The meeting discussed the use of models for data editing and the value they add. Justin shared that he uses models to edit existing data by inputting his own information and asking the model to provide the desired output. This approach creates a closed loop where the user provides input and instructions for output. The participants also discussed the process of creating custom models, highlighting the need for extensive training data and fine tuning. They suggested using pre-trained tools as a good starting point and then adapting them to create personalized data editing solutions.
(Source: Office Hours Extra - OpenAI 2/10 )
