“We’re in a world were the capability of the model is a function of how much money you put into it. At what budget should you evaluate these models?” OpenAI Researcher Noam Brown why the AI industry’s traditional benchmark grids are broken.
