OpenAI brings a new GPT-4o model to ChatGPT, adjusted based on user feedback.

OpenAI subtly announced an update of ChatGPT to a new GPT-4o model via X. However, according to the company, this is not a new model, but an updated version based on user feedback. This is leading to mixed reactions from users who want more details about the changes in terms of model behavior and user experience.

Update GPT-4o

The official ChatGPT account posted on X that a new GPT-4o model for ChatGPT has been available since last week. The company did not provide any further details about what the new update includes.

Following the subtle announcement about X, an update followed in a blog about the model, stating: “We have introduced an update to GPT-4o that we have found through experimental results and qualitative feedback that ChatGPT users prefer. It is not a new frontier model. While we would love to tell you exactly how model responses differ, figuring out how to evaluate and communicate model behavior improvements in detail is an ongoing area of ​​research in itself (and one we are working on!).”

Not everyone is happy

This rather vague update to ChatGPT sparked many mixed reactions. According to some users, OpenAI owes customers more explanations to explain exactly what has changed in terms of model behavior and user experience.

Some users on Intrepid speculated that the new GPT-4o model in ChatGPT included step-by-step or multi-step reasoning and more detailed explanations of the processes to the user in natural language. An OpenAI spokesperson told VentureBeat “that there was no new reasoning process in the model update and that ChatGPT describing its reasoning could be triggered by a specific prompt from a user.”