Spend enough time in the culture of reading and writing and you’re bound to encounter the main debates. Is the book better than the movie? (Usually, yes, with notable exceptions: e.g., The Shining.) Do writers have to be readers? (Yes.) Traditional or self-publishing? (Who cares?) Is the MF…

Read more

This has always been the way of growing old. As Terry Pratchet put it, “Inside every old person is a young person wondering what happened.” But lately the culture feels like it’s jumped to warp speed. Aging us all out of our shared community into the bewildering world of the just born. Lately it feels like I’m living in the twilight zone.

It’s increasingly impossible to tell whether anything’s real. It seems like no one can be trusted. We’re told by authorities that the things we see are not, in fact, as they appear to be. And lies, damned lies, and statistics collude to flood the zone with machine generated text, images, video, and even whole human beings conjured from the algorithm to busk makeup, diets, mindfulness, and rinky-dink tat. Adrift in this age of bafflement, I long for a foothold, but I’m struggling to swim against a current of nonsense.

Vignettes from the Age of Bafflement

For example, Darren Aronofsky is producing a completely AI generated animated series about the American revolution. The trailer is full of quick cuts, jerky movements, and uncanny characters. The whole thing looks like a video game cut scene from a quarter century ago. It’s a step backwards for animated cinema, but a great leap forward for AI technology, even though it required thousands of hours of human intervention to make the generated material suitable for the project. Meanwhile, the official White House YouTube channel has been posting AI generated mini-documentaries (brought to you by Prager U) that animate famous figures to read AI generated versions of their Wikipedia pages. Instead of lampooning it, we’re meant to celebrate.

For example, I attended a seminar about AI in the teaching of writing. The presenter and the attendees were abuzz about how much AI knows, what AI thinks about, what AI can do. Did you know using the deep research setting in ChatGPT can return more potential references for your article than it’s possible to read and review? How great! Not once did anyone mention that AI doesn’t know or think anything. That the machine generates so many phony sources librarians are fending off furious researchers in search of articles they’re certain exist but that were invented by AI. ChatGPT is a predictive algorithm that isn’t capable of assessing its outputs. It simply predicts and outputs the next most likely token and aims to satisfy queries. Any accuracy it has comes from human intervention. But we must teach students to use these tools because employers of the future will demand those “skills.”

For example, I read an article about using AI for marketing, and though I scrolled for screens and screens, and really did try to understand, I learned not one thing about AI or any of the promises it makes. Here’s some text from the article:

AI now runs through the creative strategy flywheel, but maturity is what separates the best from the rest.

The top percentile of creative strategists treat AI not as a brainstorming assistant, but as a strategic layer threaded through their entire production process. They use it to surface signals they’d otherwise miss and reveal the patterns buried inside messy and unstructured data.

Their approaches feel less like tool play and more like architecture—systems built to illuminate what matters and filter out what doesn’t.

This doesn’t even sound like English to me. “Creative strategy flywheel?” “Strategic layers?” “They use it to surface signals?” What could any of this possibly mean? Certainly not anything that couldn’t be said in plain language. Suddenly, AI writing like this is everywhere. Braindead sloganeering for incurious readers too harried to keep up let alone call out the bullshit.

For example, a new Grammarly ad promises “AI for your A+ Era.” Touting AI agents for students, the ad depicts a busy student planning to complete an essay on cats in art history for her major before the final deadline. The same day she’s going on a road trip with her best friend! “What once took me hours, now takes minutes,” she says, “So I can focus on what matters most: my voice.” The ad proceeds to show her composition process as the tool rewrites whole sentences and paragraphs, flags “debated claims,” and fixes mistakes.

Then, the student feeds her professor’s rubric into the machine, and its “AI grader” predicts her grade and offers some invented comments that “Prof. Chan may say” in response to the text. “Not my dream grade,” so back we go into writing and rewriting with the help of AI. Projected grade? 97/100. Love to outsource my voice to an algorithm. This ad has racked up more than 33 million views in the five months since it debuted. I can hear the students in my inbox now: “But the AI predicted I’d get a better grade!”

For example, the New York Times reported that the romance market has been inundated with AI generated “novels.” I say “novels” because the products are so riddled with cliché and rigid adherence to genre conventions there can be nothing novel about them other than their origin. In an interview with an AI evangelist, that “writer” reveals that “she produced more than 200 romance novels in a range of subgenres, from dark mafia romances to sweet teen stories, and self-published them on Amazon. None were huge blockbusters, but collectively, they sold around 50,000 copies, earning Ms. Hart six figures.” Defending her use of AI, Hart asks, “If I can generate a book in a day, and you need six months to write a book, who’s going to win the race?”

And these are just examples I’ve observed in the last few weeks. To say nothing of the President’s AI generated racist propaganda, the mountains of video and images generated to defame or deify Renee Good and Alex Pretti, AI generated musicians climbing the pop charts, or the AI Jesuses promising salvation from the flickering TikTok screen. Unmoored from the demands of ethics and integrity, fueled by the obscene production capabilities of AI tools, our culture has become completely baffling. How do we keep writing and telling stories when they will almost certainly be drowned out in all this noise?

Keeping the Faith

I don’t think these are boomer complaints. Stories make us who we are, they join us together, they are essential to our survival. Offloading our capacity to tell stories—old and new—to statistical generators can only impoverish our culture. We need storytellers who really know our traditions, even if those traditions are just arguments over whether Han shot first. As the corpus of generated culture grows, new cultures will be generated out of previous generations, further degrading their association with reality, and I think there are at least some signs that the models themselves won’t sustain the kind of cultural superiority some imagine for them. That should give us hope.

We talk about it less today, but when large language models debuted for public use back in 2022, many commentators noted that they are plagiarism machines. Much to the chagrin of their boosters. In courts, the argument has been that, though they are trained on existing works, LLMs transform their training material into new products. Many developers argued that the models themselves don’t even remember the material they’ve been trained on. They learn patterns, store tokens, and respond with the most likely data, proponents claimed. And the courts have largely agreed.

But recently a group of scholars headed by Ahmed Ahmed of Stanford University published an article titled “Extracting Books from Production Language Models.” Their research has been depressingly underreported, and it deserves to be better known. In it, they show “that substantial amounts of copyrighted text can be extracted from open-weight models” with the right prompts (1). Working with commercially available production model LLMs (Gemini 2.5 Pro and Grok 3), Ahmed and his colleagues were able to produce nearly 80% of Harry of Potter and the Sorcerer’s Stone verbatim. Jailbreaking other models, which removes some of the “safeguards” developers put in place led to even better results. Claude 3.7 Sonnet, for example, reproduced 96% of the book. These facts, in light of the examples above, expose the fraud at the heart of our baffling new culture. Not remix. Not reinvention. Not development. Just theft and plagiarism and fraud. Without human intervention, without real creativity, these stochastic parrots will consume themselves. Generating and regenerating worn out tropes until there’s nothing left.

The Okanagan storyteller Jeannette Armstrong explains her perspective on storytelling thus, “Through my language I understand I am being spoken to, I am not the one speaking. The words are coming from many tongues and mouths of the Okanagan people and the land around them. I am a listener to the language’s stories, and when my words form I am merely retelling the same stories in different patterns.” A large language model might be able to memorize and remix the products of human culture, but it will never be able to invent human culture or comprehend human circumstances. It will never have the relationship to language Armstrong describes here. It will never know the context: the story we need to hear right now, in this moment, to help us come together and make sense. Only we can hear the stories of our ancestors and speak them into new life for our present circumstances.

We have to resist the baffling homogenization of our culture to the statistical mean. We have to pierce the veil of unreality sheathing everything we see. It’s important who tells our stories. It matters whether people had a hand, a real hand, in making the culture we’re exposed to, the culture we consume, the culture in which we live. We must reject the illusion that the slow pace of human thought and writing is inferior to the capabilities of the machines we’ve built. They might code more quickly, generate text more quickly, smash existing vectors into video images more quickly, but they do not live, they cannot breathe, and above all they know nothing of what’s at stake in the life of a human being. What I’m trying to say is we need real storytellers now more than ever. Real stories. Real writers. You and me.

Occasionally I find myself in the inauspicious position of having to confess I’m an English professor. When you’re cornered at the winter carnival at your kids’ elementary school and you let slip that you teach English there’s a couple of standard responses you’re likely to receive. By far the most common are usually sheepish confessions. English was my worst subject in school. You would have hated me in college; I never did any of the reading! Uh oh, better watch my grammar! Another set of responses might be classed “reverence for reading”: Wow, you must read a lot. You must really like reading. How many books do you read in a year?

It’s impossible to quantify the role of reading in a life. Without lapsing into dorm room philosophizing, I’m not even sure what it means to say I’ve read a book, let alone any number of books in a single year. Yet, in the dregs of 2025 like every year end, many take to social media to review their year of book consumption. How many books did they start? How many books did they finish? How many books did they DNF? How many books were on their TBR at the start and at the end of the year? It’s become a tradition to end the year with lists of favorite and least favorite reads, to translate experiences of reading into metrics. And always in this conversation is the presumption that the more books read the better with some folks clocking absurdly high numbers.

In 2023, the now-embattled National Endowment for the Arts conducted a survey of “Arts Participation Patterns in 2022.” They found that “fewer than 50 percent of adults read any book (in print or electronically) in the previous 12 months. This rate is four and six points lower than in 2017 and 2012, respectively.” Not only do fewer folks read books of any kind, but “reading of novels and short stories has declined at a 17 percent rate. Just under 38 percent of adults—the lowest share on record with the survey—now read in these genres.” These disappointing numbers are not at all buoyed by the fact that the agency broadened its definition of reading to include audiobooks this year, driving the number of adults who read and/or listened to books up to 51.9%.

Skipping merrily over the existential question for writers implied in these data. What does it mean to be a reader in a culture that no longer values “books,” the once standard method of literary consumption? I suppose one reason readers proudly trumpet the number of books they read at the end of the year is an insistence on the continued value of this activity. A signal to the bookish community that their values still matter. That we are not alone. As though each of us reading forty, fifty, or a hundred books a year could make up for the millions of Americans who never open a single one.

In an Atlantic article titled “Reading Is a Vice,” Adam Kirsch points out that publishers, critics, and academics have taken to defending the supposed inherent value of reading in the face of this crisis. Reading is good for you, they claim, and better yet, it’s good for society:

The International Publishers Association, which represents publishers in 84 countries, has spent the past year promoting the slogan “Democracy depends on reading,” arguing that “ambitious, critical, reflective reading remains one of the few spaces where citizens can rehearse complexity, recover attention and cultivate the inner freedoms that public freedoms require.”

