AI markets have crystallized. Four years after GPT-3 made scaling laws legible to investors, the first generation of winners is now visible. Elad Gil, an early backer of Harvey, Perplexity, and Character.AI, argues that the era of backing every credible founder on every large problem is over. The field has sorted.
In LLMs, the core players are Anthropic, Google, Meta, Microsoft, Mistral, OpenAI, and X.AI. Three or four of those dominate benchmarks and enterprise adoption. Revenue for leading foundation model companies has gone from zero to billions in roughly three years. Hyperscaler partnerships, Amazon with Anthropic, Google with Gemini, Microsoft with OpenAI, have created funding incentives that are structural, not speculative. Newer entrants like SSI and Thinking Machine Labs are wildcards: potential innovators or eventual acqui-hires. Gil tracks the full taxonomy here, including voice, video, biology, and materials models, and the piece is worth reading for how he maps capital requirements to market position across each category.
The more important argument is about what comes next. Gil identifies a second wave of markets that are today as uncertain as AI code tools were in 2022, before Cursor, Windsurf, and Devin existed. He names them, but their likely winners are not yet clear. That section of the piece is where the real investment thesis lives.
[READ ORIGINAL →]