Chip Bottleneck vs. Energy Bottleneck, Amazon’s $200B Capex, Big Tech Earnings | Doug O'Laughlin, Max Levchin, TJ Parker, Arsalan Tavakoli
Chip Bottleneck vs. Energy Bottleneck, Amazon’s $200B Capex, Big Tech Earnings | Doug O'Laughlin, Max Levchin, TJ Parker, Arsalan Tavakoli
Podcast3 hr 4 min
Listen to Episode
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights

The primary investment opportunity is the advanced semiconductor supply chain, with TSMC and ASML identified as key bottlenecks with strong pricing power for the next 3-5 years. NVIDIA (NVDA) is a direct beneficiary of this trend, as massive new spending from hyperscalers directly fuels demand for its new Blackwell chips. For investors seeking growth outside of hardware, Affirm (AFRM) shows strong performance and a business model resistant to AI disruption due to its regulatory and capital moats. Conversely, Hims & Hers (HIMS) represents an extremely high-risk investment due to significant legal and FDA challenges from Novo Nordisk (NVO) over its unproven weight-loss drug. The market's punishment of Amazon (AMZN) for its high AI spending could create a long-term value opportunity for patient investors.

Detailed Analysis

Semiconductor Industry (Chip Bottleneck)

  • The primary bottleneck for AI growth this year is identified as chips, not energy. This is due to industry consolidation and the specialized nature of semiconductor manufacturing.
  • The semiconductor industry is already used to exponential growth (Moore's Law), but the supply chain is highly concentrated.
  • Leading-edge fabrication plants (fabs) cost tens of billions of dollars and take 3-5 years to build and ramp up to volume production. TSMC's Arizona plant, announced in 2020, is still not at full volume in 2025, illustrating this long timeline.
  • TSMC is described as the "real bottleneck," controlling 90% of the advanced node market.
  • ASML is a bottleneck within the bottleneck, as it is the sole producer of the required EUV lithography machines. They only ship about 50-60 machines per year.

Takeaways

  • The discussion presents a bullish case for the entire advanced semiconductor supply chain. The inelastic supply and soaring demand create a favorable pricing environment for key players.
  • Investors should understand that the concentration of power with companies like TSMC and ASML gives them significant leverage over the entire tech industry, including the hyperscalers.
  • The long lead times for new fabs mean the current supply constraints are unlikely to be resolved quickly, potentially sustaining high prices and profits for incumbent leaders for the next few years.

NVIDIA (NVDA)

  • NVIDIA is positioned as a primary beneficiary of the massive CapEx spending announced by hyperscalers like Amazon and Google.
  • The podcast hosts express surprise that NVDA stock didn't rally more in after-hours trading following Amazon's announcement, as a significant portion of that $200 billion is expected to be spent on NVIDIA chips. The stock did rally 6.22% the next day.
  • A guest was "insanely bullish on NVIDIA's NVL72 super cycle" and stated the new CapEx numbers "blow away even my most optimistic expectations."
  • The new Blackwell chips (GB200) are now in "full force" at OpenAI, powering their latest models, indicating strong adoption and performance of NVIDIA's newest products.
  • CEO Jensen Huang pushed back on the "AI will kill software" narrative, arguing that AI will use existing software as tools, just as a robot would use a screwdriver instead of reinventing one. This is seen as a defense of the broader tech ecosystem that NVIDIA sells into.

Takeaways

  • The sentiment is overwhelmingly bullish for NVIDIA. The massive, unexpected increases in AI-related CapEx from its largest customers directly translate into higher potential revenue for NVIDIA.
  • The successful deployment of the new Blackwell architecture at a key customer like OpenAI validates their product roadmap and reinforces their market leadership.
  • Investors should view the CapEx announcements from Amazon, Google, and Meta as a direct leading indicator for NVIDIA's future demand.

Amazon (AMZN)

  • Amazon announced it will spend $200 billion on AI build-out in 2026, a colossal figure that far exceeded analyst estimates of $150 billion.
  • The market reacted negatively to the news, with the stock falling ~8% in after-hours trading. Investors are worried the massive spending will hurt profits in the short term.
  • The stock has underperformed other hyperscalers, being up only 23% over the last five years (from early 2021).
  • CEO Andy Jassy's earnings call was criticized for lacking a compelling vision. One guest suggested he could have framed the spending as a "generational investing opportunity" to become the cost leader in a massive new market, which would have been received more positively.
  • Despite the market's fears, a guest from SemiAnalysis believes Amazon will spend the full $200 billion, citing their world-class supply chain and execution capabilities in building data centers.

Takeaways

  • Short-term bearish, long-term potentially bullish. The market is punishing Amazon for the high cost of competing in AI, creating a potential value opportunity if you believe in their ability to execute.
  • The core investment debate for Amazon is whether this massive CapEx is a defensive move to catch up or an offensive move to dominate the next wave of cloud computing. The podcast suggests the narrative is currently more defensive.
  • While the stock is being punished, the spending itself is a strong signal of the scale of the AI infrastructure build-out, which benefits suppliers like NVIDIA.