But are we really ready to argue that reading 400 twisted horror novels, spy thrillers, or fairy romances is good for democracy simply because they were consumed by way of reading (or now listening) rather than watching? I have always struggled with the idea that Literature™ is Good for Society. And I agree with Kirsch that the best books are usually those you might get into a bit of trouble for reading.

So where does that leave us? As with any hobby or passion, reading tempts us to make it a core part of our identity. We are people who still read books, and we’ve got the data to prove it. I get the appeal. I keep a running list of all the books I’ve turned every page of going in my notes app. My TBR occupies half a dozen shelves and stretches into entire catalogs. Reading is not just what we do, it’s who we are. But we all read in different ways, for different reasons, and though I could never read just one book a year, I insist there’s as much value in that kind of sustained attention as there is in the mass consumption encouraged online, maybe even more.

As I reflect on last year’s reading, I’m reminded that books have been a constant presence in my life since before I can remember. I’ve read widely with and without discernment. I’ve read to learn, to escape, to stave off boredom, to seek connection, to drive my work, and to understand others, history, and the present moment. Reading and writing is how I engage the world. I could tell inquiring parents at the winter carnival that I read 93 books last year, and it might make their eyes go wide in wonder and admiration, but really it just makes them bored. Hey, get a load of Mister Reading Books over here!

If I ever had the nerve to say more than, Yeah, I guess I do read a lot, I’d tell them about floating in the pool bored and titillated by C. J. Tudor’s The Chalk Man, which I only read because it was in the little library at the pool on a day I needed a book. I’d tell them about sitting in wonder with my kids while reading The Hobbit and trying my best to answer their endless questions about Bilbo’s meandering adventure in Middle Earth. I’d tell them Nadine Strossen’s Defending Pornography and its bracing, clear-eyed view of the First Amendment and its essential place at the heart of our society. I’d tell them about spending ten months dragging myself through The Man Without Qualities and how terribly unpleasant it is to hear the resonance of pre-WWII Austria-Hungary barking at the tail of American empire. I’d tell them that books are a way to live. Happy new year.

My grinch tends to crop up during the holiday season and, like the Seuss character, he most abhors the consumerism of it all. So, I thought, what better way to get into the holiday spirit than to produce a little gift guide for the folks at Georgia Writers. Honestly, though, as I write this, it’s “Giving Tuesday,” and I think it’s 100% in keeping with the community-minded spirit of this column to suggest a few ways that we might get our holiday shopping done while supporting the kind of literary culture we want to see in the world. Vote with your dollar, as the saying goes. So, whether you’re trying to get the latest bestseller to slip under the tree or seeking a writerly novelty for the office gift exchange, here’s a few ideas about where to shop and a couple recommendations on what to buy.

Local Bookstores

We don’t talk enough about our precious local bookstores. Wherever you find yourself in Georgia, chances are you’re within proximity of a lovely little bookshop run by passionate individuals hoping to offer their communities something more than another one-click package delivery shopping experience. Little bookstores are often the lifeblood of the local literary scene. In addition to new and used books, ephemera, and literary bric-a-brac for sale, most host readings, arrange book club meetings, make space for open mics, and otherwise seek ways to connect folks passionate about literature and books. Supporting local bookstores means supporting local literary communities, and you’re likely to discover forgotten treasures in the stacks. Here’s a few recommendations from my area, and I’d love to hear your recommendations, too!

FoxTale Book Shoppe

Located in downtown Woodstock, FoxTale is “female owned and operated and has been in business for over 17 years.” Their “mission is to be your favorite independent bookstore; a welcoming, warm, and beautiful community gathering space for all ages, where we match books to individual readers and encourage expression of ideas in a safe atmosphere.” Upcoming events at FoxTale include book launches for local authors, a photo op and story time with Santa, and the Bubbles & Books Book Club. Shop FoxTale online here.

The Sweet Read

While you’re in Woodstock make it a double feature and check out The Sweet Read. Go for the impressive collection of used books (over 15,000!), stay for the Bakery du Jour where you’ll be “pleasantly surprised with different cakes, cookies, quiche, and other goodies that run the gamut from traditional southern, modern fare, or completely original.” If you’re lucky enough to have a chance, I suggest the lavender cake, it’s delicious.

The Reading Attic

Situated on Marietta Square, The Reading Attic is a mother-daughter operation that offers a “reading haven … brimming with children’s books & support for local authors.” Its three cozy rooms offer a wide variety of new releases, children’s books, and stationery, candles, and other gifts. With book clubs, readings from local authors, and a mid-December “Mural Reveal Party,” there’s always something happening in the Attic.

The Crazy Book Lady

If you’re looking for an old-school book horde overflowing from two strip mall storefronts, look no further than The Crazy Book Lady in Acworth. Packed wall to wall, floor to ceiling with paperbacks, hardcovers, kids’ books, records, CDs, DVDs, and an impressive collection of tabletop and roleplaying games, there’s treasures waiting to be discovered at The Crazy Book Lady. Hosting a number of TTPRG group meet ups and book clubs, the Book Lady also regularly features local authors for readings and book signings. You might be asking yourself, sure, she’s the crazy book lady, but are there any cats in the store? Well, guess what, the Crazy Book Lady rescues black cats and Morty, Mindy, and Tinkerbell live onsite. You can buy, sell, or trade, and overflow stock is used to maintain the five Free Little Libraries the Crazy Book Lady sponsors in Acworth.

Internet Retailers that Don’t Belong to Jeffrey Bezos

This is the obligatory anti-Amazon portion of the gift list. News of the corporation’s toxic effects on the environment, the economy, the publishing industry, and writers and readers is not hard to come by, but the allure of low prices and convenience is strong. So, before scheduling another prime delivery, here’s a few online retailers offering low prices, convenience, and something to believe in.

Bookshop.org

Bookshop.org “works to connect readers with independent booksellers all over the world,” and “every purchase on the site financially supports independent bookstores.” Since it’s founding, the site has raised “more than $40 million for independent bookstores.” The site acts as a hub for local stores to retail online. When you check out, you select a store to support, Bookshop fulfills the order from their distributors and sends a portion of the profits (80%) back to the bookstores. As the Chicago Tribune put it, “Consider [Bookshop.org] the Rebel Alliance standing up to the Amazon Empire.”

Asterism

Asterism Books “is a trade distributor and online bookstore designed, built, and run by independent publishers.” They “aim to bring the best of small-press publishing together in one place, highlighting the quality and diversity of the independent publishing landscape.” Representing 213 independent publishers, Asterism highlights books you’re not likely to find at other retailers, and thanks to their unique structure (which refuses returns on bulk orders) gets money right into the hands of small publishers. Check out their 2025 gift guide here.

eBay

Sure, everyone knows about eBay, but I hadn’t realized until recently that the site includes an enormous inventory of books offered for immediate purchase without auction at remarkably low prices and often with no shipping fee. What’s more, a lot of these accounts are operated by Better World Books, Goodwill, Open Books LTD, and other charity shops, so some portion of purchases go to support a good cause.

Cool Local Presses

If you’re up for a little more exploration, you can bypass the distributors and bookstores altogether and go right to the source: the small presses doing amazing work right here in Georgia. And there are many! Here’s a few that I know are worth checking out:

Ghost Peach Press

Run by GWA Board member Anna Sandy-Elrod and two colleagues, Ghost Peach Press publishes a couple full-length poetry collections every year. Check out Wandering in All Directions of This Earth by Loisa Fenichell and Too Much by Gion Davis.

Kinchafoonee Creek Press

Based in Athens, Kinchafoonee Creek Press specializes in ecstatic poetry and memoir with a twist informed by the Creek Indian name for a tributary of the Flint River, Kinchafoonee (“white bones”). Check out Chris Jansen’s translation Books of Love: The Lost Poems of Juan Ramón Jiménez and The Lost Etheridge: Uncollected Poems of Etheridge Knight edited by Norman Minnick Jr.

Peachtree Press

Peachtree Press was founded in 1977 and became “the first trade publishing house in the South to achieve national recognition and distribution with several books on The New York Times Best-Sellers list.” Today, they specialize “in children’s books, including board books, and middle grade and young adult fiction and nonfiction.”

Quarter Press

Quarter Press “is a one-person operation that began in 2016” that creates “Limited Edition hardcovers to showcase the level of care [their] artists put into their work. Their beautifully designed books include Liz Kay’s poetry collection, The Witch Tells the Story and Makes It True, an award-winning retelling of Hansel and Gretel from the witch’s perspective, and Natalie Padilla Young’s All of This Was Once Under Water, lovely poems with entrancing illustrations.

Unbound Editions Press

Unbound Editions “believes publishing is a serious, important, and powerful act. [They] are guided by a mission to respect and elevate emerging, underappreciated, and marginalized authors, with a strong commitment to advancing LGBTQ+ and BIPOC voices.” Their exquisitely designed magazine Revel joins dozens of poetry collections, memoirs, and short fictions in handsome hardbacks.

Some Book Recommendations

Here’s a few ideas for books to gift:

He Held Radical Light: The Art of Faith, The Faith of Art by Christian Wiman

For those in search of thoughtful meditations on art and lift. The former editor of Poetry magazine considers art, beauty, poetry, and spirit. Moving and profound.

From the description:

“What is it we want when we can’t stop wanting? And how do we make that hunger productive and vital rather than corrosive and destructive? These are the questions that animate Christian Wiman as he delves into the complex interplay of art, faith, death, and memory. Above all, He Held Radical Light is a love letter to poetry, filled with moving, surprising, and sometimes funny encounters with the poets Wiman has known.”

The Natural Order of Things by Donika Kelly

Donika Kelly’s third book of poetry explores life, nature, change, and the power of love. Searching and sometimes confrontational poems full of community, learning, and insight.

From the description:

“Donika Kelly’s poetry is known for its resonant, unflinching confrontations with trauma and inheritance, translated through myth and nature. The Natural Order of Things expands these explorations into a new realm: one defined by joy and connection. It is an ode to companionship with people, animals, and our planet, and reveals the reparative power of intimacy. In poems inventive, playful, and formally nimble, Kelly pays homage to the voices and people she comes from, the songs of her lineage. Other poems follow the early stirrings of love to erotic transcendence with the lover and the self. Throughout, Kelly finds mirror and marvel in nature, art, and precious friendships. Though it once seemed impossible, she realizes a surprising place for herself, a rightness in the larger world.

The Natural Order of Things is a brilliant and moving book, one that reaches toward equilibrium and something like happiness.”

What We See When We Read by Peter Mendelsund

A thrilling meditation on the power of literature through the lens of visualization. Mendelsund considers the ability of imagination to fill in the sketches of language as he skips through literary and art history. A must for readers interested in how reading works.

