentified speaker (UNKNOWN): the
- 13:45
- Unidentified speaker (speaker_0): next step is to quantify how much of financing is contingent on these credit support agreements. Probably a meaningful amount.
- 13:54
- Unidentified speaker (UNKNOWN): You
- 13:54
- Unidentified speaker (speaker_0): know, they they're you know they they've put a lot behind these, right? So we got I haven't seen anyone quantify or analyze that. That's a very detailed exercise,
- 14:05
- Unidentified speaker (UNKNOWN): but
- 14:06
- Unidentified speaker (speaker_0): someone needs to go do
- 14:07
- Unidentified speaker (UNKNOWN): that.
- 14:08
- Unidentified speaker (speaker_0): Um the financing is still happened though because the uh capital investment is productive. That's the main thing. If capital invested is productive, meaning you earn more than the cost of the capital and the equipment,
- 14:24
- Unidentified speaker (UNKNOWN): okay,
- 14:25
- Unidentified speaker (speaker_0): the cost of capital meaning what's required to find. finance the equipment, then you're gonna make money. The ROIC is uh somewhere between 15 to 30 percent, depending on how you look at it, whatever assumptions you
- 14:38
- Unidentified speaker (UNKNOWN): make. Okay,
- 14:39
- Unidentified speaker (speaker_0): so 15% is more on the on the weaker side. Uh
- 14:43
- Unidentified speaker (UNKNOWN): there are
- 14:43
- Unidentified speaker (speaker_0): a couple other parts of the story that are really interesting to look at. One is Broadcom was forced to extend less. leverage because of how aggressive Nvidia was. That tells you a lot about how markets work. It reminds me of Chuck Prince in 2007 when he said
- 15:01
- Unidentified speaker (UNKNOWN): when
- 15:01
- Unidentified speaker (speaker_0): the music is playing, you must dance. There's this pro-cyclical aspect to uh markets and leverage. And you see it not just at
- 15:13
- Unidentified speaker (UNKNOWN): the
- 15:14
- Unidentified speaker (speaker_0): markets level. but you see it at the company level. You see this in the subprime crisis too.
- 15:20
- Unidentified speaker (UNKNOWN): Okay. Now
- 15:21
- Unidentified speaker (speaker_0): this isn't a subprime crisis, not
- 15:23
- Unidentified speaker (UNKNOWN): that.
- 15:24
- Unidentified speaker (speaker_0): But this is one of the uh
- 15:27
- Unidentified speaker (speaker_0): uh weaknesses of markets and I'm a markets guy. I'm a love I love free markets, okay? One of the weaknesses of markets is that they get carried away and they can overinvest at extremes
- 15:41
- Unidentified speaker (UNKNOWN): and
- 15:41
- Unidentified speaker (speaker_0): they can do excessive lending and excessive capex.
- 15:46
- Unidentified speaker (speaker_0): So we saw we saw that with Broadcom copycatting Nvidia. Really interesting. We saw it
- 15:53
- Unidentified speaker (UNKNOWN): at
- 15:54
- Unidentified speaker (speaker_0): like the company level, not as like this
- 15:57
- Unidentified speaker (UNKNOWN): kind
- 15:57
- Unidentified speaker (speaker_0): of trend on a chart. We saw here's a competitor that wants that's also wants to win,
- 16:04
- Unidentified speaker (UNKNOWN): right?
- 16:04
- Unidentified speaker (speaker_0): Hawk Tan is a fierce competitor, wants to beat Jensen, they can't stand each other, trying to destroy each
- 16:12
- Unidentified speaker (UNKNOWN): other. And
- 16:12
- Unidentified speaker (speaker_0): so he put his balance sheet on the line to extend credit and grow his business. Okay. This is what happened in uh
- 16:21
- Unidentified speaker (UNKNOWN): the
- 16:21
- Unidentified speaker (speaker_0): 2008 crisis. The investment banks would retain what
- 16:26
- Unidentified speaker (UNKNOWN): were
- 16:26
- Unidentified speaker (speaker_0): called the super senior tranches of mortgage-backed securities on their balance sheets and uh enable transactions to get done. They actually not only did that, they acquired the originators of subprime loans, they vertically integrated the financial stack
- 16:46
- Unidentified speaker (speaker_0): end-to-end.
