The New York Times book section is increasingly looking more like a space for advertisements than, well, a book section. Case in point: this week’s article about Tertulia, a book-recommending app that trawls social media and uses various kinds of SEO witchcraft to recommend books.
Using a mix of artificial intelligence and human curation, Tertulia aggregates book discussions and recommendations from across the web, drawing from social media posts, book reviews, podcasts and news articles to generate reading recommendations that are tailored to individuals’ tastes and interests.
In other words, Tertulia will figure out what books other people are talking about so that you, too, can talk about the same books that everybody else is talking about, so that Tertulia can tell even more people that you, too, are talking about the books that everybody else is talking about. Nothing new to see here.
But the Times’s whole cover is blown when it reports, with a straight face, that Tertulia’s goal is to become “the Netflix of books.” If you’ve paid any attention at all to publishing in the last decade, you’re probably feeling the same dull, throbbing knot of a headache forming that I get every time I read “Netflix for books.” The idea that we can have a data-driven app with algorithmically-determined book recommendations, that this will somehow be as appealing as watching TV or playing games, and that it will somehow make a profit is one of those zombie ideas that keeps coming back, shambling and eager, no matter how many times reality puts it down.
Let me count the ways:
Here is a review of four “Netflixes for Books” from six years ago, where the author is already complaining that the market is saturated with Netflix imitators that disappear after their VC money dries up. Of the four reviewed, you’ll notice that one of the is Amazon, the company that famously makes negative money from books; Scribd, the only piracy site with the cojones to ask people to pay for its stolen content; and two companies I’ve never heard of (Booksfree and Bookmate) that let you pay money to—dig this—borrow books to read at home. How cool is that? Can you imagine how great it would be if, say, every town in the developed world had a place where you could go to borrow books for a few weeks, then return them when you’re done? Frankly, you’re a sucker if you aren’t paying for these services.
Here is The Guardian in 2014 reporting on several companies, including Amazon, struggling valiantly to become Netflix for books.
Here is The Atlantic in 2014 profiling Oyster, which also tattooed NETFLIX FOR BOOKS on its forehead and played around with Peter Thiel’s money for a few years before imploding.
Here is WIRED in 2015, pointing out that the licensing structure for books means that most Netflixes-for-books are actually incentivized to make you read fewer books at a slower rate.
Basically, if you’re a book reporter—much less at the august cubicles of The New York Times—and somebody says “Netflix for books” around you in earnest, you have a professional obligation to laugh in their face and tell the public not to invest in these fools. What you don’t do, especially in the Year of Our Lord 2022 and all its hindsight, is let them explain their business model, their expectations for changing the industry, how their data models allow them to create fine-tuned, highly marketable user profiles with labels like “beach reader” or “cool dad” (yes, really) and then print their claims in the country’s biggest newspaper without mentioning that, by the way, this exact plan has failed a dozen times in a decade. People might subscribe to this app because of reporting like this, instead of spending that money on actual books in actual bookstores. The company in question might use a pull-quote from your gullible reporting, snatching credibility from your famous institution for their scheme. Oops.
You’ll notice, in the list of also-rans for book streaming, that Amazon kept coming up. That’s because Amazon was an early player in the books-as-service area, establishing Kindle Unlimited in 2014. Today, it’s pretty much the only big name left, letting subscribers pay $10 a month to access a catalog of two million ebooks. If anybody could make it work, it’s Amazon. Having had a long free trial with Kindle Unlimited last year, I can safely say that it doesn’t.
The catalog is the problem. Two million books is a lot, but a quick browse through the stacks reveals that the vast majority of them are self-published romances, thrillers, historical novels, and a kind of mutant slumgullion of all three mixed together. I don’t just mean that the quality of writing is low, though it usually is: the financial structure incentivizes bad writing. Kindle Unlimited pays authors from a shared pool based on their total number of pages read, so authors are motivated to pad out their books with extravagant descriptions, pointless exposition, and meandering subplots. In other words, it’s the 19th century all over again, with authors paid by the word and novels ballooning to 900 pages, except the self-published historical erotica crowd has yet to produce a Dickens or Eliot. If these kinds of books are what you really like, then more power to you, but for readers like me, most of Kindle Unlimited’s catalog is digital pulp.
