The recent advancements in the research area of large language models (LLMs) and their applications are significantly shaping various fields, particularly in urban analytics, geospatial predictions, and societal bias mitigation. A notable trend is the integration of LLMs with multimodal data sources, such as street view imagery and geographic coordinates, to enhance the precision and granularity of geospatial predictions. This approach is revolutionizing urban planning, disaster management, and public health by providing detailed urban environment analyses. Additionally, there is a growing emphasis on addressing and understanding the biases embedded within LLMs, especially concerning race and gender disparities in human mobility predictions. This focus is crucial for ensuring equitable societal outcomes as LLMs are increasingly applied in decision-making processes. Furthermore, the incorporation of spatial point pattern statistics into deep learning models for terrain feature classification is advancing GeoAI capabilities, offering improved accuracy in spatial relationship representations. These developments collectively indicate a shift towards more nuanced, data-driven, and equitable approaches in urban and spatial research.