The article focuses on the evolving role of UX research in the context of large language models (LLMs) and AI-driven features. It emphasizes the importance of understanding user needs when designing AI systems. For instance, a "summarize" feature must be tailored to its context, like providing concise bullet points in Gmail versus a comprehensive summary in Adobe Acrobat. The flexibility of AI systems can lead to different outputs based on how prompts are structured. NotebookLM serves as a key example, starting as a chat-based tool and evolving to include features like audio summaries. This shift highlights the necessity of defining clear and relevant prompts that align with user expectations. Effective UX research plays a critical role here, guiding teams to focus on what users genuinely require rather than what developers assume they need. In-depth interviews help uncover users' mental models and the criteria they use to evaluate AI outputs. This foundational research helps shape product development, ensuring that features are not only functional but also meaningful and valuable to users.