AI needs experiments that connect molecular changes to organ function and health over time. This essay explains why existing biological datasets often lack those connections, and proposes using stored samples, longitudinal cohorts and intervention studies to generate data that can help identify treatments for aging.
