What this covers
Tyler Cowen and Russ Roberts discuss their reading habits, favorite books, and the principles that guide their choices. The conversation spans how to extract lasting value from books, what makes certain works worth returning to, and how reading fits into learning and intellectual life more broadly. Rather than treating individual books as isolated achievements, both speakers argue for a more deliberate architecture to reading that emphasizes context, repetition, and selectivity.
The discussion moves through several interconnected claims about reading practice. Reading in clusters—working through multiple books on related subjects—builds retention and understanding far better than isolated reading, since context accumulated across several works lets ideas crystallize in ways a single book cannot. Rereading great books over a lifetime serves a different purpose: as readers mature and accumulate knowledge, the same text reveals new layers and becomes qualitatively different. A third principle cuts against completionism: refusing to finish books one doesn't love is not failure but discipline, a recognition that opportunity cost matters when a lifetime holds only roughly 2,500 books. Cowen and Roberts also contest the inflated reputation of most nonfiction, arguing that methods and ideas matter more than the books that contain them, and that YouTube increasingly rivals books as a learning medium. The conversation touches on why very long books like Atlas Shrugged and Harry Potter gain outsized cultural weight—partly through costly signaling, where length itself creates club membership value—and explores how fiction like Moby Dick can teach about human irrationality in ways analytical disciplines cannot.
Cowen argues that the value of reading comes not from isolated books or memorization but from building context by reading in clusters, rereading great works across a lifetime, and refusing to finish books you don't love.
- Reading in clusters builds context that makes any single book more meaningful and retainable
- Great books should be reread over the span of a lifetime because they change as the reader matures
- Most nonfiction books are overrated; methods and ideas matter more than individual books
Read only what you love; avoid obligation and maximize opportunity cost against other unread books.
- Cowen avoids highlighting because he could read another book in the time it takes to highlight one, and since the next best unread book is probably quite good, the opportunity cost of highlighting is too high.
“I also find I could read another book in the time it takes me to highlight. The next best book I haven't read is probably quite good. So, why should I highlight it?”
- Giving someone a book can do them harm unless it is the one book in the world they most need to read, because the recipient feels obliged to read it; gifting it to someone who will only use it for reference avoids tricking them into a reading obligation.
“unless I think the book I'm giving is the book in the world the recipient most needs to read, I feel I'm doing the person harm.”
- The single most important reading principle is not to read anything you don't love reading; aside from job requirements, the entire point of reading is to love what you are reading, and if you don't, you should not do it.
“don't read stuff you don't love reading. Simplest point that I would stress above all else. Maybe you have to read it for your job. But, look: The point of reading is that you love what you're reading. If not, don't do it.”
- Cowen folds over pages with something notable but deliberately does not mark what was notable, because on return he will find other things, effectively randomizing his second reading rather than anchoring it to his first impressions.
“typically, I don't mark what was notable because then when I go back, I'll find other things. So, I'm deliberately randomizing my second thought or search a bit.”
- Cowen dislikes the Kindle for serious reading because he remembers facts by recalling visually where they appeared on the page, and the Kindle makes everything 'the same page,' destroying the spatial cues that aid his memory.
“To remember things, I try to remember visually where it was on the page. And that helps me remember the fact. And I can't do that for Kindle. It's all the same page somehow.”
Long books create valuable cultural membership through signaling; certain works disproportionately shape technologist thinking.
- Long books like Atlas Shrugged, Harry Potter, and Piketty have outsized impact because they create an absorbing world that functions as a totem and signal of belonging to a shared culture; if the entry fee (length) were too low, the value of club membership would be diluted, so length itself adds value through economic reasoning about signaling.
“It's a world you get absorbed into. And it's a kind of totem and a signal of absorption into a culture that people share. And if the entry fee is too low, the value of club membership is diluted. So, I think it makes sense and looking at it with economic reasoning.”
- When Cowen meets a tech crowd, the books everyone has read are some mix of Tolkien or I, Robot—not the Bible, Dickens, or Proust—indicating these works are massively and disproportionately influential among technologists.
“If I get together, say, with a tech crowd, the book that everyone has read--it's some mix of Tolkien or I, Robot--not the Bible, not Charles Dickens, better or worse. Not Proust. But everyone has read those books, or almost everyone.”
- Asimov's I, Robot may go down as the most influential real book of the 20th century because it framed how to govern artificial intelligence with laws, and those laws functioned as satiric commentary on Torah law that the robots, like humans, fail to obey—making it theology as well as science fiction.
“I think I, Robot might go down as the most influential book of the 20th century. Not counting something like Mein Kampf”
- Atlas Shrugged, published in 1957, is highly prophetic and underrated as sociology, with its cocktail party scenes anticipating what we now call 'the woke' more accurately than any contemporary critic, even though Cowen finds the novel itself too long and one-sided.
