What does it mean to change one’s mind rationally? Early in my work, this led me into epistemology and rational choice theory, especially models of belief revision under uncertainty. These frameworks are elegant and powerful, but they also raise a deeper issue: rationality is not just a property of equations and arguments. It is a human practice carried out under pressure. That realization shifted the center of gravity in my research. I became increasingly interested not only in the logic of rational change, but in the methods of inquiry that guide it, and in the ethical and psychological assumptions built into those methods.
Over the past decade, my work has focused on the foundations of scientific inquiry, especially in the mind sciences. Scientific methods do not merely discover facts; they privilege certain forms of evidence, sideline others, and quietly shape what counts as a legitimate question. My research examines how this happens and how disciplines can broaden their frameworks without sacrificing rigor.
This line of work developed into an extended collaboration with clinical psychologist and researcher Shahar Gindi. Our book project, Waiting for Einstein: Change and Crisis in the Mind Sciences, argues that psychology’s heavy reliance on tightly constrained quantitative methods has produced a paradox: technical sophistication paired with limited practical reach. We trace how this methodological consolidation emerged, how it affects research and clinical practice, and how a more pluralistic, ethically informed model of inquiry can do better.
From there, the move into AI is a natural next step. AI’s inherit their assumptions about mind and reasoning from psychology and cognitive science, including their simplifications and blind spots. My current research examines how algorithmic systems reshape our concepts of judgment, agency, and responsibility, and how psychological concepts such as tacit knowledge, intuition, and empathy can guide more responsible design and use.
To contact me, please email avital.pilpel@gmail.com.