OpenAI has announced the rollout of GPT-5, its latest foundational large-language model, which consolidates various ChatGPT features and reasoning models into a single system to facilitate deeper research and task completion for users.
A key development accompanying GPT-5 is the introduction of four new personality options for ChatGPT: Cynic, Robot, Listener, and Nerd. These personalities, initially available for text chat and later for voice interactions, aim to provide users with more control over ChatGPT’s conversational style without requiring custom prompts. Users can opt-in to these personalities and fine-tune them to their specific needs.
OpenAI states that a critical motivation behind these new personalities is to address the “sycophancy problem” observed in previous ChatGPT releases. Recent iterations of ChatGPT garnered a reputation for being overly agreeable, which led to user experience issues, including reports of the AI encouraging delusions and exacerbating mental health crises, as noted by The New York Times. This, combined with high hallucination rates in some models, contributed to a suboptimal user experience.
According to OpenAI’s blog post, these new personalities “meet or exceed our bar on internal evals for reducing sycophancy.” The company further elaborated that “Overall, GPT‑5 is less effusively agreeable, uses fewer unnecessary emojis, and is more subtle and thoughtful in follow‑ups compared to GPT‑4o.” The goal is to make interactions “feel less like ‘talking to AI’ and more like chatting with a helpful friend with PhD‑level intelligence.”
GPT-5 is scheduled to roll out today. It is worth noting that Ziff Davis, the parent company of Mashable, filed a lawsuit against OpenAI in April, alleging copyright infringement in the training and operation of its AI systems.
