Stuart_Armstrong comments on LINK: TCP and decision theory