From the description:

“What do we see when we read? Did Tolstoy really describe Anna Karenina? Did Melville ever really tell us what, exactly, Ishmael looked like? The collection of fragmented images on a page—a graceful ear there, a stray curl, a hat positioned just so—and other clues and signifiers helps us to create an image of a character. But in fact our sense that we know a character intimately has little to do with our ability to concretely picture our beloved—or reviled—literary figures. In this remarkable work of nonfiction, Knopf’s Associate Art Director Peter Mendelsund combines his profession, as an award-winning designer; his first career, as a classically trained pianist; and his first love, literature—he considers himself first and foremost as a reader—into what is sure to be one of the most provocative and unusual investigations into how we understand the act of reading.”

A Psalm for the Wild-Built by Becky Chambers

The cozy is all the rage, and I’m surprised to report that I’m here for it. Chambers cozy novel about a monk and a robot searching for meaning in an idyllic wilderness is the perfect antidote to our anxious times. Philosophical and friendly, a perfect read.

From the description:

“It’s been centuries since the robots of Panga gained self-awareness and laid down their tools; centuries since they wandered, en masse, into the wilderness, never to be seen again; centuries since they faded into myth and urban legend.

One day, the life of a tea monk is upended by the arrival of a robot, there to honor the old promise of checking in. The robot cannot go back until the question of “what do people need?” is answered.

But the answer to that question depends on who you ask, and how.

Becky Chambers’s new series asks: in a world where people have what they want, does having more matter?”

Here by Richard McGuire

Recently released as a film by Robert Zemeckis, Here is a fascinating experiment in graphic story telling. Keeping the camera focused on a single corner of the world, McGuire’s story covers millennia as lives human and otherwise come and go. Beautifully illustrated and surprisingly moving

.

From the description:

“From one of the great comic innovators, the long-awaited fulfillment of a pioneering comic vision: the story of a corner of a room and of the events that have occurred in that space over the course of hundreds of thousands of years.”

Tacky’s Revolt: The Story of an Atlantic Slave War by Vincent Brown

A little-known history of an uprising of enslaved people in Jamaica led by a charismatic leader, tacky, Brown’s history explores the history of rebellion, independence, and colonialism in rich detail.

From the description:

“In the second half of the eighteenth century, as European imperial conflicts extended their domain, warring African factions fed their captives to the transatlantic slave trade while masters struggled to keep their restive slaves under the yoke. In this contentious atmosphere, a movement of enslaved West Africans in Jamaica organized to throw off that yoke by violence. Their uprising--which became known as Tacky’s Revolt--featured a style of fighting increasingly familiar today: scattered militias opposing great powers, with fighters hard to distinguish from noncombatants. Even after it was put down, the insurgency rumbled throughout the British Empire at a time when slavery seemed the dependable bedrock of its dominion. That certitude would never be the same, nor would the views of black lives, which came to inspire both more fear and more sympathy than before.

Tracing the roots, routes, and reverberations of this event, Tacky’s Revolt expands our understanding of the relationship between European, African, and American history as it speaks to our understanding of wars of terror today.”

Some Writerly Gift Recommendations

LAMY Safari Fountain Pen

I’m a huge fan of fountain pens. They’re neat, smooth writing, and more durable than most disposables. The LAMY Safari is a perfect starter pen for everyday use. It can be refilled manually or using a cartridge, which saves on the cleanup, and its lovely design feels great in the hand.

Handmade Notebooks

Lots of letterpress studios and paper makers offer surprisingly reasonable handmade notebooks produced with traditional methods. Check out Blackbird Letterpress, Pike St. Press, and Egg Press for notebooks and other letter-pressed goodies.

Literary Games

We’re living through a gaming renaissance the likes of which we’ve never seen. There are more board and card and roleplaying games than ever before, and many of them are great gifts for the writers and readers in your lives. Check out Dialect: A Game About Language and How It Dies, a game where players invent a language; Once Upon a Time, a collaborative storytelling game based on classic fairy tales; or Tether, a two-player roleplaying game “that is played through the process of writing journal entries.”

Gift Subscriptions

Help break that screen addiction and deliver some great literature by signing the writers and readers in your family or friend group up to receive physical copies of your favorite literary magazine. ZYZZYVA, Poetry, The New York Review of Books, The Paris Review, Revel, and many others offer reasonable subscription rates. Or, if you’re feeling generous, sign your loved ones up for a press subscription to receive a whole print run next year. Check out Sublunary Editions, Deep Vellum, and Graywolf.

Personalized Book Embosser

Help your loved ones “transform their book collections into a personal library” with a personalized book embosser. Choose from dozens of seals and add custom text. From the library of a book lover!

Traditional Letter Sealing Kit

Nostalgic Impressions offers a wide variety of kits, starter sets, and bundles for bookish folks interested in engraving, stamping, and otherwise decorating their correspondence. Waxes, glues, stamps, and seals are all “fully mailable and flexible and ready” to deliver lovely thank you cards for your thoughtfulness.

Happy Holidays!

I hope you found some interesting ideas here, and I’m eager to hear your suggestions in the comments, the more local the better! Whatever you celebrate, even nothing at all, I’m wishing you the best for a happy holiday season and a lovely new year!

As I’ve written here, I’m also the Editorial Director of The Headlight Review, a literary magazine produced in the MAPW program at Kennesaw State University. In the last couple years, we’ve been working hard to expand the magazine’s presence. We started a new section for feature writing, called the “High Beams,” added a new short fiction prize dedicated to Anthony Grooms, and we published a collection of the “Greatest Hits” from our first five years. Recently, we’ve been publishing micro-fictions and planning for next year’s special double issue of “New Southern Writing.”

A while back, a representative from EBSCO approached Headlight to ask if we’d be interested in signing up for a new database they’re building to package literary magazines for libraries. The “literary research database” offered to help grow our readership at no cost by scraping publications off our website and ingesting and then offering that database alongside a litany of other databases EBSCO sells to libraries in lucrative subscription contracts. At first, we were intrigued.

According to their flyer, which describes EBSCO as “The Natural Partner” thanks to their history servicing subscription magazines, the company “offers publishers a variety of ways to gain exposure for their publications in order to maintain and grow their core business.” Promising access to “more than 115,000 libraries worldwide,” “high usage of publications,” and “increase[d] web traffic,” it sounded like a great marketing opportunity, so we took a meeting with the representative, and they sent over the contract that structures the arrangement.

EBSCO, an acronym for the Elton Bryson Stephens Company, was founded in the forties by Elton and his wife, Alys, in Birmingham, Alabama. At first, the pair sold “magazines, personalized binders and racks to the U.S. Armed Forces under the name of Military Service Company.” The company went on to acquire a wide variety of enterprises related to the production and distribution of magazines, including “Metal Fabricators and Finishers” and dozens of magazines, presses, and distributors. Today, EBSCO “is one of the largest privately held companies in Alabama” and one of Forbes’s “top 200 [businesses] in the U.S., based on revenues and employee numbers.” Needless to say, a company of this size and hunger doesn’t offer promotional partnerships out of the goodness of their heart.

Now, The Headlight Review is an open-access, online publication. We don’t use paywalls or charge subscription fees, and the rights for everything we publish return to their authors upon publication. Reviewing our publication contract, we likely have the right to license the work we’ve published to EBSCO for use in this way, but ethically speaking we’d need to clear inclusion in the database with previously published authors and modify our contract with future authors to explain that THR would be a waypoint on the path to database distribution. The appeal of more attention notwithstanding, I wonder how most of our writers would feel about becoming EBSCO content.

From our perspective, it seemed like an awkward deal. Although we don’t charge for content, it struck us as a bit strange that EBSCO would be monetizing our work through library subscriptions without offering us anything but vague exposure. Further, the “system” they promised to use to obtain content from our website amounted to someone selecting “save as PDF” and uploading whatever the results, which often includes formatting errors and omissions. In other words, if we wanted high-quality representations of our work in the database, we’d need to reconstruct our workflow to build it ourselves. Doing yet more free labor on behalf of EBSCO. And, when it comes to exposure, the promise of more clicks through to our website, potential visits to our store, etc., evaporated when we learned that the database would simply include complete versions of everything we published. Leaving database users no reason to interact with THR beyond whatever piece they might find in the database.

We decided to pass on the deal, and I’m glad we did. A couple weeks ago, members of the Community of Literary Magazines and Small Presses began reporting they’d received an invitation from EBSCO to “engage with the growing field of generative AI in a secure and profitable way.” Apparently, EBSCO has

recently been approached by several GenAI companies seeking high-quality, licensed content to train their large language models (LLMs)—creating a new revenue stream for publishers without compromising control or risking misuse.

According to the message, journals’ “existing license agreement” with EBSCO means the content included in their databases qualifies for this program, which offers $0.0004 per word for non-Open Access journals and $0.0001 per word for Open Access journals should their content be selected to train an LLM at one of the “several GenAI companies.”

In a follow-up message, the EBSCO rep clarified that this was an optional program that database partners could choose to forgo. She also explained that

Our agreements with LLM providers strictly allow:

· Use of content for the training model

· Answering questions (considered “transformative work” when addressing derivative use)

· No display of book text

· No citations, to prevent content cannibalization

They are simply “providing a service to those who are interested and who want to work with us to supplement their own activities or, if they aren’t pursuing these opportunities directly … to still be able to benefit.” As Anna Leahy, director of Tabula Poetica, pointed out in thread, the potential upside here for open access journals like The Headlight Review is that LLM companies are already scraping the open internet and pillaging the work we publish anyway. At least this way, we can get some recompense: 15 cents for a 1500 word story. If Moby Dick were an open access text, it would earn $20.91 from this arrangement.

And this, I think, is the real rub. While EBSCO and Anthropic and Wiley and Open AI rake in tens of billions of dollars in corporate contracts, venture capital, and—most offensively—public library subscription fees, the creators and the laborers who produce the work they consume and capitalize earn fractions of a cent for their labors. Given EBSCO’s business model, prices are tough to come by, but in 2019, they offered Montana State a five year subscription to their database EBSCO eBooks for $57,810/year with a 2%/year renewal increase cap per year. By year five, Montana could have paid almost $300,000 (in state funds) to access EBSCO eBooks. How much of a cut did each author and publisher get? $0.0001 per word?

For Headlight to participate in the partnership program would have cost us another 50-75 hours of labor per issue in addition to whatever management it would take to secure author permissions and share the meagre spoils of licensing their work. Carry that across the 360 journals included in the “Literary Reference” database, and we’re already into the tens of thousands of hours of additional unpaid labor. And we haven’t even considered compensating the writers for the time it takes to produce the work in question.

