The research talk will focus on leveraging large language models (LLMs) for societal advancement, with applications in energy and robotics. LLMs, as a product of AI’s rapid evolution, have shown immense potential in addressing complex challenges. In energy, LLMs support sustainability by enabling advanced time-series forecasting and optimizing fusion energy research while significantly reducing computational costs through Parameter-Efficient Fine-Tuning (PEFT). In healthcare, LLMs act as interactive, interpretable agents, integrating diverse data sources including genomics, medical imaging, and clinical records to enhance diagnostic accuracy, facilitate personalized treatments, and improve accessibility to medical services. The presentation will highlight the capacity of LLMs to minimize resource consumption, improve operational efficiency, and enable equitable, scalable solutions that align with the overarching goals of social good.