The rest is meager. There are usually a few thousand books from mainstream publishers available, but these are mostly backlist titles added and removed at random. The problem, as I learned here, is that copyright law favors authors over publishers, and very few blue-chip authors want to license their books away to Amazon for fractions of a penny per page. (Funnily enough, the legal situation is exactly reversed with music, where record labels hold all the power to strike deals with Spotify and leave only crumbs for the artists. Spotify itself says that only record labels can issue takedown requests, not artists.) And so the best you can get with Kindle Unlimited, besides all the mush, is usually a few hundred popular novels and nonfiction works which might disappear at any time and are probably available for free at your library, too.
As for curation and suggestion, I probably don’t have to tell you that Amazon isn’t exactly the best place to build up your reading list. Even with every book from everywhere for sale at every price and the mighty metadata machine of Goodreads to help sort them, Amazon’s algorithmically-sorted lists are hopelessly encrusted with paid-placement, SEO nonsense, and junk.
But even if Tertulia manages to accomplish all the lofty goals The New York Times claims it will, I’m convinced that data-driven book recommendations are a solution in search of a problem. I know a lot of heavy readers, and none of them need any help at all with finding their next book. Most of them live in constant fear of the tottering pile of books-to-read on the nightstand, which in my case is often large enough to pose a significant health risk if it falls the wrong way.
Still, I’ve hung around in enough libraries and bookstores to know that a lot of well-meaning readers actually, honestly struggle with finding a good book to read. Thankfully, this is a skill that can be developed. In the interest of helping out, here are a few things that work for me.
Read book reviews. I don’t mean individual reviews of books, but The New York Review of Books or The Times Literary Supplement, or book sections in larger websites and publications. Subscribe to Kirkus—it’s free, and they review everything. Whatever kind of writing you’re interested in, there are websites and journals dedicated to covering it and presenting the best and latest in the field.
Keep a mental map of the territory. You should know the landmarks of your favorite genres or subjects. If you want to know, for instance, what kind of American sci-fi is good, it’s worth investing some time in the classics. Once you’ve figured out that you really like books by Ursula Le Guin or Philip K. Dick, you know that you’re much more likely to enjoy the hundreds of books like them produced in their wake.
Follow good readers. When somebody cares deeply about good books, they tend to point them out for the benefit of others. The internet is full of talented readers and writers with excellent taste who make recommendations and references to books all the time. If you like what they write, follow up on their references and sources, going upstream to the things that they like.
Know where the good books are and who makes them. This can be a store with a killer selection or a section of the library where your interests are particularly strong and browsing is likely to pay off, but it also means knowing a little bit about publishers, imprints. Experience tells me I can safely ignore most books published by the Big Five, but anything from Open Letter Books is worth a look because they make good stuff.
Know your reading goals. I’ve written about this before, and I stand by it: the best readers have projects and stick to them. When I want to know more about something, I start looking for books that sound interesting and I make a list to pick up at stores or libraries.
Keep a list. I mean an actual list, whether it’s on paper or in Google Keep or whatever. I’ve turned most of my internet browsing into a fine-tuned reading-recommending machine, but it’s all pointless if I can’t remember the title or author of a book that I heard about three months ago. I have a file in Obsidian with dozens of books, mostly obscure, academic, and expensive, that I update a few times a week. Most of the time, I don’t actually run out and get the book at all and my interest in it drifts away, but maybe a year or two down the line I’ll get interested in the subject again and there it is, ready for me to track down and find.
If this seems like a lot to do, it’s only because I’m a serious reader and an English teacher. I do this for business and pleasure, so I invest a lot of time and energy into being a good reader with good things to read. Your own investment should match your own goals. There are no shortcuts, though: your book-seeking skills are only as good as the time and energy you invest in them. What you should never do is put your faith in a bunch of geeks with a business plan who promise to do the work for you while you look at some very promising products made by their sponsors. Your reading is yours: own it.
Blog Roll
Here are my posts on the Musement Blog since my last newsletter. This essay had a quick turnaround, since my summer break officially started yesterday, but hopefully there will be more blogging for the next installment.
A language AI at Google has tricked one engineer into thinking its sentient.
Wittgenstein rates Kafka, completing the circuit between the 20th century’s gloomiest philosopher and its gloomiest writer. The two never met, though both were present at an airshow in Brescia in 1909. Kafka wrote about it. Guy Davenport wrote an entrancing short story about that.
Happy reading!