Bremmer's baseline bet for the next five years is that the Open AI model will win (against other approaches like the Chinese developing a separate model), though there is substantial uncertainty around this forecast.