- 16:46
- Unidentified speaker (UNKNOWN): end
- 16:46
- Unidentified speaker (speaker_0): to end.
- 16:48
- Unidentified speaker (speaker_0): So again, this is not that. I just want to reiterate. My main point is
- 16:52
- Unidentified speaker (UNKNOWN): that
- 16:53
- Unidentified speaker (speaker_0): uh
- 16:54
- Unidentified speaker (speaker_0): cycles are cycles. You have excess
- 16:58
- Unidentified speaker (UNKNOWN): uh
- 16:58
- Unidentified speaker (speaker_0): and you also overcorrect on both sides too. So Mark is trying to figure out where do things settle for earnings and
- 17:08
- Unidentified speaker (speaker_0): um
- 17:09
- Unidentified speaker (UNKNOWN): you know
- 17:09
- Unidentified speaker (speaker_0): ROIC and compute and the cost of financing and
- 17:13
- Unidentified speaker (UNKNOWN): you know
- 17:14
- Unidentified speaker (speaker_0): what's gonna happen.
- 17:15
- Unidentified speaker (speaker_0): with this uh open weight ecosystem. Here's another
- 17:20
- Unidentified speaker (UNKNOWN): uh
- 17:20
- Unidentified speaker (speaker_0): irony around the whole thing. So allegedly,
- 17:25
- Unidentified speaker (UNKNOWN): it
- 17:25
- Unidentified speaker (speaker_0): looks like this,
- 17:26
- Unidentified speaker (UNKNOWN): and
- 17:27
- Unidentified speaker (speaker_0): Nvidia indirectly sold chips to China through this Singapore affiliate.
- 17:33
- Unidentified speaker (speaker_0): Those
- 17:34
- Unidentified speaker (UNKNOWN): uh
- 17:35
- Unidentified speaker (speaker_0): racks were then used to power open openweight weight models models by deep. by DeepSeek and Kimmy. DeepSeek and Kimmy
- 17:43
- Unidentified speaker (speaker_0): and
- 17:43
- Unidentified speaker (UNKNOWN): uh
- 17:44
- Unidentified speaker (speaker_0): Alibaba's model, Quinn, they then published these openweight models, which now are about twenty five percent of token share consumption, which is up quite a lot from single digits like two years ago. A lot of software companies are switching to these open weight models. I've mentioned that, I've talked about that before. Okay.
- 18:01
- Unidentified speaker (UNKNOWN): Uh
- 18:02
- Unidentified speaker (speaker_0): you're You're kinda kind of silly if you don't do that because Claw's gonna try to eat your lunch per the earlier point. Yeah? Like you're supposed to do these open weight models provided you can do it in a sandbox environment that's controlled. So you're not shipping your data to China inadvertently. There's like a attack vector uh that can be built into these open weight models.
- 18:22
- Unidentified speaker (speaker_0): Okay, let's take a step back. So Nvidia
- 18:26
- Unidentified speaker (UNKNOWN): are ar
- 18:26
- Unidentified speaker (speaker_0): arguably gave chips to China. China created open weight models which are now being in the US,
- 18:32
- Unidentified speaker (UNKNOWN): and
- 18:32
- Unidentified speaker (speaker_0): those open weight models are now being used to create a flourishing dynamic ecosystem for AI and American innovation. How funny is that?
- 18:42
- Unidentified speaker (speaker_0): It's lowered the cost of compute for a lot of
- 18:46
- Unidentified speaker (UNKNOWN): uh
- 18:46
- Unidentified speaker (speaker_0): startups, these open weight way models. It's made it
- 18:49
- Unidentified speaker (UNKNOWN): uh
- 18:49
- Unidentified speaker (speaker_0): it's made it more accessible. it's made it more competitive, it's keeping anthropic on their toes.
- 18:55
- Unidentified speaker (UNKNOWN): Uh
- 18:56
- Unidentified speaker (speaker_0): all that's a good
- 18:57
- Unidentified speaker (UNKNOWN): thing.
- 18:58
- Unidentified speaker (speaker_0): So
- 18:59
- Unidentified speaker (UNKNOWN): kind
- 18:59
- Unidentified speaker (speaker_0): of funny. You know, I think Jensen was
- 19:01
- Unidentified speaker (UNKNOWN): right. Like
- 19:02
- Unidentified speaker (speaker_0): big picture, you know, China loves to subsidize and make things free, but that's to our benefit. Does it create dependency? Yes, it can. So you gotta got to be hyper-aware hyper aware about
- 19:13
- Unidentified speaker (UNKNOWN): it.
