The book "Algorithms to live by - The computer science of human decisions" [1] has an hypothesis on this in the chapter on explore vs exploit. It is that in th

The book "Algorithms to live by - The computer science of human decisions" [1] has an hypothesis on this in the chapter on explore vs exploit. It is that in the early part of life, we focus on exploration (extreme case: young kids), which includes meeting lots of new people and having many friends. Then as we age, we focus more on the "exploitation" part: focusing on what we liked best. This leads to reducing the circle of friends to a reduced core of very good friends. Of course, it would be reductive to only consider this angle (e.g.: having kids tends to reduce quite a lot social activities, and often happen in this age range). But it's an interesting perspective, and the CS part of it may resonate with the HN crowd ;) [1] http://algorithmstoliveby.com/ http://algorithmstoliveby.com/