“her cocktail party scenes, her account of what we now call 'the woke'. It's what? Published in 1958 or around then? And she saw this in the 1950s. She so nails it.”
Great books reward rereading over years; translation quality, format, picture books, and YouTube deserve serious evaluation.
- Cowen is always reading something in German and Spanish at any point in time, very slowly, partly as language practice, finding short stories easier and more comprehensible in foreign languages than novels.
“I'm always reading something in Spanish and German at any point in time, but very slowly.”
- For modern books, if the author can't grab the reader fairly quickly, the book may in fact not be good, because enough other readers sample the same way that failing to engage early is a negative signal; this rule does not apply to older books, which were not written to grab readers up front.
“I just start at the beginning and see if it grabs me, and enough other people do that that if the author can't grab you fairly quickly, it may in fact not be a good book. Now, older books are quite different. They were not written to grab people up front.”
- Proust's In Search of Lost Time was poorly served by its original English translation (Scott Moncrieff), reads better in other languages, and is now much better in English via Lydia Davis's newer translation; Cowen read the whole work in German over two years and calls it possibly his best reading experience ever.
“Language is a problem. The original English translation was not very good. Lydia Davis has done a better one. It's better translated into other languages. So, I actually spent two years reading it in German in the 1980s.”
- YouTube is in many ways becoming more potent than books, so most people, including Cowen himself, should evaluate and improve their YouTube consumption rather than dismiss it as inferior to reading.
“YouTube is in many ways becoming more potent than books. So, just evaluate your YouTube consumption and see if you could improve it, would be another tip.”
- Drawing on the Soviet School of Chess as captured in Kotov's How to Think Like a Grandmaster, the way to become better at almost anything is to do exercises with actual feedback that might prove you wrong—for example, annotating a chess game and comparing your analysis to a grandmaster or computer—rather than passively playing through games or staring at the board.
“the way you become better is by doing exercises with actual feedback that might prove you wrong. So, try to annotate a chess game and then compare your ideas against--then a grandmaster, now it would be a computer--and that's much better than just playing through games or staring at the board”
- The very best books should be reread after many years because as the reader gets older and knows more, the books become very different and, in Cowen's view, much better; even then, they should be read twice in a row chapter by chapter rather than finished and restarted, because they are too dense to absorb in one pass.
“The very best books, as you are older and know more, they become very different for the most part. I think they become much better. But even then, you should read them twice in a row when you're reading them.”
- Picture books are greatly underrated as learning tools: pulling a library picture book simply titled after a subject like Venice teaches most people more than reading a conventional history, because it is concise, richly illustrated with photos and maps, to the point, and not partisan or post-modern.
“if you just go to your public library and pull down a picture book--it's probably just titled Venice--most people will actually learn more doing that, reading the picture book.”
Understanding comes from reading clusters and accumulated context across many sources, not single books.
- Reading several books on the same topic in a cluster is more valuable than reading any one of them, because retention and understanding come from accumulated context; rereading a book immediately twice in a row teaches little, but reading four or five other books on the topic causes many more pieces of the original to fall into place.
“I like to read books in clusters; and overall, it's a good book. Much of what Bull says will have much more meaning to me after I've read four or five other books on the 19th century Irish land question.”
- Books are overrated as a learning vehicle; in writing his book Talent, Cowen read remarkably few books but a very large number of articles, reinforcing his view that one should learn methods, read in clusters, avoid obsessing over single books, and read around a project to build context.
“I read remarkably few books to write that book. I read a very large number of articles. And this is maybe getting back to my view that books are overrated.”
- Cowen attributes his recall not to raw memory but to investing in context: he would not remember random numbers better than a median person, but by sampling a topic from many sources he accumulates context that makes information stick.
“I think it's that I'm good at context more than I have a good memory. If you gave me a string of random numbers to remember, I don't think I would remember them better than the median human being.”
Fiction teaches cultural irrationality and human behavior more effectively than political science or analytical methods.
- To understand events like Russia's attack on Ukraine, fiction often serves you better than political science or international relations, because novels like Moby Dick (obsessive quest, believing badly in a screwed-up idea) or Russian fiction teach how a culture produces irrational behavior in ways analytical disciplines do not.
“If you're trying to understand, say, the current war--Russia attacking Ukraine--I think fiction often does you better than to read political science and international relations.”
- Moby Dick is mostly about man's quest for God, with Melville conducting an ideological tour of the theological universe sandwiched between the story of chasing a white whale, and also containing substantial material on law, science, the America of that time, and abolitionism.
“It's mostly about man's quest for God, I think, and Melville doing a kind of ideological tour of the theological universe, and that is sandwiched in between a story about chasing a white whale, and there's plenty in the book about law, and science, and America of that time, and abolitionism.”