OpenAI Transforms Creative Writing with Latest ChatGPT Update

ChatGPT has often been praised for its impressive writing abilities. However, I personally believe that it will never truly match human creativity in this area. That said, OpenAI has recently released an update designed to enhance the creative writing capabilities of its chatbot. Now powered by GPT-4o and with the potential to utilize the new o1 model, users can expect to engage in more natural and compelling conversations with ChatGPT.

It’s important to note that this update does not apply to the o1-mini and o1-preview models.

This means that even users of the free version of ChatGPT will have access to this exciting update.

What New Features Can We Expect After ChatGPT’s Latest Update?

For your information, ChatGPT is now available on Windows and includes support for advanced voice mode.

Following this latest update, ChatGPT’s sophisticated voice capabilities rely on the rapid and simultaneous inference of text, voice, and audio.

This new feature is only possible with the GPT-4o model, marking an enhancement that previous models, including GPT-4 Turbo, did not have.

Furthermore, when comparing GPT-4o to its predecessors, this recent update significantly improves reasoning skills.

But can this be compared to the current o1 model, known for its absolute reasoning capabilities?

GPT-4o received an update 🎉

With this upgrade, the model’s creative writing abilities have significantly improved—resulting in more natural, engaging, and tailored content that enhances relevance and readability.

It also excels in handling uploaded files, offering deeper insights and more comprehensive responses.

— OpenAI (@OpenAI) November 20, 2024

ChatGPT-4o Becomes More Accessible Post-Update, Yet It Has Its Limits

Indeed, following the latest update, ChatGPT, utilizing the GPT-4o model, is now accessible to all subscribers, regardless of whether they are paying or on the free tier.

However, there are limitations to its usage; it does not allow for unlimited content generation.

In other words, if you are among the free users of ChatGPT, you will notice that the new model is only available sporadically.

Eventually, ChatGPT will redirect you to the GPT-4o mini model. Those subscribed to ChatGPT Enterprise and Team will experience a processing speed approximately five times greater than that of the free version.

Lightweight and Responsive, GPT-4o Mini Surpasses Gemini 1.5 Flash and Claude 3 Haiku

Speaking of the GPT-4o mini model, it has been trained on the same dataset as GPT-4 while consuming fewer resources than its predecessor.

Due to its compact nature, reduced resource consumption, and efficient performance, GPT-4o mini proves to be more responsive. It integrates seamlessly across a diverse range of applications, particularly in code generation.

“GPT-4o is now the smallest, most powerful, and cost-effective model we have ever created,” stated Sam Altman in a recent post.

In benchmarks, GPT-4o mini even outperforms some of the most advanced models from Meta, Anthropic, and Google, including Llama-3-8B, Claude 3 Haiku, and Gemini 1.5 Flash.

This results in an impressive score of 82% on the MMLU tests conducted by Artificial Analysis, compared to 79% for Gemini and 75% for Claude 3 Haiku.

Our blog thrives thanks to our readers. When you make purchases through links on our site, we may earn an affiliate commission.