Chit Chat Stocks Podcast

Episode: Big Tech Investing Expert Rihard Jarc Tells Us Who Is Actually Winning the AI Race

Date: October 29, 2025
Hosts: Ryan Henderson, Brett Schafer
Guest: Rihard Jarc, Technology Investor & Founder of Uncover Alpha

Overview

In this episode, Ryan and Brett host returning guest Rihard Jarc, a technology investor and AI startup founder, to dissect the state of competition among "big tech" companies in the modern AI arms race. The discussion ranges from Amazon’s calculated moves to Google’s stacked advantage, Meta’s monetization strategies, and the financial dynamics fueling Nvidia and OpenAI. Jarc brings a critical, data-driven perspective—often referencing insider data and structural market risks—that leaves listeners with a nuanced view of which companies are best positioned for sustainable success in AI.

Key Discussion Points and Insights

1. Amazon’s AI Position and Strategy

[02:21–11:04]

• Falling Behind? There’s a widespread narrative that Amazon’s AWS is trailing Microsoft and others in AI.
  • Jarc: “...they are playing it more safely than many of the other cloud providers...” (02:21)
• Anthropic Partnership: Amazon owns a significant (10–20%) stake in Anthropic. AWS Bedrock routes the majority of Anthropic’s traffic, but Amazon is diversifying by hosting open-source models too.
  • Interdependence: Anthropic heavily relies on AWS, but Amazon is “forcing” Anthropic to use their custom Trainium chips—helping Amazon improve their in-house silicon.
  • Risk Aversion: Amazon’s cautious approach may stem from Jeff Bezos’ public comments about a possible AI bubble.
  • Potential Upside: As other cloud providers max out data center capacity, AWS stands to benefit from overflow demand.

2. Value and Structure of Tech Giants’ AI Partnerships

[11:04–13:34]

• Pseudo-Ownership: Compared to Microsoft’s arrangement with OpenAI, Amazon’s stake in Anthropic is symbiotic but not controlling.
• Risk Management: Tech giants limit direct ownership to avoid the costs and cash burn; they let external investors fund the AI startups (who then spend that money on their clouds).
  • Jarc: “...after restructuring... Microsoft is going to get like a third maybe or even 30% of the business of OpenAI only...” (08:13)
  • Dilution Risk: Future capital-raises will further erode their equity stakes.

3. Google/Alphabet’s Full-Stack AI Advantage

[10:14–16:42]

• Google’s Unique Structure:

  • DeepMind for model development (e.g., Gemini, Veo).
  • GCP (Google Cloud Platform) for cloud infrastructure—helping optimize at scale.
  • Proprietary TPUs (Tensor Processing Units): The only mature, large-scale in-house silicon alternative to Nvidia. Most internal workloads, including Gemini, are powered by TPUs—not Nvidia.
  • Distribution & Data: The “full stack” from silicon to app to end-user.
  • Quote: “...it's only them and Google, which are kind of like mature enough to offer this kind of stuff.” (13:34)

• Consumer Momentum: Recently, Gemini’s iOS app topped download charts (even over ChatGPT), and Google’s AI products serve “quadrillions” of tokens—a testament to scale.

  • Jarc: “...the trajectory of those tokens is really high because they are serving with AI overviews and AI mode also these AI workloads...” (16:42)
  • Competitive Threat: If Gemini Free proves superior or more cost-effective, Google could undercut OpenAI.

4. Meta’s (Facebook’s) AI Monetization Path

[19:26–27:54]

• Advertising Still King: AI’s biggest role remains ad targeting—improving creative content generation for advertisers, especially smaller businesses, and boosting CPMs (cost per thousand impressions).

• Messaging Apps/e-commerce: “Monetizing WhatsApp” via e-commerce and direct sales integration (Meta taking a cut from transactions, not just ad revenue).

• Customer Service Disruption: Generative AI can automate customer service, a $0.5 trillion market, providing Meta with new sales streams.

• Smart Glasses & Voice AI: Meta's AR devices stand to benefit from advanced AI navigation and assistance in real-world contexts.

• Quote: “With AI you can replace all of that industry or at least most of it. And I think that's a big lift for Meta and brings new revenue which is like more sales revenue again.” (22:16)

5. Microsoft and the Rise of “Neo Clouds”

[27:54–35:07]

• Azure’s Growth: Microsoft Azure has gained share—partially due to joint deals (Office, cloud Integration) and OpenAI partnership.
  • Risk in OpenAI Dependency: OpenAI’s immense compute demand means Azure must juggle serving both them and other clients.
  • Market Fragmentation: New “Neo Cloud” players (CoreWeave, LambdaLabs, etc.) emerge, often equipped with Nvidia GPUs, creating margin pressure for the traditional hyperscalers.
• Shared Infrastructure: Microsoft outsources capacity (e.g., CoreWeave) to fulfill demand while avoiding overcommitment.
  • Defensive Strategy: If there’s a bubble and it bursts, Microsoft may pick up distressed assets.

