↵ select ↓ ↑ navigate esc close

marginal revolution read out: Will Baird, book maven (from my email)

marginal revolution read out ·

Listen: https://serve.buildtall.systems/360dec5a32394f1cfd1994eba1ca46b96e0e99858019ce62e6542b6462c1a2fc.mp3

Source: https://marginalrevolution.com/marginalrevolution/2026/11/will-baird-book-maven-from-my-email.html?utm_source=rss&utm_medium=rss&utm_campaign=will-baird-book-maven-from-my-email


This is marginal revolution read out. Today's post is Will Baird, book maven (from my email), from Marginal Revolution. The post is a reader letter from Will Baird to Tyler Cowen and Alex Tabarrok.

Will Baird, book maven (from my email) opens with Will Baird announcing that he has built a project called Marginalia Library. Using AI and the archives that Tyler and Alex have posted, he found every time a book has been mentioned on Marginal Revolution and created a searchable, browsable catalog. The catalog covers 13,315 mentions of 7,968 books across 7,306 posts. It draws from the entire Marginal Revolution corpus, not just posts in the Books category. The site does not directly quote anything from Marginal Revolution other than post titles, which link to the relevant post where a book was mentioned. In addition to compiling mentioned books, Will Baird had Claude determine whether books were evaluated positively, negatively, or mixed, recommended, or described as self-recommending. This let him identify books Tyler and Alex have both evaluated: there are 98 books they have evaluated similarly, and 10 where they have differed. The project was created almost entirely with Claude Code and the Claude API, though Will Baird used GPT-6 Luna for a first pass over the entire archive. He checked about 2,000 links and 650 labels by hand. He says there are still some mistakes, and each record has a contact address so users can send corrections. He has been a long-time reader of Marginal Revolution and has discovered some of his favorite books by reading about them there. This project was also a way for him to improve his book discovery and learn more about using AI. It is, he says, a labor of love: Marginalia Library does not and will not have ads, tracking, or affiliate links, and he plans to keep it current from the weekly Marginal Revolution AI Archive updates.

Marginalia Library, the first linked item, is the catalog itself. The fetched page shows what a record looks like. For example, it lists Elena Ferrante's The Neapolitan Novels as recommended, with a positive evaluation from Tyler, three mentions from 2015 to 2019, and a topic of friendship and social life. That gives a sense of how the project turns a loose phrase in a blog post into a structured, searchable entry with form, place, period, evaluation, and mention counts. It is the practical core of Will Baird's offer to readers who want to find books by how and when they appeared on the blog.

books you’ve both evaluated is the patterns page linked from the email. Here the catalog moves from individual records to aggregate patterns. It reports 13,314 mentions in 7,306 posts, about one post in six across the blog. Tyler and Alex mention books at much the same rate: 18 percent of Tyler's posts and 17 percent of Alex's. The busiest year was 2017, with 769 mentions. The most-mentioned book is The Bible, with 97 mentions, and the most-mentioned author is William Shakespeare, with 106 across 33 works. The data on staying power shows that Macbeth has been mentioned for 22.8 years, from September 2003 to July 2026, while Interest and Prices by Michael Woodford returned after 22.7 years of silence. The page also tracks genre shifts: fiction fell from about 22 percent of the blog's own book mentions in 2003 to 2007 to about 12 percent in 2023 to 2026, while nonfiction rose. Geographically, 37 percent of the books are about or set in the United States, and 13 percent in the United Kingdom. By period, the twentieth century is the most-covered, with 3,576 books, and by decade the 2000s lead. The catalog also finds that 61 percent of book mentions are of new books published that year or the year before, and 78 percent of the books were first published since 2000. Fourteen percent of mentions are books first written in a language other than English, with German and French leading. Comparing the two bloggers directly, Tyler states a view in 68 percent of his book mentions and Alex in 45 percent; economics is 43 percent of Alex's book mentions against 24 percent of Tyler's. The page closes with the 108 books both evaluated: they agree on 98 and differ on 10. Among the disagreements are Liberal Fascism, Phishing for Phools, The Big Short, and The Infinite Machine. Among the long list of agreements, Capital in the Twenty-First Century gets a negative from both, Atlas Shrugged gets a positive from both, and Wealth of Nations also gets a positive from both. Will Baird's own commentary for this link is simply that this process let him identify those books; it is the data, not the interpretation, that is the headline.

The Fatal Shore, the final item, is a book record in the catalog. It is Robert Hughes's 1986 history of the transportation of convicts to Australia from 1787 to 1868. The catalog classifies it as nonfiction, specifically narrative and argument, with subjects including prisons, punishment and criminal justice, and colonialism and decolonization. Its places are Australia and the United Kingdom, and its period runs from 1787 to 1868. Overall the catalog rates it positive, and Tyler Cowen's evaluation is positive. It has two mentions in two posts from August 2005 to June 2018, both by Tyler. In one mention he indicated that he had read the book. Will Baird says he is already reading The Fatal Shore after finding it in this process, which is exactly the kind of discovery the project is meant to produce.

The post as a whole is a small demonstration of what happens when AI indexing meets a long-running blog's reading history. It turns twenty-plus years of passing book mentions into a searchable, browsable record, and it gives readers a way to see not just what was recommended, but how often, by whom, and over how many years. Will Baird's letter is also a useful example of a reader using AI to build something with no ads, no tracking, and no affiliate links, purely for the sake of better book discovery.

  1. Will Baird, book maven (from my email)
  2. Marginalia Library
  3. books you’ve both evaluated
  4. The Fatal Shore