AI in research methodology

This paper explores the transformative impact of Generative Artificial Intelligence (GenAI) on scientific research methodology. It contrasts the traditional linear research approach with emerging AI-driven paradigms, highlighting the closed-loop automation demonstrated by frameworks like DOLPHIN. We...

Descripción completa

Detalles Bibliográficos
Autor: Vilasis-Cardona, Xavier
Tipo de recurso: artículo
Fecha de publicación:2026
País:España
Institución:Universitat Ramon Llull (URL)
Repositorio:DAU Arxiu Digital de la Universitat Ramon Llull
OAI Identifier:oai:dnet:dau_________::c9e6e3ba6f0029392b3ab1af1de45129
Acceso en línea:http://hdl.handle.net/20.500.14342/6214
Access Level:acceso abierto
Palabra clave:Generative AI
Research methodology
00
004
Descripción
Sumario:This paper explores the transformative impact of Generative Artificial Intelligence (GenAI) on scientific research methodology. It contrasts the traditional linear research approach with emerging AI-driven paradigms, highlighting the closed-loop automation demonstrated by frameworks like DOLPHIN. We identify five primary use cases for GenAI in science—Literature Review, Gap Finding, Hypothesis Generation, Research Question Refinement, and the Socratic Opponent—and analyze four key tools (Elicit, ResearchRabbit, Scite, Consensus) facilitating these tasks. Finally, we address critical risks such as hallucinations and methodological monoculture, alongside the strategic perspective of the European Commission regarding AI in science.