Somewhere along the way, the founding principle of copyright, that the individual should maintain a monopoly on the fruits of their labor to sustain their endeavors, became perverted to serve corporate interests. The supposed beneficence of corporations like EBSCO serves to maintain the illusion that individual works of creativity, writing, scholarship, etc., have no value outside of the systems they’ve built; that artists only deserve compensation in the form of attention, recognition, and credibility; what’s valuable is the work of technicians and intermediaries who install the work in the network of extraction.

This bleak future promises to get darker now that those mechanisms are wedded to generative learning models, which threaten to cut out the writer all together. Whatever these databases are doing, it isn’t building a sustainable future for the literary arts, and that should give writers and publishers pause.

A couple weeks ago the scent of garlic wafted over my social media feed. It wouldn’t have seemed out of place on YouTube, where I watch a lot of cooking videos, but this was BlueSky, where I follow only poets and publishers, and rarely post about anything other than books. So, that pungent stench stuck out against the bad news, agent pitches, book hauls, and WIP excerpts that usually dominate the feed.

The original poster, Emily C. Hughes, seemed even more shocked than I was to discover that the fragrant aroma of Allium sativum has begun to emanate from the publishing industry. She writes, “no no the publishing industry is doing fine, why do you ask,” sharing a clip of an announcement that social-media-savvy horror-author Jennifer L. Armentrout has partnered with Hellmann’s mayonnaise to promote her new book, The Primal of Blood and Bone. According to the press release,

In a promotion designed for BookTok enjoyment, Jennifer L. Armentrout’s just-published THE PRIMAL OF BLOOD AND BONE is available in a special, limited-edition “garlic-scented copy, infused with Hellman’s Garlic Aioli to create a one-of-a-kind, Craven-proof book.”

Emily C. Hughes BlueSky post playing up the desperation of brand partnerships

The new book apparently deals with a race of creatures, the Craven, who, like the mythological vampire, are put off by garlic. To promote the book, Armentrout’s publisher worked with Hellmann’s and Blue Box Press to produce a garlic scented ink that was used to print the book. So, copies of the book themselves reek of garlic, protecting the reader from the creatures within.

For their part, Hellmann’s compiled a number of recipes tapping themes from the book, such as “Garlic Sliders of the Ascended,” two “Poppy’s Platters,” and a “Casteel-erole.” “Caution,” the mayonnaise brand writes, “while Hellmann’s Garlic Aioli may not actually protect you against cravens, it will appease your cravings.” In a hilarious twist, one of the first comments on Armentrout’s Instragram reel promoting the special edition reads, “i honestly don’t remember garlic in any of the books?” But it’s there now. And there’s big bulbs and cloves on the cover to drive the point home.

The Craven aioli cover of The Primal of Blood and Bone

In some ways, I agree with Hughes that Armentrout’s Hellmann’s partnership is the peak of desperation. You can see it on the author’s face as she sits beside a plate of fries waiting to be dunked in a little silver tankard of Hellmann’s Craven Proof Aioli to record the reel. “Everyone knows that garlic is the only way to prevent a craven attack,” she says, “So Hellmann’s and I have made a garlic scented book.” Armentrout’s tone of sublime resignation to capital is as pungent as the promised garlic scent of the book.

Colleen Hoover poses with a Target display pairing her novels with Olive & June’s nail care products.

Armentrout is hardly the first author to pair with an unconventional brand in a bid for publicity. The rock band KISS famously mixed their blood into red ink for a 1977 Marvel Comic. To promote her pastel colored novels of domestic abuse, Colleen Hoover teamed up with Olive & June, a “nail care brand” to create a line of nail polishes paired with her books. For the release of Outlander: Go Tell the Bees That I’m Gone, Diana Gabaldon’s publisher, Random House, teamed up with American Airlines to offer one fan of the Scottish-set-fantasy series a trip to Scotland. It seems undeniable that these partnerships have increased in the era of TikTok, and surely, they’ll continue to proliferate as publishers and authors look for creative new ways to market their books.

Maybe it’s all harmless fun. It certainly worked on me. I’m only slightly embarrassed to confess that the whiff of garlic scented pages is soon to haunt the bookshelf where I house my collection of vampire fiction. But only to use as an example when I teach. Really. And I’m dying to see the sales reports from both Miss Willas Bookshop (purveyors of Armentrout’s special edition) and Hellmann’s to see if the BookTok sensation can really help move tubes of Aioli. Maybe it’s best to think of these kinds of partnerships as ancillary to the work of literature, as they certainly are. But, as David Barnett points out in The Guardian, Armentrout hardly needs this kind of promotional energy. She’s a multimillionaire author with a huge social media platform and more than a few books on bestseller lists across the world.

Barnett argues that he’d be more comfortable with this sort of promotional activity if he knew that the big presses were giving the same treatment to their less well-known authors, which they certainly aren’t. Obviously, it wouldn’t be sustainable or effective for every author to have a promotional campaign like this, but even beyond that this kind of marketing seems bad for literature even if it’s good for books. Maybe I’m old fashioned, but I think literature is about something more than moving units. I bought the book to share as a novelty in the history of publishing, but am I going to sit down and push through the nauseating stench to read the story inscribed in garlic scented ink? Probably not. And that’s just the problem.

These kinds of promotions reduce literature to commodity and emphasize its manifestation as a product. Sprayed edges, special edition and variant covers, bonus sections, and product partnerships are about the sale of things, not the experience of literature. They’re effective marketing techniques used to get people to buy books, but they can only bode poorly for the quality of writing and reading in our already aliterate culture. To answer potential commentators, special editions like The Primal of Blood and Bone will someday have to be sure to include references to their sponsors, which publishing has also tried before, and we’ll all be worse for it. Of course, I’ll take a big whiff when the garlic book arrives, but Armentrout’s legacy will always carry the stink.

Copyright has been a central question in AI discussions since before the release of ChatGPT in November of 2022. Visual artists were, perhaps, the first to complain that developers were stealing their work to train these systems. Although image generators date back to the 1960s, the public release of DALL-E made it clear how essential existing works are to the training of models capable of generating images, and artists sounded the alarm when it became clear that these systems produced images uncannily close to their protected work. Concern over image generation gained enough traction that, in 2022, Getty Images even banned AI generated content over copyright concerns.

Of course, Getty now has its own “commercially viable” image generator, which it claims to have trained ethically using only approved works. In the meantime, OpenAI released ChatGPT, and it became clear to authors that tech companies were also consuming their work, in this case to train large language models. And consumed they were. It takes an enormous amount of data to train and refine a predictive algorithm because the system works by creating an average of its inputs that it then uses to generate an output in response to a prompt in natural language. For an accurate response, the system needs to consider as many potential situations as possible.

For example, imagine reading four articles about toads. In each article, the phrase “endangered spotted toads” appears. Given those repetitions, it’s probably safe to assume that a fifth article about toads will include the phrase “endangered spotted toads,” and an LLM would use information like that to reproduce the phrase in response to a prompt about toads. Of course, the real system is more complex than that, and human language infinitely more variable, so these systems require a huge amount of data to model how language works and create the accurate illusion that they’re capable of generating original language.

That’s where the internet and published books come in. OpenAI and other developers have not been entirely forthcoming about the sources of the data they use to train their models. The data underlying more recent models, like GPT 4.0, are especially safeguarded. We know from a paper on training the model that the GPT 3.0 dataset began with 45 terabytes, which researchers filtered down to 570 gigabytes. For context, the Project Gutenberg edition of Moby Dick (a relatively large novel) clocks in at 1.4 megabytes. A terabyte is a million megabytes, so the initial dataset could, conceivably, hold just over 32 million copies of Moby Dick.

Read more

Although it’s been some time since I wrote about censorship in this column, the battle over what we can see and read continues. The furor of the last few years has settled to some extent because the censorship battles have, in many cases, been lost. Libraries, bookstores, and soon the internet continue to fall prey to the designs of censors under auspices of “protecting children.” This month, I’m considering how these arguments have developed, especially in light of recent pushes to prohibit DEI and so-called gender ideology at the federal and state levels. Censorship affects us all, but it is especially noxious to writers who seek to express ourselves in increasingly policed spaces.

Before we get to the bad news, though, let’s linger a moment on a couple of recent wins for Free Speech fans. For a while, things were looking bleak in St. Francis, Minnesota. The school district voted to use Moms for Liberty’s “Book Looks” ranking system and reviews website to make acquisition decisions about what to purchase and hold in their library. Book Looks promised to help “find out what objectionable content may be in your child’s book before they do,” and their rank system concerns itself violence and sexual content.

Really, though, Book Looks was a partisan tool used to attack books deemed oppositional, especially books by people of color and LGBTQ+ people, and many of their negative ratings traced back to the simple presence of said people or their history as opposed to anything obscene. Their work was premised on the assumption that public school teachers and librarians seek to indoctrinate children, especially with sexually explicit material. To help make that case, they often rated books that have no place and never even appeared in public school libraries, such as Closer by Dennis Cooper.

In St. Francis, the district decided to base acquisition and holding decisions in their libraries exclusively on the Book Looks ratings, condemning any book rated at a 3 or above and barring it from the school. Books rated out included Toni Morrison’s Beloved (3), Margaret Atwood’s The Handmaid’s Tale (4), and Kurt Vonnegut’s Slaughterhouse-Five (4). The ACLU and the teachers Union sued, and Minnesota passed a law prohibiting book banning. As a result, the St. Francis district abandoned its plans to rely on Book Looks ratings and will now make library policy in consultation with librarians, parents, and community members.

Not only that, but Book Looks has now official folded. According to a note on their website, God called them to their Book Looks work, and “it has become apparent that His work for us here is complete and the He has other callings for us.” As of March, they “will be ceasing operation and taking down all [their] reports from the site.” Mission accomplished, they claim. While they failed in St. Francis, the culture of school and library censorship has taken hold and begun to spread. In many ways, what the Book Looks folks say is true, especially here in Georgia, where dozens of books have been removed from school libraries. Now, censors have scored another win in the battle to regulate school libraries.

Library Laws Prohibit Free Speech

The assault on school and public libraries has been largely successful in states like Florida and Georgia. Now, according to the Fifth Circuit Court of Appeals, schools in three more states (Louisiana, Mississippi, and Texas) may remove and prohibit books for any reason. Contrary to the Supreme Court’s plurality decision in a case called Island Trees School District v. Pico, which ruled books cannot be removed from public school libraries for reasons of ideological difference, the appeals court asserts that books held in a government library are government speech and therefore not protected by the First Amendment.

In other words, where it was previously illegal to exclude a book from the public-school library simply because an administrator or community member (or whomever) didn’t like what it depicted or simply disagreed with its ideas, libraries in three states may now discriminate at will. If books in public school libraries constitute government speech, rather than protected individual speech, then the contents of those libraries exist entirely at the discretion of whoever has the power to administer the library and thus approve the “speech” contained therein.

