Skip to content

← All Q&A

Why do you argue that foundation models have a first-mover disadvantage?

InvestmentLLM Moats

Drawn from Lutz Finger's Forbes column, LinkedIn writing, and Cornell teaching. Sources are cited inline so you can read the originals.

Later entrants build on frontier research at tiny fraction of cost.

Foundation models have a first-mover disadvantage because AI models can train other AI models. Stanford’s Alpaca model was fine-tuned for less than $100, utilizing existing GPT-3 that had cost over $4.6 million. That’s 46,000 times cheaper. Research into frontier models presents an economic disadvantage. If we apply the same 18x discount to Trump’s Stargate $500 billion announcement, someone could do the same value creation with just a $27 billion investment. Models are difficult to protect from other companies or countries.

DeepSeek - New Economic Rules And Regulatory Challenge · Forbes


Have a follow-up? hello@lutzfinger.com. Or pick another question: all Q&A →