Natural Machine Learning (NML) Shapes the Adaptive Immune System

The adaptive immune system responds to the state of the body to compute networks of immune molecules and cells that protect and maintain the organism. These immune phenotypes are learned by a process of natural machine learning (NML). NML is based on training, prompts, and feedback. Early training is guided by the transfer to the fetus of maternal IgG, by immunological homunculus self-immunity, and by the acquisition of healthy microbiomes among other factors. Signals from the body prompt immune system computation, and repetitive feedback from resulting immune responses enhances the learning process. Mature immune systems respond to training-based treatments such as intravenous IgG, T cell vaccinations, and heat shock proteins. NML clarifies the organization of the immune system and extends immunology beyond classical reduction to clonal selection. NML supports new avenues of research and a novel rationale for treating immune pathology.