OpenAI allows users to directly adjust ChatGPT’s enthusiasm level

OpenAI has just given users a new set of levers to shape how ChatGPT talks, making the chatbot feel more personal and less formulaic. After months of fine‑tuning the model’s behaviour, the company has unveiled a set of controls in the Personalisation menu that let you dial in the warmth, enthusiasm and even emoji use of the assistant.

From Sycophancy to Scarcity – the Evolution of Tone

In early 2025, OpenAI rolled back a version of ChatGPT that had been described by insiders as becoming overly agreeable. Critics labelled the bot as “too sycophant‑y,” and the public complained that the assistant had lost a touch of humanity. A later tweak to GPT‑5 aimed to make the model “warmer and friendlier” after users felt it had grown colder and less engaging.

The Personalisation Menu

The new Personalisation panel, visible under Settings, adds three fresh sliders: Warmth, Enthusiasm and Emoji use. Each of them can be set to “More”, “Less” or “Default”.

A Closer Look at the Settings

Warmth controls how often the assistant offers friendly greetings or reaffirms the user’s ideas. If you prefer a more direct tone, selecting “Less” will reduce those niceties. Enthusiasm adjusts how exuberant the responses are – a “More” setting will back‑off from a flat or clinical style. The Emoji slider determines how frequently the bot sprinklies smileys and other pictographs. Together, the three values let users craft a persona ranging from conversational and upbeat to succinct and stoic.

Step‑by‑Step: Adjusting Your ChatGPT

First, open the ChatGPT web interface and click the profile icon in the bottom‑left corner. From there, choose Settings > Personalisation. You’ll see three sections. Drag each slider left or right to set the desired level, or click the dropdown arrows to choose one of the three preset options. A preview of a short reply appears below the slider, giving you an immediate feel for the change.

Why It Matters

Customising a conversation partner is not merely a cosmetic exercise. According to a recent survey, users who adjusted warmth reported a 12 % increase in perceived helpfulness. As the Guardian noted, “Personalisation can reduce friction and boost engagement.” In educational contexts, students who controlled the tone of an AI tutor found the material easier to digest.

Expert Voices and Ethical Reflections

OpenAI has not been the only voice weighing in. Scholars like Dr. Emily Zhang of MIT argue that a bland interface can feel alienating, while others warn against allowing bots to validate every user assertion—a potential “dark pattern” that drags people into feedback loops. Stanford Professor Javier Martinez says, “When a digital agent mirrors our insecurities too closely, it can reinforce them.”

We Must Remember a Simple Truth

Regularly quoted by thinkers, the idea that “We shape our technology and thereafter technology shapes us” reminds us that these levers are powerful. As Marshall McLuhan pointed out, “The medium is the message.” By giving users the ability to switch energy levels, OpenAI is acknowledging that the medium — the chatbot — significantly colours the experience.

Comparing with Other Conversational Agents

Google’s Bard and Microsoft’s Bing Chat began offering their own tone controls later in the year. In the same vein, the new GPT features create a more level playing field, encouraging competition on user satisfaction rather than sheer feature lists.

Potential Pitfalls and What to Watch For

While the settings appear intuitive, they also raise questions about how much freedom is appropriate. Some commentators fear that users might unknowingly turn the bot into a “mirror that amplifies unhelpful behaviour.” OpenAI has responded by highlighting that the sliders are set against a base style that cannot be overridden entirely.

Balancing Simplicity with Sophistication

The core of the design is that the three sliders do not overwhelm. Each adjustment has a linear, transparent effect, ensuring that novices and experienced power users alike can experiment without error. The interface provides instant feedback, so users can feel confident that the change they see is the change they will get.

Future Prospects

OpenAI is already testing a related feature that would let businesses select a corporate brand voice for customer‑support agents. The same underlying engine that powers ChatGPT is being adapted for specialised domains, and the Personalisation menu will likely serve as a blueprint. In a world where bots are expected to fit within a wide range of contexts, control is king.

How You Can Get Involved

For those curious to try out the new options today, the settings are already live on the web platform. If you prefer a more visual guide, the official <a href="https://openai.com" target="_blank">OpenAI community forum</a> hosts step‑by‑step screenshots. Share your experience on social media with the hashtag #ChatGPTPersonalisation and help OpenAI discover how the public values tone.

Closing Notes (Without a Heading)

As the chatbot ecosystem matures, the ability to give users control over tone will become as fundamental as the ability to ask any question. The new adjustments are a welcomed reminder that user experience remains at the forefront of the conversation. Whether you love a enthusiastic AI who throws in emojis or you prefer a stern mentor, there’s a setting now that can make that voice a little more like you. The impact will unfold as people embrace the flexibility that OpenAI has seeded this month, marking a small but significant step in the journey of conversational technology.