Alphabet (GOOGL) remains a high-conviction core holding as Google Cloud achieves a massive 63% growth rate and the company uses its proprietary TPU chips to bypass global hardware shortages. Investors should maintain exposure to the semiconductor stack through NVIDIA (NVDA), AMD, and Micron (MU), as the demand for compute is currently described as infinite with no signs of slowing. A significant "geographic wealth transfer" is occurring in rural land and data center infrastructure; look for opportunities in firms like Blackstone or Brookfield that are financing the massive energy and land requirements for AI. While OpenAI is a dominant force, retail investors should adjust liquidity expectations as their IPO is likely delayed until 2027 due to a strategic pivot toward enterprise services. For those looking at the next frontier of software, monitor the "orchestration layer" represented by startups like Blitzy, which automate large-scale enterprise coding and legacy system modernization.

Alphabet / Google (GOOGL)

• Alphabet reported a record quarter with $109.9 billion in revenue (22% year-on-year growth) and $62.6 billion in profit. • Google Cloud is a major growth engine, hitting $20 billion in revenue with a 63% growth rate, outpacing competitors AWS and Azure. • AI is now the primary driver of results across the entire Google ecosystem, particularly in ad targeting, which has allowed revenue to grow even as search volume flattened. • The company is vertically integrated, utilizing its own TPUs (Tensor Processing Units) for internal needs and selling compute capacity to other labs.

Takeaways

• Cloud Dominance: Google Cloud has successfully transitioned from a potential "excised" business unit to a top-tier competitor, making it a core pillar for investors. • Vertical Integration Advantage: Google’s ability to design its own chips (TPUs) provides a significant moat against the global GPU shortage. • Ad-Tech Resilience: AI-driven ad targeting is extending the lifespan of Google’s search revenue, countering fears that AI chatbots would immediately destroy the search business model.

OpenAI

• OpenAI is diversifying its infrastructure, ending its exclusivity with Microsoft Azure to run on AWS, Google Cloud, and Oracle. • The company reportedly missed internal revenue and user growth targets for 2025, leading to a potential delay of its IPO until 2027. • GPT 5.5 is mentioned as being "dramatically amazing" and potentially matching or exceeding the capabilities of Anthropic’s "Mythos" model, particularly in cybersecurity benchmarks.

Takeaways

• Infrastructure Diversification: OpenAI’s move to "date everyone" (AWS, Oracle, etc.) suggests that compute availability is the primary bottleneck for the company, not loyalty to Microsoft. • IPO Timeline: Investors should adjust expectations for a 2027 liquidity event. The delay is attributed to the difficulty of predicting revenue in an exponential market and meeting public reporting standards. • Enterprise Shift: OpenAI is pivoting from a consumer-heavy focus to an enterprise focus, where the "highest value per token" is found.

Blitzy

• Blitzy recently raised $200 million at a $1.4 billion valuation. • The company focuses on autonomous software development for large-scale enterprise codebases (millions of lines of code), rather than just simple code completion. • It acts as an "orchestration layer," using multiple frontier models (Claude, GPT, Gemini) against one another to drive higher code quality.

Takeaways

• Enterprise AI Adoption: Blitzy represents the "Organizational Singularity," where AI is used for massive refactoring and modernization of legacy systems in Global 2000 companies. • High-Touch Model: Unlike "headcount-light" startups, Blitzy is scaling rapidly (aiming for 300 employees) to provide "forward-deployed engineers" to help enterprises adopt AI.

Investment Themes & Sectors

The "Innermost Loop": Chips and Energy

• Demand for compute is described as "infinite." This is driving a massive bull run for the entire semiconductor stack. • Key Tickers Mentioned: NVIDIA (NVDA), AMD (AMD), Intel (INTC), Samsung, Micron (MU), and SanDisk. • Insight: Even "old" chips (like the A100) are not being retired because demand is so high. The "innermost loop" of the economy is now defined by whoever controls chips and the energy to power them.

Data Center Infrastructure (Ocean, Space, and Rural Land)

• Ocean-Based AI: Companies like Panthalasa (backed by Peter Thiel) are building data centers on the ocean to utilize wave energy and natural saltwater cooling. • Orbital Data Centers: StarCloud is raising capital to put H-100s into orbit, powered by solar energy, to bypass land-based regulatory and cooling constraints. • Rural Wealth Transfer: 67% of new U.S. data centers are being built in rural areas. This is described as the "biggest geographic wealth transfer since fracking."

Private Equity (PE) as an AI Catalyst

• Major PE firms (Blackstone, TPG, Brookfield) are partnering directly with AI labs (OpenAI, Anthropic). • Insight: AI will be forced into legacy companies "top-down" via PE owners to drive EBITDA growth, bypassing internal corporate resistance.

Risk Factors

• Geopolitical "Gatekeeping": The White House is considering vetting AI models before release. There is a risk that "frontier labs" might self-censor or create a "compliance moat" that stifles smaller competitors. • Talent as National Security: The case of Manus AI (where China blocked an acquisition by Meta) suggests that AI researchers and code are now viewed as national assets, potentially leading to a "Cold War" for talent. • Compute Constraints: Even the largest companies (Google, Microsoft) are compute-constrained. This "harsh mistress" means that if a company doesn't secure its compute futures now, it may be unable to compete in 2-3 years.