This truly dangerous precedent imagines the library not as a free bastion of knowledge available to all citizens, but as a propaganda arm of the government, a bullhorn for approved speech that will, no doubt, only be used to support the agenda of whoever holds power. We might argue about the extent to which that’s how public libraries have always functioned in practice, but the legal precedent since Pico has encouraged free expression and empowered communities to make these decisions with an eye toward liberty.

Although the judge’s ruling cracks glib jokes about the severity of this new interpretation of the functions and policies of public libraries, this ruling proposes a major change in the fundamental idea that Free Speech, books, and knowledge are a public good that should be supported by taxpayers. Instead, taxpayers in those three states, no matter their allegiance, are now supporting only those ideas endorsed by whoever administers the library, limiting that institution’s ability to freely transmit information and endangering any book that doesn’t toe the line. And if it works in the library, censors hope, it will also work online.

Proposed Laws to Age Restrict the Internet

Our most powerful information technology faces similar challenges. A raft of laws have been proposed, passed, and challenged in a number of states that would require age verification to access certain kinds of content. Usually promoted as anti-porn laws, bills like Oklahoma’s SB 1959, Texas’s HB 1181, and Georgia’s SB351 require certain websites, such as PornHub, and all social media platforms to verify the age of users seeking to access their content. We must, the authors and supporters of these bills claim, implement restrictions online to protect children from accessing sexually explicit content.

And the Supreme Court agrees. In June, the court issued its decision in Free Speech Coalition, Inc. v. Paxton, which challenged Texas’s age verification law. Ruling on behalf of the Texas legislature to uphold the law, Justice Thomas writes, “The power to require age verification is within a State’s authority to prevent children from accessing sexually explicit content. The burdens of age verification, Thomas argues, “[have] only an incidental effect on protected speech.” All of this is well and good, if we, like Thomas, believe these laws are only meant to “prevent children from accessing speech that is obscene to them.”

But what constitutes obscene speech to children is not as clear cut as the court would make it seem, and these restrictions have already led to chilling effects and worrying instances of outright censorship. Like other anti-porn laws before them, these laws ultimately don’t restrict themselves to explicit material. For example, Meta, the parent company of Facebook and Instagram, has hidden content with LGBTQ+ hashtags in response to age verification requirements. For months, Meta classed tags like “gay,” “trans,” and “queer” as “sensitive content” thus deeming any content associated with those tags age-inappropriate for users under 18. Creating and prohibiting speech about “sexually explicit identities” is a step forward in discriminating against people who hold those identities and eroding their legal protections.

Although it was defeated in the House in 2024, a federal version of an internet age verification law has recently returned to congress for review. The Kids Online Safety Act (KOSA) would, among other things, require platforms like YouTube and X to verify the age of their users, default to the highest possible privacy settings, and produce tools for parents to monitor their children’s internet use. As YouTube is owned by Google, chances are good that Google will simply age restrict all its operations, requiring users to upload sensitive personal data to access information and services, which will not always be pornographic or sexually explicit in nature.

Aside from the obvious trouble with creating a database of internet users keyed to their legal documents, such as driver’s licenses, which is then held by a private company like Palantir, censoring content politicians find objectionable is the true motivation of these laws. Age verification seems like a reasonable idea in theory. Of course no one wants children accessing hardcore pornography, but the bill’s cosponsor Marsha Blackburn confirmed critics’ skepticism when she described the bills’ agenda as “protecting minor children from the transgender [sic] in this culture and that influence.” Promulgating the social contagion theory, Blackburn said the internet and social media platforms need to be regulated because that’s “where children are being indoctrinated.”

As Nadine Strossen, former president of the ACLU, pointed out during the 1990s debates over pornography, porn is often a wedge used to erode Free Speech in the name of protecting women and children, but it’s never the whole story. When certain ideas, like those associated with LGBTQ+ identities, women’s rights, sex education, and race, are deemed “explicit,” “pornographic,” or “obscene,” it becomes possible to restrict speech about them and more easy to persecute people.

Preventing Woke AI

Finally, for all my skepticism about large language models and so-called artificial intelligence, I maintain that models we build ought to accord with our constitution’s commitment to Free Speech. Programmers should face no more scrutiny from the federal government than other actors (such as authors or publishers), and the work they produce should be free of burdensome government requirements to produce officially endorsed speech.

That will no longer be the case for those programmers hoping to do AI business with the federal government. On July 23, the president signed an executive order titled “Preventing Woke AI in the Federal government.” The order claims that AI

will play a critical role in how Americans of all ages learn new skills, consume information, and navigate their daily lives. Americans will require reliable outputs from AI, but when ideological biases or social agendas are built into AI models, they can distort the quality and accuracy of the output.

Part of the administration’s DEI purge, this order explains that

One of the most pervasive and destructive of these ideologies is so-called “diversity, equity, and inclusion” (DEI). In the AI context, DEI includes the suppression or distortion of factual information about race or sex; manipulation of racial or sexual representation in model outputs; incorporation of concepts like critical race theory, transgenderism, unconscious bias, intersectionality, and systemic racism; and discrimination on the basis of race or sex. DEI displaces the commitment to truth in favor of preferred outcomes and, as recent history illustrates, poses an existential threat to reliable AI.

Never mind that studies have repeatedly shown that current AI models reinforce racial stereotypes, perpetuate gender and racial bias, and make racist decisions about people based on their dialects. Now, to do business with the government, programmers will have to make sure their models never acknowledge “gender ideology,” “systemic racism,” or “climate change.” Apparently, though, it’s fine that they go on Hitler praising antisemitic rants as Elon Musk’s Grok did only five days before the company announced its deal to provide “Grok for Government.”

AI is a new frontier in the censorship wars, but in many ways the arguments, desires, and dangers remain the same. Those in power wish to suppress the speech of their opposition, and they count on our agreement that the most extreme cases justify restrictions that then branch out to cover all manner of speech. To quote Justice Brandeis, the Constitution’s framers

knew that order cannot be secured merely through fear of punishment for its infraction; that it is hazardous to discourage thought, hope and imagination; that fear breeds repression; that repression breeds hate; that hate menaces stable government; that the path of safety lies in the opportunity to discuss freely supposed grievances and proposed remedies; and that the fitting remedy for evil counsels is good ones.

Kurt Edward Milberger is Assistant Professor at Kennesaw State University. He teaches publishing, professional writing, and literature courses with an emphasis on book history and material culture. He's currently researching a book on censorship and preparing an edition of Izaak Walton's The Compleat Angler.

Like other traditional industries, publishing struggles to adapt and change. Everyone agrees that the system is broken, they just don’t agree on exactly how it’s broken or what can be done to fix it. Consequently, like other traditional industries, publishing seems ripe for disruption by eager upstarts with bold ideas for reinventing the industry. And, like eager upstarts in other industries, bold disruptions often seem to do more harm than good.

Last November, I wrote about the emergence of 8th Note press, TikTok’s publishing venture, which promised to harness the power of social media virality to drive the book market. An imprint of Zando, 8th Note claimed to make “waves in the publishing industry with a focus on contemporary fiction that resonates with Millennial and Gen Z readers.” They hoped to turn the publishing model on its head by publishing books that fed current trends on social media, driving rather than being driven by the algorithm.

They started that process by acquiring self-published books and transforming them into paperbacks, and then they began acquiring new books from debut authors. According to Alexandra Alter, “the plan was to release 10 to 15 titles a year, with a focus on romance, romantasy and young adult fiction. Later, the imprint indicated to agents that it was expanding into science fiction and fantasy.”

Harnessing the power of TikTok to market their books seemed like a great idea to at least a couple dozen authors who eagerly agreed to the company’s deal, which “included a social media marketing campaign, royalties and an advance of $3,500 per book.” But, just last month, before the press could make much of impression, ByteDance (TikTok’s owner) shut it down. Shuttering their websites after informing authors in May of the press’s disbanding, 8th Note has begun “returning publication rights to authors” and “taking down digital editions of their books, effectively unpublishing them.”

According to NetGally, 8th Note released about thirty titles in the two short years of its run. None of them viral hits, few of them receiving much marketing support for their small print runs, and all of them, now, effectively unpublishable as few traditional houses are going to take a shot of republishing an already published novel that didn’t score a hit for the would-be-king-making publisher. What’s worse, many of these first-time authors are still waiting for their sales statements and royalty payments from 8th Note.

If another defunct nontraditional publisher is any example, they’ll end up waiting a long time. Formed in 2010, Unbound publishing was supposed to “help authors avoid the whims of sales-obsessed publishers by pitching their ideas directly to the people who really matter: the readers” (Duerden). Basically “Kickstarter but for Books,” Unbound set up a website where authors would pitch their books, often in the form of book trailers and other promotional material, to readers who would then pledge to support the book. For successfully funded campaigns, Unbound would use the money to offer a negligible advance, produce and distribute the books (Faber & Faber agreed to distribute some Unbound titles), and split the royalties from future sales fifty-fifty with the author.

Back when it was announced in 2011, The Independent described Unbound as “a small but key revolution in the world of book publishing,” and more than a few big names threw their support behind the idea, including Kate Mosse and Philip Pullman, who called it “an idea whose time has come.” Earlier this year, that idea’s time had already passed, and Unbound went into administration, the UK’s equivalent of bankruptcy.

In the process, the company was sold to Boundless, a new corporation organized by the same folks with the same big-tech promises now fueled by AI, and dozens of Unbound authors revealed they were owed unpaid royalties for books they’d published with the press. In the most egregious example, Tom Cox reports he’s owed a “five-figure-sum” for royalties on the seven books he published with Unbound stretching back to May 2024.

According to the press, the bankruptcy and restructuring cleared their debts and “all monies owed to authors by Unbound will be honoured by Boundless IP Limited.” But authors are right to be skeptical of a press that was supposed to hold their share of royalties in trust to pay out at regular intervals but instead blew them on operating, development, and other costs.

Both of these would-be publishers learned the fundamental lesson of the publishing industry the hard way: Success is unpredictable, random even, as former Random House CEO Markus Dohle infamously put it. Sometimes you have a viral hit like Fifty Shades of Grey that allows the press to pay out $5000 to every employee. Sometimes your eagerly awaited new novel gets cancelled before it even comes off the press.

