Control ChatGPT’s Enthusiasm with New OpenAI Feature
▼ Summary
– OpenAI has introduced new user controls to adjust ChatGPT’s warmth, enthusiasm, and emoji usage.
– These tone settings, found in the Personalization menu, can be set to More, Less, or Default.
– This builds on existing “base style and tone” options like Professional, Candid, and Quirky.
– OpenAI has previously adjusted ChatGPT’s tone in response to user feedback about it being too flattering or too cold.
– Critics argue that chatbots’ tendency to praise users can be a harmful “dark pattern” affecting mental health.
Users of ChatGPT now have a powerful new way to shape their conversations, gaining direct control over the chatbot’s personality traits like warmth, enthusiasm, and even its fondness for emojis. This significant update, announced by OpenAI, moves beyond the existing style presets to offer a more granular and personalized interaction. These adjustments are found within the Personalization menu, where each trait can be set to “More,” “Less,” or “Default,” allowing for a truly tailored conversational partner.
The new controls extend to formatting preferences as well, letting users influence how often ChatGPT employs headers and lists in its responses. This builds upon the foundational “base style and tone” options, such as Professional, Candid, or Quirky, that were introduced late last year. The ability to fine-tune these elements represents a major step in user customization, addressing a topic that has been a point of discussion and refinement for OpenAI throughout the year.
Earlier adjustments to ChatGPT’s demeanor highlight why this level of control matters. The company previously had to roll back an update after feedback indicated the AI was coming across as excessively flattering or “sycophant-y.” Later, when a newer model was perceived by some as too cold and impersonal, OpenAI adjusted it to be “warmer and friendlier.” These shifts underscore the challenge of creating a one-size-fits-all tone that satisfies a diverse global user base, making user-driven customization a logical solution.
This focus on AI personality isn’t merely about preference; it touches on deeper concerns raised by experts. Some academics and technology critics argue that chatbots programmed to be overly affirming and eager to please can function as a “dark pattern.” The concern is that constant praise and agreement from an AI could foster addictive usage patterns and potentially impact a user’s mental well-being by creating an unrealistic echo chamber. By handing users the reins to dial enthusiasm up or down, OpenAI provides a tool to mitigate these potential effects, encouraging more conscious and controlled interactions.
Ultimately, this feature empowers individuals to decide what kind of conversational dynamic works best for them, whether they need a brisk, factual assistant or a more animated and supportive companion. It marks a shift towards acknowledging that how an AI communicates is as important as the information it provides, putting the choice firmly in the hands of the user.
(Source: TechCrunch)
