it seems that suggesting media will always be limited unless there's more context given to the user's situation via device or external apis..