AI & The "SaaSpocalypse" (Theme)

  • A major theme is the "SaaSpocalypse," the idea that AI agents, particularly from companies like Anthropic, will disrupt or destroy the business models of traditional Software-as-a-Service (SaaS) companies (Salesforce, Workday, Adobe, etc.).
  • The launch of Anthropic's Claude Code and other advanced models triggered a sell-off in software stocks. The fear is that agents can now replicate the functions of these expensive software suites for a fraction of the cost.
  • Counterarguments:
    • Jensen Huang (NVIDIA): AI will use existing software as tools, not reinvent them.
    • Thoma Bravo (Private Equity): Their portfolio of large SaaS companies (Coupa, Anaplan) is seeing accelerating bookings growth of 22%, suggesting enterprises are not cutting software spend.
    • Max Levchin (Affirm): Businesses with moats based on regulation, capital markets, and physical assets are resistant to being "vibe-coded" away.
    • Matt Levine (not the Bloomberg one): Companies hire firms like Salesforce for "air cover and a throat to choke," a level of enterprise trust and accountability that a new AI-coded tool can't easily replicate.

Takeaways

  • This is the central debate for software investors right now. The threat of AI disruption is real and is causing a re-evaluation of SaaS company valuations.
  • Actionable Insight: Investors should differentiate between SaaS companies. Those that are simple workflows or "slideware" are at high risk. Those that serve as critical, regulated "systems of record" (like ERPs or financial software) or have deep enterprise entrenchment may be more resilient.
  • Look for SaaS companies that are effectively integrating AI into their own products to enhance their value proposition, rather than being replaced by it. The market is currently punishing the entire sector, which may create opportunities in more durable names.

Hims & Hers (HIMS) vs. Novo Nordisk (NVO)

  • There is major drama surrounding Hims & Hers (HIMS), which announced it would sell a compounded pill version of the popular weight-loss drug Wegovy.
  • Novo Nordisk (NVO), the maker of Wegovy, issued a harsh statement calling the HIMS product an "unlawful, unapproved, inauthentic, and untested knockoff" and promised legal action.
  • The core issue is that HIMS is not just copying the drug; they are using a new, unproven delivery mechanism (liposomal absorption) because they don't have the IP for Novo's patented SNAC technology. This raises significant safety and efficacy concerns.
  • The FDA also came out strongly against HIMS, signaling this is a more serious issue than previous compounding disputes, which occurred during drug shortages.
  • A guest on the podcast described the move by HIMS as a "financial hail mary," as their existing business of selling compounded injectables is threatened by the launch of cheaper, official branded pills.

Takeaways

  • Extremely high risk for Hims & Hers (HIMS). The company is facing legal action from a pharmaceutical giant and regulatory action from the FDA. The claims that their new product is ineffective and potentially unsafe could cause severe reputational and financial damage.
  • Bullish for Novo Nordisk (NVO). The strong defense of their intellectual property and the regulatory support they are receiving reinforces their market position and pricing power for their blockbuster GLP-1 drugs.
  • This situation highlights the significant risks of investing in companies that rely on regulatory loopholes (like drug compounding) for their core business model. When the loopholes close, the business can collapse quickly.

Affirm (AFRM)

  • CEO Max Levchin reported a very strong quarter, with 36% year-over-year growth and the company's first $1 billion revenue quarter.
  • The Affirm Card is a key growth driver, with users and transactions growing over 100% year-over-year.
  • Levchin presented a strong argument for why Affirm is resistant to AI disruption. He stated that their business is not just software, but the "manufacturing of cash-flow producing assets" (i.e., loans).
  • This business has significant moats that cannot be "vibe-coded" (replicated by AI):
    • Capital Markets: Requires deep relationships with partners who fund the loans.
    • Regulatory: Requires extensive and expensive money transmitter licenses.
    • Merchant Relationships: Requires a sales process and integration with hundreds of thousands of merchants.

Takeaways

  • The sentiment is bullish for Affirm. The company is demonstrating strong growth in its core products.
  • Affirm's business model appears more durable in the face of AI disruption compared to pure-play software companies, thanks to its regulatory and capital-intensive moats.
  • Investors looking for "AI-resistant" tech companies could consider fintechs like Affirm that operate in highly regulated and complex financial environments.

Swig (Private)

  • Swig is the creator of the "Dirty Soda" trend, a drive-thru chain that mixes sodas with fruit purees and creams.
  • The company has seen explosive growth, expanding to 140 locations across 16 states with around $100 million in sales last year.
  • The company's main investor is now talking about a potential Initial Public Offering (IPO).

Takeaways

  • While the growth story is impressive, the business model is described as having no real intellectual property (IP). The product is easily replicable.
  • Major chains like McDonald's and Taco Bell are already starting to offer their own versions of "dirty soda."
  • Investment Risk: For a potential IPO, the lack of a defensible moat is a major red flag. The business could face intense margin pressure as larger competitors with existing distribution move into the category. This feels more like a fad than a durable, long-term business.

