There is currently no “perfect” recommendation system on the market. hn's algorithm can discover "useful" information with a high probability. No one can define what you consider useful information but yourself. Similarly, for content that has received a lot of attention in the past, there is no fixed standard to evaluate whether the content deserves attention again. In my opinion, hn prefers users to contribute valuable information rather than presenting all valuable information to users.