Predicting human movement is crucial for urban planning, traffic management, and even retail strategies. Traditionally, this has been a tough nut to crack, requiring vast amounts of data and complex models. But what if we could predict your movements using just a few bits of information about you? New research explores how large language models (LLMs), the same tech behind ChatGPT, could revolutionize how we understand human mobility. The study uses readily-available socio-demographic data, like age, job, and education, as a starting point. By feeding this information to LLMs, researchers were able to generate surprisingly accurate daily activity chains – sequences of where a person goes and what they do throughout the day. These AI-generated activity chains were then compared to real-world travel data from the National Household Travel Survey (NHTS) and the Southern California Association of Governments' activity-based model (SCAG-ABM). The results? Impressively low discrepancy scores, especially for common activities like work and home life. This suggests that LLMs can, in fact, infer realistic daily routines based on minimal personal information. This LLM-driven method could bypass the need for enormous datasets, making mobility modeling more accessible and adaptable across different locations and populations. Imagine easily predicting traffic flows in a new city or simulating how a community might react to changes in public transport. While this research shows exciting potential, there are limitations. Current models struggle to predict complex, multi-day activity chains. Future research will delve into multi-day activity prediction and further refine the model, potentially leading to more accurate and insightful urban planning tools.