Here are the slides from my session on generative AI in research for participants in the African Doctoral Academy programme. It was an updated version of a similar talk I gave earlier this year.
Abstract
This presentation explores the applications and implications of Generative AI (GenAI) in research. The talk covers the fundamental nature of GenAI as a next-word predictor with multimodal capabilities, highlighting its increasing competence and ubiquity across various domains.
The presentation distinguishes GenAI from traditional search, emphasising its ability to generate content rather than retrieve it from a database. It discusses the importance of prompt engineering in establishing the context for AI-generated responses. And I suggest that we treat GenAI as an expert collaborator while maintaining awareness of its limitations.
Various use cases for GenAI in the research process are outlined, including literature review, idea generation, summarisation, data collection, writing assistance, grant writing, and data analysis. The presentation also addresses the ethical implications of using GenAI in academic contexts, touching on issues of authorship, originality, and transparency.
The talk concludes by exploring the future of AI in research, both as a tool and as a subject of study. It emphasises the continued importance of human input in problem selection, interpretation, and accountability in the research process. The presentation encourages researchers to consider the biases, limitations, and potential applications of GenAI in their respective fields while reflecting on the unique value that human researchers bring to the table in an AI-augmented research landscape.