- 19:14
- Unidentified speaker (speaker_0): Uh
- 19:14
- Unidentified speaker (UNKNOWN): But
- 19:14
- Unidentified speaker (speaker_0): if China wants to make on tap intelligence on demand free for Americans or very low cost and 90% cheaper than an American frontier model.
- 19:27
- Unidentified speaker (speaker_0): And again, the revenues don't even go to China. These open way models, they're hosted on American infrastructure. It's open source. Okay? They're
- 19:35
- Unidentified speaker (UNKNOWN): They're
- 19:35
- Unidentified speaker (speaker_0): hosted on open code.
- 19:37
- Unidentified speaker (UNKNOWN): Okay, they're
- 19:38
- Unidentified speaker (speaker_0): hosted on host a in lot of places. Uh
- 19:42
- Unidentified speaker (UNKNOWN): so
- 19:42
- Unidentified speaker (speaker_0): Grok hosts them, Nvidia's acquisition, for example. Okay. So I think he got the right bet overall. Uh and
- 19:52
- Unidentified speaker (UNKNOWN): and
- 19:52
- Unidentified speaker (speaker_0): it looks like China's just trying to undermine the economic dominance of the United States with these open way models and try
- 19:58
- Unidentified speaker (UNKNOWN): to, you know,
- 19:59
- Unidentified speaker (speaker_0): undercut these frontier labs, which just winds up helping a American innovation.
- 20:04
- Unidentified speaker (UNKNOWN): It's
- 20:04
- Unidentified speaker (speaker_0): kind of a funny thing, you know. The way capitalism works, it just drives costs lower and drives quality higher. Okay? That's what's happening. And
- 20:15
- Unidentified speaker (UNKNOWN): you
- 20:15
- Unidentified speaker (speaker_0): know, the system
- 20:17
- Unidentified speaker (UNKNOWN): uh
- 20:17
- Unidentified speaker (speaker_0): is working. You're not really getting any dependency on China here. I mean if they stop refining their models, it's unfortunate, but it's not a give back, yeah.
- 20:27
- Unidentified speaker (speaker_0): So uh what else? Uh
- 20:39
- Unidentified speaker (UNKNOWN): yeah,
- 20:39
- Unidentified speaker (speaker_0): we'll see how this shakes out. I think there's more to the story. We're still learning.
- 20:43
- Unidentified speaker (UNKNOWN): Uh
- 20:43
- Unidentified speaker (speaker_0): the point I'll keep coming back to is I don't expect much in the way of enforcement actions. Trump uh
- 20:50
- Unidentified speaker (UNKNOWN): said
- 20:51
- Unidentified speaker (speaker_0): the other day that he told Jensen to fight for his job.
- 20:56
- Unidentified speaker (UNKNOWN): Okay.
- 20:57
- Unidentified speaker (speaker_0): Now Trump should be independent from the DOJ.
- 21:02
- Unidentified speaker (speaker_0): In theory.
- 21:05
- Unidentified speaker (speaker_0): Okay. He can influence and set priorities with a phone call.
- 21:12
- Unidentified speaker (speaker_0): He's got a twenty-five twenty five million dollar donation for from Jensen in the
- 21:16
- Unidentified speaker (UNKNOWN): you know
- 21:17
- Unidentified speaker (speaker_0): from October of last year. Would you see an enforcement action from the DOJ? That's the question. That comment from Trump suggests that you're not gonna see
- 21:27
- Unidentified speaker (UNKNOWN): an
- 21:28
- Unidentified speaker (speaker_0): enforcement action. It's unfortunate that you gotta look at the world this way,
- 21:32
- Unidentified speaker (UNKNOWN): but you know,
- 21:33
- Unidentified speaker (speaker_0): you kinda have to do that.
- 21:34
- Unidentified speaker (UNKNOWN): So
- 21:35
- Unidentified speaker (speaker_0): I don't think you're gonna
- 21:36
- Unidentified speaker (UNKNOWN): get an
- 21:36
- Unidentified speaker (speaker_0): enforcement action from the DOJ.
