TMTPOST --  On the second day of OpenAI’s 12-day presentation series, the company introduced its Reinforcement Fine-Tuning Research Program. This initiative aims to enable developers and machine learning engineers to create fine-tuned expert models. The new model customization technology allows developers to use dozens to thousands of high-quality, task-specific models and rank the model's responses based on reference answers. This advancement enhances the model's ability to derive solutions for similar problems and improve accuracy on specific tasks. OpenAI encourages research institutions, universities, and businesses to apply for access, with expectations for positive outcomes in fields such as law, insurance, healthcare, finance, and engineering. The model excels in tasks where the results have objective “correct” answers, which most experts would agree on.