As Anthropic and OpenAI race toward record-breaking IPOs, Microsoft CEO Satya Nadella has argued in a viral X post that frontier AI models aren’t where the real value sits. The winners, he says, will build proprietary “learning loops” trained on their own data and judgment—not chase the best model. Here’s why his “human capital vs token capital” pitch matters, and who’s pushing back. As Anthropic and OpenAI race toward record-breaking IPOs, Microsoft CEO Satya Nadella has argued in a viral X post that frontier AI models aren’t where the real value sits. The winners, he says, will build proprietary “learning loops” trained on their own data and judgment—not chase the best model. Here’s why his “human capital vs token capital” pitch matters, and who’s pushing back.
