How does OpenAI compete?

Toni Caron-Brown and Benedict Evans discuss the complexities of competition in the AI space, particularly focusing on OpenAI. They explore the challenges of differentiating AI models, given that they are largely based on similar science, data, and infrastructure. Evans suggests that companies can compete by either investing heavily in infrastructure or by innovating above the model layer through new features, applications, or improvements to the chatbot interface itself. The conversation touches on the disconnect between the advanced capabilities being developed and the actual needs and usage patterns of most users, questioning whether the focus is aligned with real-world applications. They also consider whether AI's impact will primarily be through new applications or by enhancing existing workflows, and draw parallels to the early days of the internet, highlighting the uncertainty and open questions that still surround the technology's future.