ChatGPT erases AI 'traces'.

OpenAI claims to have fixed the ChatGPT bug that consistently used hyphens, opening up more customization options so users can control the writing style as they wish.

OpenAI says its ChatGPT tool can now limit or remove hyphens in text when requested by the user. This has been a subject of debate for months, as hyphens are often a distinguishing feature of AI-generated text.

This method of identification is increasingly being applied in checking assignments, emails, online comments, customer service communications, and forum posts.

Many critics argue that the hyphen makes text look "artificial" and suggests the writer is relying on chatbots. On the other hand, some argue that the hyphen was already widely used in English even before major language models became prevalent.

However, the fact that chatbots almost always use this type of punctuation has made the phrase "ChatGPT hyphens" annoying, even though it is not a reliable indicator of AI-generated text.

For a long time, users have reported that ChatGPT continues to automatically insert hyphens even after they have requested the system to stop. This issue has put OpenAI under considerable pressure, especially as debates about AI text recognition capabilities continue to grow.

"If you ask ChatGPT not to use hyphens in its custom instructions, it will do as requested," OpenAI CEO Sam Altman wrote on X. He also described the update as a "small but welcome victory."

According to OpenAI's explanation, refining the model helps ChatGPT better comply with the requirements to avoid hyphens if users explicitly set this in the personalization settings. This means the chatbot won't automatically remove hyphens completely, but users can control the frequency of their appearance.

The latest update is OpenAI's attempt to meet the growing expectations for ChatGPT's customization capabilities, as many users want the AI's writing style to be closer to natural human writing.

Source: https://znews.vn/chatgpt-xoa-dau-vet-ai-post1603036.html