A change has just emerged in ChatGPT that very few expected. This is a minimal adjustment, but one that many people will appreciate.
OpenAI has clarified that, finally, ChatGPT is able to stop using long dashes whenever the user expressly requests it..
This typographic detail had become, during all this time, an almost unequivocal sign that a text had been written by artificial intelligence. It has appeared in many school papers, emails and practically every corner of the Internet.
The proliferation of these long scripts generated criticism of those who used them, accusing them of relying too much on artificial intelligence tools. The fact that chatbots were unable to avoid repeating this punctuation turned the so-called «ChatGPT script» into an almost unwanted brand.
The problem persisted even when users gave specific instructions for ChatGPT to stop using long dashes: the model still included them.
Now, Sam Altman, CEO of OpenAI, has announced the solution: ChatGPT may stop using these scripts if the user so indicates..
What the user should do is to ask ChatGPT to stop using them in its responses and make sure custom instructions are turned onso as not to have to repeat this indication constantly.
That is, the elimination of the long dash will not be applied by default, and it will be up to each user to give this specific instruction to ChatGPT.
With this small adjustment, OpenAI attempts to respond to one of the most frequent complaints in recent months, offering users greater control over the model’s writing style.