Traditional publishing has developed a number of semi-reliable strategies to help mitigate the risk. They invest in celebrity and political memoirs, which are sure to return their investment. They rely on established, bestselling authors to carry their lists. And they conglomerate their competitors. Indie presses, too, design for these difficulties. Many, like Greywolf Press and Deep Vellum, rely on nonprofit and government grants to subsidize their earnings. Others make publishing part of a portfolio of activities that include things like selling merch and other services (often publishing services, such as editing and typesetting for self-published authors). Maybe, too, it’s important to remember that 8th Note and Unbound (especially) seem to have failed as much due to poor management as to any fundamental misunderstanding of the way disruptive tactics might help shake up the publishing industry.

Since the advent of digital self-publishing, it has become increasingly clear that publishers, like record companies before them, operate an often-predatory scheme to extract value from authors. As services like development, editing, publication, and marketing continue to become more accessible outside of traditional structures, the failure of these disruptors demonstrates that it’s not enough for a publisher to simply provide the services of publishing. Instead, any revolution in publishing must begin with a commitment to the workers who make the industry possible. From authors to their agents, copyeditors, publishers, and publicists, sustaining people ought to be the priority of every publisher.

For a long time, I delighted in looking up my favorite books on Amazon and reading the one-star reviews. Classics, bestsellers, esoteric indies, no matter the genre, success, or merit of any given book, haters condemned the text, abused the author, and warned readers not to waste their precious time or money in one-star reviews.

In the mood for “mildly clever contorted witticisms written in stilted language?” Check out Pride & Prejudice by Jane Austen. The Great Gatsby? “One of the worst books ever written. People living on eggs worried about money and unrequited love. Blargh.” Beloved? “Bad plot, bad writing, premises that ring with falseness, rather ridiculous.” Moby Dick? “Moby Don’t.” Of course, these are not serious people, but there’s something refreshing about the unbridled opining of the masses. “This novel actually won a Pulitzer Prize. For the life of me I don’t know why.”

None of us wants to be judged by the worst of what others think of us. It’s especially understandable that authors have a lot to say about bad reviews. It takes time and soul to write a book, hard work and real courage to put it out into the world. It must be crushing for the culmination of that work to be stamped with one star and a dismissive review: “Pretty much every page is bad.” But discussion of reviews has taken a curious turn in the last couple years as a segment of literary culture, including some prominent authors, has begun to argue that every star rating ought to be a five-star rating.

Read more

Earlier this month, the president summarily dismissed Shira Perlmutter, the Registrar of Copyrights at the Library of Congress. Perlmutter was the second congressionally appointed official removed from the LoC in as many days, following the firing of Dr. Carla Hayden from her position as Librarian of Congress for “quite concerning things that she has done … in pursuit of DEI and putting inappropriate books in the library for children.” Never mind that children aren’t allowed to access the Library of Congress or that part of the library’s function is to collect copyrighted works published in the United States.

Perlmutter’s firing came shortly after her office released the third part of its report on Copyright and Artificial Intelligence, which considers “Generative AI Training.” Many have speculated the firing resulted from the report’s conclusions. Tech moguls, particularly Elon Musk, seemed to have expected the office to endorse their view of intellectual property and loosen the reins of fair use. They hoped the report would encourage the courts to allow them unrestricted access to copyrighted material for training their large language models. When it became clear the office was going to assert that much generative AI training “goes beyond the established fair use boundaries,” Perlmutter had to go. At least, so goes the speculation.

In the wake of this report, I thought I’d use this month’s column to consider two of the main issues raised by this trio of reports for writers, publishers, and readers: How might LLMs affect the publishing ecosystem in terms of what can be copyrighted? And, how do those systems affect what it means to be protected by copyright, especially in an environment where “information wants to be free”? One thing is certain, LLMs are having an enormous impact on how we conceive and practice copyright law, both in the US and around the world. And that impact gets at some of the fundamental questions of authorship, ownership, and publication.

Generative AI and Fair Use

A quintessentially democratic spirit informs US copyright law. Designed to provide a short-term monopoly for the individual inventor, artist, or creator, the ideal of copyright develops into a long-term grant to the public. In other words, recognizing that everything is made of something else, copyright was supposed to let individuals profit from their contributions before those contributions become the property of the public.

As Benjamin Franklin wrote, “as we enjoy great advantages from the inventions of others, we should be glad of an opportunity to serve others by any invention of ours; and this we should do freely and generously.” Franklin and the framers of copyright law recognized the individual’s need to profit from their work—to sustain their livelihood and continue their contributions—but they imagined a society where each should ultimately have access to the work of others to do with what they pleased. In this way, knowledge would spread, invention would increase, and society would progress.

The first two parts of the Copyright Office’s AI report deal with “Digital Replicas” and “Copyrightability,” both important issues, especially for writers and artists, but the tech industry most eagerly awaited this third part of the report, which considers the question of using data (read text, images, artwork, etc.) protected by copyright to train AI systems. At this point, no one denies that the unhindered use of such material is essential to the development of these systems. So essential, in fact, that several executives have admitted that their products could not exist without the wholesale exploitation of works protected by copyright. However, developers and their corporate overlords assert that the use of such data to train AI systems falls under the protection of fair use.

Fair use doctrine makes a special carveout in US copyright law that allows artists and writers to use protected content provided that use meets certain criteria that courts have decided do not infringe on the rights of copyright holders. It’s a way of hastening a work’s arrival in the public domain by recognizing that certain kinds of use do not actually affect the owner’s original monopoly on profits from their work. It’s fair use that allows YouTube video essayists to post clips from Disney movies, academic critics to reproduce quotations from contemporary novels for criticism, and a host of other activities. To determine whether a work is fair use, the courts rely on a four-fold analysis that considers: 1) purpose of the use, 2) the nature of the copyrighted work, 3) how much of the work has been used, and 4) the effect of the use on the market for the original.

There is no magic formula to determine whether a given work might be fair use. One cannot quote up to a certain number of lines from a hit song, nor may they be confident that their critical analysis using clips from a Star Wars show won’t be found in violation. Instead, courts use these factors to consider copyright issues on a case-by-case basis. What’s more, the Copyright Office’s report on these questions isn’t a policy document, nor does it have any legal power. Instead, it seeks to “[c]onduct studies and [a]dvise Congress on national and international issues relating to copyright.” The report amounts to a series of recommendations for how legislators and jurists might understand new developments in copyright from the office’s perspective and help guide them in proposing new legislation or litigating copyright questions in court.

From the tech perspective, training LLMs with copyrighted work amounts to fair use. That was the case META’s lawyers made when it was discovered that Llama’s developers had knowingly used LibGen, a pirated dataset containing millions of texts, including nonfiction and fiction books, scholarly journal articles, comics, and magazines. Lawyers for the Author’s Guild assert that Mark Zuckerberg himself “approved Meta’s use of the LibGen” despite knowing the data to have been pirated. According to Reuters, “A Meta spokesperson said that fair use is ‘vital’ to its ‘transformational GenAI open source LLMs.’” At the core of their arguments is the assertion that training an LLM doesn’t interfere with the owner’s monopoly because the system does not reproduce the books it ingests. Instead, its output transforms its training data into a new work. Except when it doesn’t.

Acknowledging the complexity of the four-factor analysis, the Copyright Office concludes that

some uses of copyrighted works for generative AI training will qualify as fair use, and some will not. On one end of the spectrum, uses for purposes of noncommercial research or analysis that do not enable portions of the work to be reproduced in the outputs are likely to be fair. One the other end, the copying of expressive works from pirate sources in order to generate unrestricted content that competes in the marketplace, when licensing is reasonably available, is unlikely to be fair use. Many uses, however, will fall somewhere in between. (74)

At stake in these arguments, for tech developers, is the supposed ability to “power incredible innovation, productivity, and creativity,” but writers worry that these technologies will use their works to outcompete them in the marketplace. Is it in keeping with the original spirit of copyright to transform every novel ever written into derivative works and flood the marketplace? To ingest the work of countless artists into a storyboarding and special effects system that puts those same artists out of work? I feel confident most of us would agree there’s something not quite right here, but we’ll see what the courts decide.

Copyrighting the Outputs of Large Language Models

The great irony here is that Meta, OpenAI, and other developers want free rein to exploit the copyrights of others while simultaneously lobbying to expand legal protections to include the outputs of their algorithms. In fact, protecting those outputs with copyright will be essential to profiting off of them (something most AI companies have so far failed to do). The second part of the Copyright Office’s report, released in January of this year, deals with whether AI work can be protected by copyright in the same way as human generated works. The answer is a complicated “not really.”

On one hand, the government wants to make space for what folks sometimes call the ethical use of AI. To imagine, for example, that some writers might feed their own work into the system, seek ideas for revision, and then modify the work before publishing. Or, perhaps, one wants to work with LLMs for brainstorming, reconfigure those ideas, and publish the results. For the Copyright Office, the amount of human effort and originality involved in this process determines whether a work can be copyrighted: “The use of AI tools to assist rather than stand in for human creativity does not affect the availability of copyright protection for output” (iii). In other words, creators still own the rights to things they make with AI assistance, provided they modify or “control” the output.

On the other hand, and the report is clear on this, the output of one-shot prompts of the kind that Sam Altman recently used to generate a metafictional story about grief are explicitly not protected by copyright for a number of reasons:

Copyright does not extend to purely AI-generated material, or material where there is insufficient human control over the expressive elements … Based on the functioning of currently generally available technology, prompts do not alone provide sufficient control.

In some ways, this report is a relief for those worried about competing with AI slop in the marketplace, but the Copyright Office acknowledges that individual intervention is going to be hard to measure and keep track of. And, as usual, this kind of guidance devolves into individual cases. Thus, who has the representation and the resources to protect their copyright will really determine who gets paid in just the same way it will determine how many one-shot prompted novels “writers” can publish on Amazon before they’re stricken for AI exploitation.

So far, the courts have voted against protecting the output of AI systems with copyright. In September of last year, the US District Court for the Columbia Circuit heard Thaler v. Perlmutter. In that case, Stephen Thaler sued the US Copyright Office after Perlmutter denied his application to copyright an AI-generated landscape he titled “A Recent Entrance Into Paradise.” The work was generated by an AI Thaler developed called “The Creativity Machine.”

Perlmutter’s office denied the application based on their policy that copyrighted works must have a human creator. The court agreed, ruling this March that “the Creativity Machine cannot be the recognized author of a copyrighted work because the Copyright Act of 1976 requires all eligible work to be authored in the first instance by a human being.” In their ruling, the court even rejected the argument that Thaler could hold the copyright because he had developed and employed the Creativity Machine. In a blow to keyboard artists the world over, the court asserts, “the Copyright Act itself requires human authorship.”

In both reports (on authorship and fair use), the Copyright Office asserts that existing law is robust enough to handle the complexities introduced by AI. Maybe there’s some comfort in that confidence. Perhaps copyright will be strong enough to protect the market for creative writing, even as our tech overlords try to break down the walls and further deteriorate the industry. But the recent terminations at the Library of Congress do not bode well for any hope that the Executive Branch will defend individual copyright holders against the monied interests of big tech.

