EI Search: How to Search for Knowledge in the Age of Generative AI

Abstract

Many widely available search engines prioritise quick access to information, without acknowledging that knowledge is situated across different disciplines. EI Search is the first search engine to empower users’ curiosity about the nature of knowledge and boost epistemic insight into how disciplines work. Unlike conventional search tools that treat knowledge as a monolith, EI Search prompts users to explore complex questions through up to ten disciplines, each offering distinct perspectives. EI Search builds on previous work with the Epistemic Insight Discipline Wheel (DW), an educational graphic that visualises disciplines as interconnected yet distinct fields. Displaying the Discipline Wheel in schools and universities has been shown to empower staff and students to wonder how different disciplines approach a complex topic. However, the DW’s static format limits its capacity to teach epistemic insight and help students to engage with the architecture of knowledge itself. Motivated by the opportunity presented by Generative Artificial Intelligence (GenAI), we reconceptualised the DW into an interactive tool—EI Search—whereby entering a question and selecting disciplines by clicking on them means that users receive pen portraits by GenAI about the approach that each discipline takes. In Phase 1, we developed multifaceted prompts for GenAI to model scholarly knowledge and discipline-specific norms. Phase 2 involved co-creative refinement with educators, technologists, and students, leading to a prototype tailored for educational contexts. In Phase 3, we adapted the tool for secondary schools, gathering data on student interactions. Our findings demonstrate that EI Search effectively ignites epistemic curiosity, fosters agentic learning, and builds disciplinary and interdisciplinary epistemic insight. The research takes place at a pivotal point in the evolution of search engines. We make recommendations about how and why the next generation of search tools should respond to the arrival of GenAI. We also make recommendations for education based on what this research reveals about the power of EI Search to boost knowledge literacy and open users’ minds to the value of taking different perspectives.