ChatGPT users can now customize the chatbot’s warmth, enthusiasm, and emoji usage social media post from OpenAI.
These options (as well as similar changes to the exploit of headers and lists in ChatGPT) now appear in the Personalization menu and can be set to More, Less, or Default. Allow users to further customize the tone of ChatGPT, in addition to the existing ability to set a “basic style and tone” – including the professional, genuine and quirky tones available in OpenAI added in November.
ChatGPT’s tone has been an ongoing issue this year, as OpenAI pulled one update because it was “too flattering” and later tweaked GPT-5 to be “warmer and friendlier” after some users complained that the recent model was colder and less warm.
Some scientists and critics of artificial intelligence suggest that chatbots’ tendency to praise users and confirm their beliefs is a “dark pattern” that causes addictive behavior and may have a negative impact on users’ mental health.