- 21:39
- Unidentified speaker (UNKNOWN): You know,
- 21:40
- Unidentified speaker (speaker_0): the DOJ and
- 21:42
- Unidentified speaker (UNKNOWN): the
- 21:42
- Unidentified speaker (speaker_0): FTC, you know, they're not into uh antitrust and you know, taking
- 21:49
- Unidentified speaker (speaker_0): uh corporations to court. That's not one of their priorities,
- 21:53
- Unidentified speaker (UNKNOWN): right?
- 21:54
- Unidentified speaker (speaker_0): Uh
- 21:56
- Unidentified speaker (UNKNOWN): what
- 21:57
- Unidentified speaker (speaker_0): else? Hopefully that was uh helpful.
- 21:59
- Unidentified speaker (UNKNOWN): We'll
- 21:59
- Unidentified speaker (speaker_0): have some exciting news in the coming weeks, the next two weeks or so. So stay tuned for that, okay? Uh lips are sealed. My favorite holidays coming up Thanksgiving. I hope you guys have good travel plans lined up. We had some flurries come out today.
- 22:14
- Unidentified speaker (UNKNOWN): The
- 22:14
- Unidentified speaker (speaker_0): first time my kids are really thrilled and excited to see that. I said, Do you like the snow?
- 22:17
- Unidentified speaker (UNKNOWN): They're like,
- 22:18
- Unidentified speaker (speaker_0): Yeah, I really like the snow. I was like,
- 22:20
- Unidentified speaker (UNKNOWN): you
- 22:20
- Unidentified speaker (speaker_0): know, I'm the guy that's gotta get rid of the snow. There's only so many things machines can do.
- 22:24
- Unidentified speaker (UNKNOWN): In
- 22:24
- Unidentified speaker (speaker_0): front of the house to where the car, we have this
- 22:26
- Unidentified speaker (UNKNOWN): kind
- 22:27
- Unidentified speaker (speaker_0): of brick sidewalk. That
- 22:28
- Unidentified speaker (UNKNOWN): okay
- 22:29
- Unidentified speaker (speaker_0): that takes human labor. We don't have a drone snow clearing machine
- 22:34
- Unidentified speaker (UNKNOWN): here. Uh
- 22:36
- Unidentified speaker (speaker_0): but it is a fun it's a fun season. It's nice to have the seasons. I like the contrast. It it creates and cultivates gratitude for other parts of the season actually. That's the way I look at it anyway. Um But make sure you do get your sunlight Uh
- 22:50
- Unidentified speaker (UNKNOWN): and
- 22:50
- Unidentified speaker (speaker_0): uh we'll check in again. Check out the Lumita Ledger newsletter this weekend.
- 22:55
- Unidentified speaker (UNKNOWN): What
- 22:56
- Unidentified speaker (speaker_0): QQQ is up like what six points in Sunday? We talked about that overpositioning, excessive fear, zero percentile, long short exposure, having a nice rally here. I think uh
- 23:06
- Unidentified speaker (UNKNOWN): you know
- 23:06
- Unidentified speaker (speaker_0): a pullback after up three percent day is very natural, it's healthy digest, digestion and that's gonna happen. Uh
- 23:13
- Unidentified speaker (UNKNOWN): the
- 23:14
- Unidentified speaker (speaker_0): interesting thing is uh here's another thought is
- 23:16
- Unidentified speaker (UNKNOWN): with
- 23:17
- Unidentified speaker (speaker_0): all this stuff around Jensen, like everyone forgot about the Democrats in the midterm and like our data center gonna go down. It's like what's going on with Jensen in a way it it takes the uh maybe the policy pressure off of the Dems
- 23:29
- Unidentified speaker (UNKNOWN): in
- 23:29
- Unidentified speaker (speaker_0): a certain way. It's
- 23:31
- Unidentified speaker (UNKNOWN): to
- 23:31
- Unidentified speaker (speaker_0): talk about data centers, what are they gonna do next?
- 23:34
- Unidentified speaker (UNKNOWN): I
- 23:34
- Unidentified speaker (speaker_0): don't know. I don't know. We'll see. A little bit more speculative there, but we'll see how this uh unfolds. Gonna be interesting. Alright? Stay warm out there. Be well.