There are 15,257 books in that list so it might not provide quite the catchy insights you may think from the title alone.

If the books average 100 pages and he read a page every 40 seconds, that works out over 2 years of continuous reading for 16 hours per day. So this isn't real.

From the source HTML:

<!-- Below is a list of books automatically extracted from hundreds of sources on the Web. -->

the beginning of Googlebot! more from the linking page

  Extracting Patterns and Relations from the World Wide Web
   by Sergey Brin.

"We demonstrate a technique for extracting relations from the WWW based on the duality of patterns and relations. We experiment with it by extracting a relations of books. WebDB Workshop at EDBT '98 (postscript)."

Is that a pre-google list kept up for posterity along with the rest of his Stanford page? Because that's a lot of books for someone that age even to have read, let alone to be his favourite subset of the books he has read.

A bunch of 'em seem to be listed twice, in immediate sucession, though, which is weird.

The source says the page was created on Jan 14th, 1998 and it was created automatically from hundreds of sources on the web.

OK, mystery solved. I didn't think he could have possibly read quite that many books by his early twenties, especially given that he was doing a bunch of other stuff at the time.

However, yeah, having Mein Kampf on a list entitled "My Favorite Books" is a bit embarrassing.

I don't think these are "Sergey's Favorite Books", I think this is a meta-list Sergey put together based on the phrase "My Favorite Books"

Going up a level to his main directory you get:

# Extracting Patterns and Relations from the World Wide Web

by Sergey Brin. We demonstrate a technique for extracting relations from the WWW based on the duality of patterns and relations. We experiment with it by extracting a relations of books. WebDB Workshop at EDBT '98 (postscript).

So why is this being up voted if it provides little value?

Because the title is linkbait, and people are automatically upvoting based on how appealing the title sounds, rather than based on the content.

This indicates a more general problem for HN: how do you get people to vote based on content, rather than title?

I'm not sure I have a good, practical solution.

If you figure out how to convince the public not to vote on appearances, please share it in time for the next general election... it'd put a lot of politicians out of work.

So who has been running Google since the last 13 years?

I swear this is some sort of puzzle...the duplicate listings make it seem dynamically generated.

"Mein Kampf" ... really?