[D] Publication rat race for PhD in ML

Currently, PhD programs in ML at top universities require applicants to have multiple first author papers and strong recommendations from reputable researchers. This kind of publication records would probably qualify you for faculty positions if they are in other fields. How could students even achieve those results without already being in top schools themselves?

Edit: My main point here is related to the recommendation requirements. Students outside of top universities have almost no chance to get recommended by well-known researchers …

author faculty fields kind machinelearning multiple papers phd publication race rat recommendations records researchers results students universities