@Hyolobrika @craig_groeschel it's fairly well covered in here from what I can tell. The individual memory T-cells don't need to live for years, only their clones

> One way to resolve this apparent paradox is to view cellular behaviour from a kinetic viewpoint. As Michie et al. observed in 1992, “… memory is maintained by long-lived clones rather than individual cells with a long lifespan” [17]. The pivotal concept here is that it is the population that must be long-lived, not necessarily the individual cell. Hence, we need to refer to “cells conferring long-lived-memory”, rather than “long-lived memory cells”. This concept is widely accepted and very familiar to people studying the dynamics of T cell memory but has been slow to be adopted into some areas of immunology.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5371741/