This paper provides a comprehensive analysis of the development, capabilities, and applications of generative artificial intelligence (GenAI) and large-scale language models (LLMs), focusing on their implications for research and education. It traces the conceptual evolution from artificial intelligence (AI) to machine learning (ML), deep learning (DL), and finally to the Transformer architecture that forms the foundation of modern generative systems. It examines technical aspects such as prompting strategies, word embeddings, and probabilistic sampling methods (temperature, top-k, top-p), as well as the emergence of autonomous agents, considering the opportunities, limitations, and risks these factors pose. It critically evaluates the integration of GenAI across the research process, from ideation and literature review to study design, data collection, analysis, interpretation, and dissemination. While focusing specifically on geography research, the discussion extends to broader academic contexts. It also addresses educational applications of GenAI, including lecture and lesson design, instructional delivery, assessment, and feedback, presenting geography education as a case study. This study focuses on the ethical, social, and environmental issues raised by GenAI, including bias, intellectual property rights, governance, and accountability, as well as emerging technological strategies for mitigating the environmental impact of LLMs. Finally, it considers the short- and long-term future of GenAI, including scenarios for continued adoption, regulation, and potential decline. By situating GenAI within both academic practice and educational contexts, this study contributes to the critical discussion on the transformative potential and social responsibility of GenAI.