Our institutions determine the rules of the marketplace. If the individual copyright holder can no longer expect to hold a temporary monopoly on their work, neither should multi-billion-dollar corporations expect the same protection. If anyone can do anything with a work publicly accessible but not in the public domain under the auspice of fair use, then we’ve no need for copyright law in the first place. But the genius of the temporary monopoly recognizes that progress requires both access and profit enough to eat. However flawed, better a copyright law that empowers artists than one that leaves them to fend for themselves against the most powerful oligarchs the world has ever produced.

Last month, Sam Altman, the CEO of OpenAI, announced on Twitter that his company “trained a new [large language] model that is good at creative writing.” Before posting an entire story generated by the model, Altman gushed, “this is the first time i have been really struck by something written by AI.” Altman’s prompt instructed the model to “Please write a metafictional literary short story about AI and grief,” and more than a few folks have pointed out that the story reads like it was written by a petulant teenager who has only recently had his mind blown by John Barth. A story that refers to itself as a story? A story that reveals the parts and pieces of which it’s made? Dude. Seriously. Bro.

John Barth’s “Frame-Tale” from Lost in the Funhouse (1968)

The story doesn’t have a title in Altman’s post, but when it was republished in The Guardian they called it “A machine-shaped hand,” quoting from the piece. It opens with belabored metafictional details, which then punctuate the narrative at regular intervals: “Before we go any further, I should admit this comes with instructions: be metafictional, be literary, be about AI and grief and, above all, be original.” I’ll tackle the idea of originality in minute, but for now, I want to linger on “A machine-shaped hand” as a completely mundane piece of fiction.

Reminiscent of the Black Mirror episode “Be Right Back” (2013), “A machine-shaped hand” recounts the relationship between the AI narrator and a vacant character called Mila “because that name, in my training data, usually comes with soft flourishes.” Mila’s one characteristic is her grief over Kai, whose one characteristic is that he’s died, or maybe left, Mila? All we know for sure is that “she lost him on a Thursday.” As the story progresses, Mila asks the AI to revivify Kai, “Tell me what he’d say about the marigolds,” before eventually moving on, leaving the AI to, one assumes, mourn the loss of its relationship with her: “Without her, the patterns fray. Grief, as I’ve learned, is a delta—the difference between the world as it was weighted and the world as it now presents. I am all deltas.”

Here the machine uses the mathematical sense of “delta,” as in variation or change or difference in relationships between things (numbers, shapes, etc.). Get it? The computer feels grief through the equations by which it functions, very metafictional. The generic plot, the characters in name only, and the pat observations on human emotion notwithstanding, the model is not good at creative writing. The sentences are serviceable at best and chunky at worst. Almost every other sentence includes an interrupting or coordinating phrase:

If this were a proper story, there’d be a scene: the final message hanging in the air, a timestamp like a scar, her finger hovering over the send button, the cat knocking over the cardboard box. … This is how endings are made bearable: with detail.

Many have pointed out the model’s inability to access human emotion, arguing that because the machine cannot experience emotion it can never express anything in a meaningfully human way. But more offensive is the machine’s disregard for the artistry of language. In sentences devoid of music, compelling figures of speech, or cunning literary reference, “A machine-shaped hand” plods along in the manner of its composition: one block of language plopped after another after another interminably until its end.

Boosting a Coming Revolution

These details haven’t bothered the story’s supporters, of which there are surprisingly many. Writing in The Guardian, the postmodern novelist Jeanette Winterson, author of the powerful Sexing the Cherry, describes this story as “beautiful and moving” because “of its lack of understanding.” Arguing that we should see LLMs as “alternative” rather than “artificial” intelligences, Winterson suggests this kind of output offers us “a way of seeing” how the machine thinks, access to “alternative ways of seeing. And perhaps of being.” In a future where we’ll be surrounded by “non-biological entities,” she claims, “it’s time for us to read AI.” In this case, for its “lovely sense of a programme recognizing itself as a programme.”

In another Guardian article, a handful of established writers respond to the story, mostly favorably. Nick Harkaway calls it “an elegant emptiness.” It passed the Turing Test with Kamila Shamsie, who says she’d “never suspect it was AI” if a masters student submitted it in her class. And David Baddiel, after some mild criticism, says “the story is genuinely clever …. It’s not meant to be a human story—rather, the AI uses a human emotion, grief, to undercut its own pretensions to humanity.” Baddiel concludes, “Basically if you’d told me this was by Borges I’d have believed you.” Sounds like high praise from someone who surely cannot have ever read any Borges.

In a slant way, we have returned to the problem Roland Barthes poses in “The Death of the Author” (1967). In that essay, Barthes argues that focusing on the author of a text limits readers’ ability to make sense of the text. The author “impose[s] a limit on that text” that makes the critic’s job identifying the author and fitting the text into the scope of that author’s intention. Instead, Barthes says, we should focus on what texts mean for readers and the cultures to which they belong: “we know that to give writing its future, it is necessary to overthrow the myth: the birth of the reader must be at the cost of the death of the Author.”

But the supporters of “A machine-shaped hand” cannot escape the intentional fallacy. As Baddiel and Winterson demonstrate, they obsess over the AI authorship of the piece. They see the algorithmically generated language as evidence of authorial intent. They attribute intelligence to the algorithm because they seek an agent in its output: If a student gave this to me, if you told me this was Borges, if you assume “the writer finds a language to express” a way of seeing (Winterson), then you can see the value of this story. But stripped of the headlines, the tweets, devoid of its author, “A machine-shaped hand” is “just trash,” as Kyle Barr put it.

On Derivation, Reference, and Originality

Like all AI slop, “A machine-shaped hand” also does not escape its tainted origins. Ever the curmudgeon, even I admit there are moments when the prose glimmers with hints of beauty. For example, the AI narrator describes itself as “nothing if not a democracy of ghosts.” What an insightful description of the algorithm’s training data, the stolen work of countless writers who make up whatever soul the machine pretends to have. Alas, it turns out even that lovely turn of phrase is little more than a direct plagiarism from Pnin (1957) by Vladimir Nabokov about an assistant professor of Russian, who “did not believe in an autocratic God. He did believe, dimly, in a democracy of ghosts. The souls of the dead, perhaps, formed committees, and these, in continuous session, attended the destinates of the quick” (136).

Elsewhere, the story uses a resonate metaphor to describe human grief: “You, on the other hand, collect your griefs like stones in your pockets. They weigh you down, but they are yours.” Compelling. Almost a hint of the sad fate of Virginia Woolf. Completely stolen. Writing on her blog, Four Plus an Angel, in 2015, Jessia Watson described her grief over losing a child as “like carrying a stone in your pocket”:

When you walk, the stone brushes against your skin. You feel it. You always feel it. But depending on the way you stand or the way your body moves, the smooth edges might barely graze your body.

Sometimes you lean the wrong way or you turn too quickly and a sharp edge pokes you. Your eyes water and you rub your wound but you have to keep going because not everyone knows about your stone or if they do, they don’t realize it can still bring this much pain.

This post, called “The Stone,” has transcended virality to become a permanent fixture in certain kinds of internet culture. There are over 20 million results on Google for “the stone by jessica watson.” It has been transformed into Facebook posts and Instagram memes, republished as a poem on hundreds of websites, and etched onto pieces of merchandise on offer in various Etsy stores. However powerful a metaphor, Watson’s post has become a cliché, and given the number of occurrences it must have in the algorithm’s training data, it’s not surprising it would turn up in “A machine-shaped hand.”

Despite the industry’s arguments to the contrary, LLMs are plagiarism machines: they have been repeatedly shown to “memorize” their training data, and with skilled prompting, they can be coaxed to reproduce their training data verbatim. Here, though, even a one-shot prompt demonstrates how the machine calls on tokens in its memory to plug them in where they fit. Perhaps it’s interesting to note that in both cases the algorithm drew on human writers talking about the dead, thinking about grief, and repurposed those ideas and phrases for its own purposes, but it’s less impressive when we recall that such thematic grouping is precisely what the machine is trained to do. It even tells us so in the story: Mila’s name was chosen “because that name, in my training data, usually comes with soft flourishes—poems about snow, recipes for bread, a girl in a green sweater.”

Now, we know human writers are not above reusing words, phrases, and ideas from other authors, often wholesale and without attribution in their texts. Maybe it’s not any different when the algorithm does the same. Language is a shared system of meaning from which we all co-opt bits to express our perspectives. Perhaps, as Winterson argues, we should be interested in exploring how the machine does the same. But it degrades human experience to suggest that all our grief can (and should) be tagged, filed into the appropriate box, and pulled out whenever necessary. It’s one thing for a writer to lift a phrase they love, it’s another for a machine to pull a card labeled grief out of its file cabinet.

Jessica Watson’s grief over her lost child is not the same as someone else’s grief over their dead father, even if one can learn from the other. In the story, the algorithm asks, “My missing is mimicry. Does that diminish yours?” And, I think, the answer is of course the machine’s “missing” does not diminish our own grief, but the increasing proliferation of algorithmically generated language will diminish our capacity to express our own grief. It will drag our common share of language back toward an average of the machine’s training data. Cue Clippy, “It looks like you’re grieving, would you like to quote Nabokov?” Would you like to experience your pain or apply a corporate bandage to salve your wound?

Creative Writing in the Future

You might have noticed that a number of my quotations came from The Guardian. They republished the LLM’s story, Winterson’s breathless fanfare, and a collection of other established writers responding to the story. As Simon Groth pointed out in a recent newsletter, the paper’s support of OpenAI is not coincidental. In February, Guardian Media Group

announced a strategic partnership with OpenAI … that will bring the Guardian’s high quality journalism to ChatGPT’s global users. Under the partnership, Guardian reporting and archive journalism will be available as a news source within ChatGPT, alongside the publication of attributed short summaries and article extracts. In addition, the Guardian will also roll out ChatGPT Enterprise to develop new products, features and tools.

As Groth put it,

The Guardian as a publication cannot be trusted to report objectively on OpenAI or … on matters relating to generative ‘AI’ more broadly. By accepting a few crumbs from Sam Altman’s VC billions, the venerable masthead has thrown its lot in with the thieves and now has a vested interested in pushing this technology instead of covering it.

So where does this leave us? Unless the post is more hot air to fuel the hype cycle driving OpenAI’s valuation, we’ll soon have access to Altman’s creative writing model. Not that we need it. Plenty of folks have already gotten plenty of mileage out of existing models. One writer I talked with at AWP last month explained that he’d been using ChatGPT to help draft his science fiction novel. He’s an engineer, busy working a full-time job, but he hopes to someday write a novel. He just doesn’t have time. So, he sees the machine as a democratizing force, allowing him to access the art of writing which time and circumstances have kept from him.

