OpenAI has made ChatGPT a better writer. The latest update to the underlying GPT-4o model has boosted AI's creativity.

This new update also helped it leapfrog Google Gemini and regain its top spot in the LLM field. This is a virtual place where users rank the output of two models, not knowing which is which until the end.

OpenAI does not share much about the new update other than that “creative writing capabilities have been leveled up” and that the writing is “more natural and engaging” with “more tailored writing for improved relevance and readability.”

OpenAI has recently added a variety of new features, including bringing Advanced Voice to the web and improving DALL-E image generation. There are rumors of a version of Sora coming soon that will bring AI video to the rapidly evolving AI platform.

The latest update is more “behind the scenes” than flashy new features like a new UI or impressive canvas. It is a change in the way GPT-4o works, one that makes it more creative than the previous generation.

The new version of GPT-4o, also called GPT-4o, is better at handling files uploaded to ChatGPT and provides more insight into their contents.

It was released in secret on the LLM chatbot arena at lmarena.ai (formerly LMSys arena). On this platform, models compete anonymously to see which model scores better against human users.

OpenAI introduced Lmarena.i as an “anonymous-chatbot” last week and quickly surpassed Google's latest model, Gemini-Ex-1114. lmarena is about X. Creative writing, coding, mathematics, etc., he said that it was a “remarkable improvement” over previous versions.

Have you noticed any changes in ChatGPT over the past few days? I have tried it and it seems to be a more complete and engaging answer. This upgrade in creativity may also explain why DALL-E's images have improved.