The consensus two years ago was that the launch of ChatGPT would benefit OpenAI exclusively, but many successful AI applications have emerged from startups instead.

OpenAI, the company behind ChatGPT, is likely paying Apple significant amounts through subscriptions, which contrasts with Facebook's earlier practices of not sharing revenue with platforms.

The rules of AI businesses are being rewritten; for instance, OpenAI's investment in training models like GPT-4 was speculative but necessary for innovation.

The development of ChatGPT was rooted in a long-term vision of creating safe AI systems that align with human values, utilizing reinforcement learning with human feedback.

The launch of ChatGPT has significantly raised public awareness about AI, making it accessible beyond just large corporations.

People trust ChatGPT a lot, which is interesting because AI can hallucinate, and it should be a technology that users are cautious about.

We were one of the launch partners of OpenAI when they first launched ChatGPT-4, and we immediately recognized its potential to help our mission.

While many view AI primarily through the lens of ChatGPT, I believe AI is underhyped because the advancements in reinforcement learning since then have been extraordinary, enabling complex planning capabilities.

The future of AI will likely involve a range of applications beyond chat, including reasoning, coding, and task automation, which are essential for OpenAI's profitability.