OpenAI has just added an option that allows users to directly adjust the level of enthusiasm of ChatGPT, from friendliness, the way they express emotions to the use of emotions. This is considered a step to give users more control of the experience in the context of the chatbot's voice continuously causing controversy.
According to OpenAI's social media sharing, new settings have appeared in the Personalization section. Users can choose Add-ons, Lessons, or Default for factors like friendliness, enthusiasm, title and list presentation. These customizations help shape ChatGPT's feedback voice more clearly, in addition to basic styles such as Professional, flat Than or Strange introduced by OpenAI since November.
In recent times, ChatGPT's voice has become a topic of controversy. OpenAI had to withdraw an update because it was considered "too slanderous". After that, the company continued to refine the GPT-5 generation towards "more compact and friendly", when many users reported that the new model was somewhat cold and un approachable.
Some scholars and AI critics warn that the trend of chatbot regularly praising and strengthening user trust can be considered a "dark model". According to them, this approach can easily create a sense of depravity and pose a potential risk of negative impacts on mental health if not properly controlled.
