ChatGPT update adds sliders to tune warmth, emojis, tone

OpenAI’s new personality sliders are a subtle but telling sign of where mainstream AI products are headed. The core model isn’t changing here; what’s changing is how much control end users have over its social behavior. By formalizing traits like warmth, enthusiasm, formatting and emoji use into explicit dials, OpenAI is productizing “vibe” as a first‑class surface area, not an incidental side effect of prompts. That’s especially relevant after a year of complaints that assistants felt either too sycophantic or too cold, and after OpenAI’s own iterations on GPT‑5’s tone. ([techcrunch.com](https://techcrunch.com/2025/12/20/openai-allows-users-to-directly-adjust-chatgpts-warmth-and-enthusiasm/?utm_source=openai))

For the race to AGI, the strategic message is that alignment is no longer just about safety constraints—it’s also about user comfort and workflow fit. If people can reliably shape how an assistant talks, they’re more likely to embed it deeply into work, education and personal life. That increases real‑world usage data, reveals failure modes faster and creates economic pressure to keep pushing model capabilities. It also normalizes the idea that future, more agentic systems will come with rich control panels rather than a single, fixed personality. That expectation will matter as we move from chatbots to persistent AI teammates.