(A paper.li being an algorithmically generated “newspaper” of links pulled from your twitter timeline (or based on a given hastag). We’re increasingly finding it a useful way to catch links that we would otherwise miss, especially since we don’t have the time to be on twitter 24/7.
We do occasionally go in and tinker with the layout, i.e., delete links that aren’t of interest, and move up links that are. We’re not sure if the paper.li algorithm takes that kind of feedback into account, but we suspect it does. It does seem like the algorithm is getting better at picking out links that we’re interested in..the machine is learning.)