With a new reasoning model that matches the performance of ChatGPT o1, DeepSeek managed to turn restrictions into innovation. What will really matter in the long run? That’s the question we ...