OpenAI User Reports Significant Performance Issues with GPT-6 Model, Seeks Company Response
A paying user of OpenAI's GPT models, Adelle, has reported significant degradation in the performance of the new GPT-6 model, particularly concerning its ability to assist with creative writing tasks. Adelle, who has relied on GPT as a creative partner for three years, states that GPT-6 fails to generate appropriate text, often inserting 'romantic phrases' or generic movements instead of the detailed frameworks previously provided by older models. The user also notes that the model struggles to complete assigned tasks and, instead of declining requests, generates content that deviates significantly from the user's intent. Adelle has spent several days attempting to adapt the model without success and is concerned about the potential loss of a valuable creative tool, questioning if OpenAI is monitoring these complaints and planning any changes to the post-training process.