6. Nvidia—King of the Supply Chain, with Caveats

[36:01–47:21]

• Circular Commitments: Nvidia funds AI startups (e.g., OpenAI) who then use that money to buy more Nvidia chips.

  • Jarc: “...the last lender or last investor is Nvidia.” (36:51)

• Late-Stage Signs? Such creative financing signals a possibly overheated, late-stage market.

• Rapid Depreciation: Fast product cycles mean GPUs are usable (at top value) for only two to three years vs. the six years cited in company filings—raising questions about capex, amortization schedules, and true profitability.

• Quote: “If this is true, if this turns out to be true, then the amortization expense should be double of what it is today. And what this means is that every company that is in this space is not accounting their cost correctly...” (44:00)

7. OpenAI—Both a Bellwether and a Bubble?

[51:51–57:56]

• Burning Billions: OpenAI plans to burn $100B+ through 2030, seeking trillions in future commitments.

• Brand Power: OpenAI is the “Google verb for LLMs.”

• Financing House of Cards? If capital dries up, it could trigger a domino effect impacting Nvidia, AMD, Oracle, and the whole supply chain.

• Microsoft’s Risk Management: Microsoft has the right to match external offers to OpenAI but is hesitating at these capital levels—perhaps sensing unsustainable overextension.

• Quote: “...the numbers just don't make sense at this point anymore for me. And that's why I'm more cautious.” (57:00)

Notable Quotes & Memorable Moments

• On Valuation and Bubbles:

  • “If you look at Jeff Bezos... He did a recent interview where he kind of said that he believes that we're in a bubble, at least an industrial bubble, so basically an AI bubble.” (03:41) – Rihard Jarc
  • “We are at a quite late stage of the cycle where, if you will, the last lender or last investor is Nvidia...” (36:51 – Jarc)

• On Google’s AI Stack:

  • “Google is the only one that is not totally beholden to Nvidia...” (10:14 – Jarc)

• On the Market’s Risk Structure:

  • “Neo clouds do have... if you think about it just from an AI workloads perspective, they're not that far behind from many of the hyperscalers...” (32:22 – Jarc)

• On GPU Depreciation and Financial Risk:

  • “...the topic of what the correct usefulness of life for GPU says or accelerators is, should be on top of minds of investors. Because especially like the business models of many of these Neo clouds, core weaves and other, if you change it from six to three, they are already making losses, but the losses are even bigger.” (44:35 – Jarc)

Timestamps for Key Segments

| Segment | Timestamps | |-----------------------------------------------|-------------------| | Amazon & Anthropic's AI Partnership | 02:21–11:04 | | Strategic AI Ownership Models | 07:35–11:04 | | Google’s Full-Stack AI & TPUs | 11:04–16:42 | | Meta’s Monetization Path | 19:26–27:54 | | Microsoft, Azure & Neo Cloud Competition | 27:54–35:07 | | Nvidia, Circular Deals & GPU Depreciation | 36:01–47:21 | | OpenAI’s Capital Burn & Systemic Risk | 51:51–57:56 | | Rihard's #1 Pick for the Next Decade (Google) | 58:35–62:30 |

Rihard Jarc’s 2030 Outlook & Favorite Investments

If you had to pick one big tech AI winner to own into 2030:

[58:35–62:30]

• Favorite: Google (Alphabet)

  • Full-stack advantage—from chips to cloud to consumer.
  • TPUs offer a unique pricing and negotiation lever over Nvidia.
  • Google can win on price once the focus shifts from training to inference.
  • Ad business and GCP have built-in synergies.
  • “My number one is Google and the reason is... it has the full stack. And I think the TPUs are something which will be Google's most important assets to date.” (58:53)

• Key Reasoning:

  • If OpenAI/Anthropic stumble, Google is most resilient.
  • Inference cost advantage will matter as markets mature.
  • DeepMind alone could justify massive valuation.
  • Despite potential AI search disruption, Google still has the infrastructure, advertiser relationships, and data advantage.

Final Thoughts

Jarc’s summary:

• AI infrastructure is king—companies with full stack and efficient proprietary chips (Google) have the edge.
• Risk is growing as capex and depreciation schedules become decoupled from business realities.
• OpenAI’s and Nvidia’s futures are deeply entwined, but also systemically risky.
• Markets may be underestimating the fragility of the ecosystem if financing dries up.

About The Guest

Rihard Jarc writes "Uncover Alpha," a Substack focused on deep dives into AI tech companies, sub-segments like TPUs, and leverages alternative data such as employment records, interviews, and non-standard datasets.

This summary skips all ads and non-content segments. For anyone seeking a data-driven, skeptical, and insightful take on who’s winning the AI race, this episode is a must-listen.