Imagine an app that knows where you want to go before you even think about it. That's the promise of next Point-of-Interest (POI) recommendation systems. These systems use your past location data to predict your next destination, offering personalized suggestions for restaurants, shops, or entertainment venues. However, traditional POI recommenders face challenges, especially when dealing with new users or understanding the "why" behind their recommendations. A new research paper introduces "GenUP," a method that generates natural language user profiles from location check-ins on social networks like Foursquare. These profiles capture your preferences and routines in a human-readable format, making recommendations more transparent and personalized. Instead of relying on complex calculations comparing your past movements to other users, GenUP creates a concise summary of your typical behavior. This profile acts as a prompt for a large language model (LLM), like those powering ChatGPT, to predict your next move. This approach is not only more efficient than previous methods, but also addresses the "cold-start" problem that plagues traditional systems. New users can simply provide a short description of their interests, allowing the system to generate relevant recommendations even without a detailed history. Testing GenUP on datasets from cities like New York, Tokyo, and Moscow showed promising results. The LLM, guided by the personalized profiles, accurately predicted next POIs, even outperforming some complex, resource-intensive methods. While the research primarily focuses on location prediction, the concept of natural language user profiles opens exciting possibilities for other recommendation systems. Imagine personalized movie recommendations explained by your taste in genres, or product suggestions based on a descriptive profile of your lifestyle. This innovative approach represents a shift towards more human-centered and interpretable AI systems, enhancing our understanding of how and why these systems make the choices they do. While challenges remain in refining the profile generation and ensuring data privacy, the ability to predict human behavior with increasing accuracy and transparency is a significant step forward for personalized AI.