Clinicians recognize the heterogeneity of age-related macular degeneration (AMD) in presentation, progression, and treatment response, as well as the challenges in distinguishing it from other macular degenerations. As part of the 2024 Ryan Initiative for Macular Research meeting, a group of clinician-scientists and basic scientists were convened to consider the question of whether AMD should be classified as a single disorder or a spectrum of conditions. To answer this question, we reviewed research on several "dimensions" that constitute AMD risk or pathogenesis: genetics, ancestry, retinal imaging findings, diet and environment, aging, and outer retinal molecular and cellular pathways. The group reached a consensus that AMD represents a heterogeneous collection of disease states arising from the interplay of these dimensions. This heterogeneity can be conceived of as a "cloud" of AMD phenotypes. Defining subtypes within this "cloud" requires longitudinal cohorts of well-genotyped and phenotyped patients who progress from no AMD through late AMD, analyzed by unsupervised learning. Comparing the AMD subtypes that emerge from this analysis, especially -omics data from each subtype, will illuminate biology that is applicable to certain subtypes of AMD patients and molecular pathogenic mechanisms that universally apply to all AMD. This knowledge will, in turn, drive improved drug development.