I have been catching up with reading during the precious downtime between jobs, on a large Kindle Scribe that allows me to read full screen PDF. Yesterday I was rereading The Story of the Human Body: Evolution, Health, and Disease by Daniel E. Lieberman (2013). I jumped to Part III, Chapter 10: The Vicious Circle of Too Much (location 4207 of 11,117 | 38%). It has some interesting discussion about how human species became so efficient at storing energy as body fat due to natual selection and why obesity becomes an issue in modern days due to low quality high energy diet. This is the knowledge I already had when I started long distance running since COVID.
At the lunch break, I went to Hacker News and got sucked into a rabbit hole. The discussion has more than 500 comments and someone posted links to more documents about the background of the document they are discussing about. Out of curiosity, I downloaded the one day old PDF/epub from Anthropic, the three year old PDF from Arxiv, and the two month old gist. When I open them on Kindle Scribe with the smallest font, Claude's Constitution (ePub) has 1,1584 locations, Claude's Constitution (PDF) has 84 pages, arXiv:2212.08073v1 [cs.CL] 15 Dec 2022 has 34 pages, Claude 4.5 Opus Soul Document (gist) has 1,736 locations.
I can't help noticing that each these documents is about 1/10 of a full book. They do take a lot of energy to create and consume. I am sure the percentage of people who commented on that Hacker News rabbit hole and actually read through at least one of those three documents is very small, our brain just don't have energy to consume that amount of information. I have spent weeks to finish the Lieberman book and I am sure it took years for Dr. Lieberman to write it.
Now we are in the age of cheap low quality information, nobody has energy to digest and so we just store them everywhere as information fat.