1: \begin{abstract}

2: It is known that a group shift on a polycyclic group is necessarily of finite type. We show that, for trivial reasons, if a group does not satisfy the maximal condition on subgroups, then it admits non-SFT abelian group shifts. In particular, we show that if a group is elementarily amenable or satisfies the Tits alternative, then it is virtually polycyclic if and only if all its group shifts are of finite type. Our theorems are minor elaborations of results of Schmidt and Osin. %Though some of our statements are non-trivial to obtain from literal statements in the literature, we claim no real contribution to the theory of groups or group shifts, and mainly point out some connections which may not have been made explicit. %We do the construction of counterexamples directly in the subshift category. % for his more abstract notion of Markov shifts, but he does not seem to provide subshift models. %\footnote{I didn't read very carefully, maybe he does.}

3: \end{abstract}

4: