markets

How I'm thinking about AI stocks — mid 2026

My current framework for the Big 7 AI names and what I'm watching.


These are my personal views. Not financial advice.

The structural thesis

The Big 7 AI names (Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, Tesla) aren't all the same bet. I separate them into:

Infrastructure picks — Nvidia is the clearest. If AI training and inference scale, they win. The risk is a competitive moat question at the chip level and geopolitical exposure to Taiwan.

Platform picks — Microsoft (OpenAI integration into Azure and Office), Alphabet (Gemini + search moat + TPU infra), Amazon (AWS + Bedrock). These are bets that the cloud incumbents capture AI spend as a layer on top of existing enterprise relationships.

Application picks — Meta is interesting here. Their open-source model strategy (Llama) and ad-targeting improvements from AI are separate value drivers that the market tends to conflate.

What I watch

  • NVDA earnings: GPU demand tells you real AI capex vs. hype
  • Hyperscaler capex guidance: When Azure/AWS/GCP raise capex, it flows to Nvidia and to inference infrastructure broadly
  • AI product adoption metrics: Copilot seats, ChatGPT revenue, Gemini integration depth — real usage vs. headlines

Where I'm positioned

Long Nvidia as the primary AI infrastructure bet. Long commodities (silver especially) as a parallel infrastructure play — AI data centers are driving significant silver demand for electrical connections.

More updates as my thinking evolves.

← back to markets