proprietary datasets

“OpenAI’s GPT-5 will be released incrementally, with each new version building on the previous one, and it will be trained on even more data, including”