From Solitude to AI Companionship-Three Phases of a Writing Life and What They Reveal About Rethinking Scholarly Support in Academic Medicine.

I describe three phases of a physician-scientist's writing career: solitary writing, human collaboration, and engagement with generative artificial intelligence (AI) based on large language models. I use these phases to argue that academic medicine should rethink how it supports scholarly work. The central claim, offered from a single case as hypothesis rather than proof, is that a significant obstacle to scholarly productivity can be structural rather than cognitive: the unavailability of appropriate intellectual companionship at the moment of need. In the first phase, characterized by productive solitude, I built an original scholarly voice, but it led, over time, to epistemic loneliness and exhaustion. In the second phase, collaboration with a scientific editor restored my energy, but our interaction was constrained by scheduled availability. The third phase introduced generative AI as an unusually accessible intellectual companion that allowed me to engage at the moment of curiosity, bridged clinical science and the medical humanities, and made it less costly to admit uncertainty. I describe concrete examples in the text and in a supplementary appendix that serves as an extended disclosure of AI use. I argue that generative AI, grounded in deep research, teaching, and clinical experience, extends rather than replaces human thought and that the experience likely speaks to hybrid scholars of many kinds, not to physician-scientists alone. This essay draws implications for faculty development, equitable access, responsible AI use, and the formation of scholarly voice among learners and identifies research questions for the field.