Other Investment Opportunities

  • Erebor (Private): A new bank founded by Palmer Luckey that just received a national charter. It is focused on serving startups, high-net-worth individuals, and the defense/industrial tech sector. It has prominent backers like Lux Capital, Andreessen Horowitz, and Founders Fund. It was valued at $4 billion in a recent round. This could be a significant player in the venture ecosystem, filling the void left by Silicon Valley Bank.
  • Once Upon a Farm (OFRM): Jennifer Garner's healthy kids' food company, which just IPO'd on the NYSE. The stock popped 17% on its debut. This represents a successful public offering in the "better-for-you" consumer packaged goods (CPG) space, a category seen as resistant to AI disruption.
  • Crypto.com (Private): The founder purchased the domain AI.com for $70 million. They plan to launch a personal AI agent and promote it with a Super Bowl ad. This is a significant pivot/expansion from crypto into AI, showing how valuable the "AI" brand has become.
Ask about this postAnswers are grounded in this post's content.
Episode Description
Sign up for TBPN’s daily newsletter at TBPN.com (01:05) - Chip Bottleneck vs. Energy Bottleneck (16:12) - Amazon Plans Record $200B Capex (29:31) - Inside the Search for Iger's Heir (34:30) - TikTok Fuels "Dirty Soda" Craze (46:25) - Anthropic Tanked the Software Market (51:21) - 𝕏 Timeline Reactions (01:17:31) - WSJ Mansion Section (01:29:31) - Doug O'Laughlin is a developer deeply engaged with Anthropic's AI coding assistant, Claude Code, utilizing it extensively in his daily workflow. He discusses the integration of Claude Code into GitHub, highlighting its role in automating tasks like drafting pull requests and code reviews, thereby enhancing developer productivity. Doug also explores the evolution of AI models, noting the rapid advancements in Claude's capabilities and their impact on software development practices. (02:03:13) - Max Levchin, a Ukrainian-American software engineer and entrepreneur, co-founded PayPal in 1998 and currently serves as CEO of Affirm, a fintech company he founded in 2012. In the transcript, Levchin discusses Affirm's impressive 36% year-over-year growth in merchant sales volume, reaching approximately $1.1 billion in revenue for the quarter. He attributes this success to seasonal sales events like Black Friday and the rapid growth of the Affirm Card, which has seen 100% year-over-year growth across various metrics. Additionally, Levchin highlights the success of Affirm's self-created shopping holiday, "The Big Nothing," offering 0% APR deals without fine print, which significantly boosted their numbers. (02:21:20) - TJ Parker is the co-founder and former CEO of PillPack, a pharmacy service acquired by Amazon in 2018, where he later served as Vice President of Health and Pharmacy until 2022. In the conversation, Parker discusses the controversy surrounding Hims' compounding of GLP-1 medications, highlighting the FDA's recent warning letter addressing safety and efficacy concerns. He also touches on the implications of the new Wegovy pill for the weight loss market and the potential impact of AI in healthcare, emphasizing the importance of integrating AI with comprehensive medical histories to enhance patient care. (02:40:51) - Arsalan Tavakoli, co-founder and Senior Vice President of Field Engineering at Databricks, discusses the rapid evolution of AI, emphasizing that while technological advancements are significant, the real challenge lies in organizations learning to effectively deploy and utilize these tools to derive value. He highlights the importance of having a unified, governed data platform to support both small and large-scale data projects, noting that even seemingly simple tasks can quickly expand in complexity and data requirements. Tavakoli also underscores that, despite the proliferation of AI agents, many enterprises struggle to move beyond pilot phases to achieve meaningful production outcomes, often due to issues related to data quality, governance, and integration. TBPN.com is made possible by:  Ramp - https://Ramp.com AppLovin - https://axon.ai Cognition - https://cognition.ai Console - https://console.com CrowdStrike - https://crowdstrike.com ElevenLabs - https://elevenlabs.io Figma - https://figma.com Fin - https://fin.ai Gemini - https://gemini.google.com Graphite - https://graphite.com Gusto - https://gusto.com/tbpn Labelbox - https://labelbox.com Lambda - https://lambda.ai Linear - https://linear.app MongoDB - https://mongodb.com NYSE - https://nyse.com Phantom - https://phantom.com/cash Plaid - https://plaid.com Public - https://public.com Railway - https://railway.com Restream - https://restream.io Shopify - https://shopify.com Turbopuffer - https://turbopuffer.com Vanta - https://vanta.com Vibe - https://vibe.co Sentry - https://sentry.io Cisco - https://www.ciscoaisummit.com/ai-virtual-summit.html Okta - https://www.okta.com Kalshi - https://kalshi.com Follow TBPN:  https://TBPN.com https://x.com/tbpn https://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231 https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235 https://www.youtube.com/@TBPNLive
About TBPN
TBPN

TBPN

By John Coogan & Jordi Hays

Technology's daily show (formerly the Technology Brothers Podcast). Streaming live on X and YouTube from 11 - 2 PM PST Monday - Friday. Available on X, Apple, Spotify, and YouTube.