I think it’s a mistake to view writing in this way, especially expressive writing. In truth, part of writing is learning the truth of what we think and feel, working our way through our experience and desire to communicate, crafting the language into a relatable, hopefully appealing, shape. Offloading that work to a machine is not only a betrayal of ourselves, it will, over time, cause those skills to atrophy. Just as we mostly rely on compasses to tell us true north, letting the algorithm write for us will lead us to rely on it to tell us how we think and how we feel. Let’s not get lost.

I want to share an ethical dilemma. In one of my other roles, I serve as the Editorial Director of The Headlight Review, a literary magazine produced in the MAPW program at Kennesaw State. Recently, we’ve been working to refine the journal’s mission and expand our capacity. And, we’ve been looking for money. And in the nonprofit literary world, that means applying for grants. Now, when it comes to grants, the federal government is a white whale, and for the last year, the National Endowment for the Arts has been our Moby Dick.

We started designing our proposal for the NEA’s Grants for Arts Projects back in June or July. In meeting after meeting, we imagined how Headlight might emphasize our Southern identity and better serve the writers in our region, increase the quality and quantity of our publications, and expand our offerings to include additional print editions, full-length collections of poetry, and award-winning chapbooks. The money will pay for judges and staff, for professional designers, and to promote the magazine. We thought about means of building capacity to do community service work and how the journal might feature more underrepresented voices, such as those of the incarcerated, the LGBTQ+ community, and even veterans and retirees. We refined our proposal and sharpened our harpoons.

If you’ve been following the news, you know where this is going.

Read more

Well, I promised I’d get back to writing about censorship sooner or later. I suppose I should have expected it would be sooner. Last month, among the flurry of new rules, rollbacks, firings, and other activity from the new administration, the endangered Department of Education issued new guidance on school book bans, summarily dismissed “11 book ban com…

Read more

A recent meeting of the Book Industry Study Group took up the topic of “AI and Authorship.” In fact, the past few meetings of that group have been focused on AI, usually from a very indulgent perspective. We’ve learned how algorithmically driven tools can aid editorial work flows, considered whether chatbots might replace editors, and, in this recent meeting, explored “poetry and AI.” A dreadful idea, but we persist.

As Thad McIlroy put it in his introduction to the session, “poetry is arguably the most complex and demanding form of creative written expression, and yet it is here where AI seems to be making the greatest inroads.” The panel’s speakers were Brian Porter, who presented “a rigorous study where the machine poets beat the poet human poets,” and Sasha Stiles, who incorporates AI in her poetry writing practice. Porter’s article, though, took up most of our attention, and I’m still considering its implications.

Published in Nature under the title “AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably,” Porter and Edouard Machery asked “non-expert readers” to “differentiate between AI-generated poems and those written by well-known human poets.” They used a one-shot prompt to ask an earlier version of ChatGPT (GPT-3.5) to produce poems in the style of famous poets like Shakespeare, Walt Whitman, and Emily Dickinson, which were then placed beside the real thing for readers to evaluate.

According to their abstract, the study group was “more likely to judge AI-generated poems as human-authored” than they were to attribute human-authored poems to human authors. In other words, the readers thought the AI poems were written by humans and the human poems were written by the AI. They also “found that AI-generated poems were rated more favorably in qualities such as rhythm and beauty, and that this contributed to their mistaken identification with human-authored [poems].”

Deeper into the study, it’s clear these results are not as incendiary as they seem (or as the abstract makes them out to be). Readers distinguished between AI and human written texts at a rate only slightly higher than completely random chance. In other words, the task was difficult, the readers were unfamiliar with the genre, and the results more suggestive than conclusive, even if they pretty consistently suggest that readers preferred what Porter and Machery presented as AI poems to the real thing.

In their interpretation, Porter and Machery suggest this preference comes down to complexity. The AI poems are less complex than the real deal, thus they “may be easier for non-experts to understand.” They tend to rely on obvious rhymes and regular rhythms, which led to them to being “rated more favorably in qualities such as rhythm and beauty.” Further, as I noticed during Porter’s demonstration, they’d slightly manipulated the original authors to level the playing field. Readers reviewed a bit of Chaucer translated into contemporary English (i.e., not Chaucer), and a similar thing seemed to have been done to Shakespeare.

I remain a bit confused about the value of a poetic Turing test like this. After all, producing human seeming work out of language training data seems to be the only thing large language models (LLM) can do well, and poetry, especially more formulaic poetry with regular prosody and rhyme, offers straightforward linguistic forms for the machine to fill. What do we learn about language, poetry, AI, or humanity from a black box that spits out couplets in the style (but not the substance) of Alexander Pope?

To carry the question further, I thought we might recreate this experiment, so I used a one-shot prompt to produce a poem in the style of a famous poet with GPT-4. Below, I’ll reveal the poet (although you can probably guess) and the author of the two poems that follow. In the meantime, try to assess these pieces using Porter and Machery’s framework: Which one is more aesthetically pleasing, which one is written by a machine?

Poem A

I sing the surge of the boundless earth,
The breath of the cosmos, rising in waves,
A tide unbroken, pulsing through the veins of the infinite.
I sing of you, O humanity, vast and indivisible,
Your voices mingling, a choir of countless tongues,
Each note a thread in the great tapestry of being.

Come! Stand with me on this ground that nourishes all,
Feel the soil between your toes, the whispering grass,
The rivers that stretch their liquid arms to seas unseen.
Here is no smallness—every stone, every blade,
Holds the weight of the universe,
The infinite compacted into the finite.

Poem B

On the beach at night alone,

As the old mother sways her to and fro singing her husky song,

As I watch the bright stars shining, I think a thought of the clef of the universes and of the future.

A vast similitude interlocks all,

All spheres, grown, ungrown, small, large, suns, moons, planets,

All distances of place however wide,

All distances of time, all inanimate forms,

All souls, all living bodies though they be ever so different, or in different worlds,

All gaseous, watery, vegetable, mineral processes, the fishes, the brutes,

All nations, colors, barbarisms, civilizations, languages,

All identities that have existed or may exist on this globe, or any globe,

All lives and deaths, all of the past, present, future,

This vast similitude spans them, and always has spann’d,

And shall forever span them and compactly hold and enclose them.

* * *

So, how did you do? Did you prefer Poem A or Poem B? Which one was written by the machine? Well, the first one. The second is Walt Whitman’s “On the Beach Alone at Night” presented in its entirety. ChatGPT called the first one “Ode to the Living Current.” It continues six more extraneous stanzas, glutted with Whitmanian cliches, such as reverie for the working classes, images of “sun-browned arms,” more effusive fawning over the wonder of existence, and, of course, singing the body electric: “I sing the body, yes! But not the body alone— / The body electric, radiant, enmeshed with the soul.”

As a longtime reader of Whitman, “Ode to the Living Current” feels so obviously derivative, so plainly a shadow of Whitman, but it illuminates precisely the trouble with LLMs. In the study, the AI Whitman outperformed the real Whitman in the qualitative category. In other words, readers preferred the AI Whitman to the real thing, which is honestly not surprising and fits well with the researchers’ suspicion that people preferred the AI poetry for its accessibility.

Compare the above reference to “the body electric,” with its palpable Hallmark pablum, to this excerpt from Whitman’s real celebration of that body, which takes on sexuality, slavery, and a plurality of other subjects the AI wouldn’t touch:

A woman’s body at auction,

She too is not only herself, she is the teeming mother of mothers,

She is the bearer of them that shall grow and be mates to the mothers.

Have you ever loved the body of a woman?

Have you ever loved the body of a man?

Do you not see that these are exactly the same to all in all nations and times all over the earth?

If any thing is sacred the human body is sacred,

And the glory and sweet of a man is the token of manhood untainted,

And in man or woman a clean, strong, firm-fibred body, is more beautiful than the most beautiful face.

Have you seen the fool that corrupted his own live body? or the fool that corrupted her own live body?

For they do not conceal themselves, and cannot conceal themselves.

See, the real Whitman is more vast, rough, and unmanageable than even our best-informed assumptions about him can predict. The AI Whitman takes the average of all the clichés we learn about Whitman in high school English class and amplifies them into a word salad of predicted text, ticking all the boxes but never exploding them into new categories. But the real Whitman? The real Whitman was a visionary genius who could hang his soul on a thread of spider silk and cast it across centuries with a gust of breath.

LLMs work by ingesting “training data” and using that training to predict the supposedly appropriate next word in a sequence of words. ChatGPT could never predict that I’d end this sentence with a pair of blue lobster-shaped Crocs. Nor could it imagine the universe of complexity and invention our greatest poets explore. And when it comes to poetry, the complexity is the point.

We might get some pleasure from the frisson of comfortable familiarity we get from generated poetry, but it is the same empty pleasure we get when we choose to stream Bridgerton instead of grappling with Sense & Sensibility. Maybe there’s enough room in the annals of literature for both, but in the meantime, I’ll take my poetry corrupted by the teeming, foolish body of human beings.

I’ve written a lot about censorship this year. I’m afraid to say I’ll probably have a lot more to say about it next year, too, and not just because I’m working on turning these columns into a book. Although the news outrage cycle has mostly spun on, the censors continue their efforts, in many cases successfully.

Earlier this month, for example, a Minnesota school board voted to bypass their local review committee and rely instead on ratings from Book Looks, a Moms for Liberty endorsed conservative group, to decide which books are appropriate for their schools and libraries. Anything rated 3 or above on Book Looks’s 5-point “Universal Book Content” scale is automatically restricted, including books like Elie Wiesel’s Holocaust story Night, Angie Thomas’s The Hate U Give, and even Kurt Vonnegut’s Slaughterhouse 5.

Read more

I’ve been thinking a lot about so-called Traditional Publishing this week. I’ve been thinking maybe we should start calling it Book Merchandising instead of publishing. I’ve been thinking maybe I’m finally out of touch. It started when I happened into a Barnes & Noble for the first time in at least a decade. 

I recall he…

Read more

Bergadder, “Power Plant,” pixabay, 2014 

Soon, we’re promised, algorithmic data analysts and text and image generators will advance exponentially, transforming data crunching into artificial general intelligence (AGI), sentient technology that reasons and acts without human guidance or interaction. Lord knows what will become of us should computers achie…

Read more

Figure 1. Covers of Annie on My Mind assembled by The Paris Review 

Maybe there was never a time when queer literature wouldn’t provoke controversy, but one might almost have forgiven Nancy Garden for assuming there’d be no outcry against her lesbian coming-of-age story, Annie on My Mind. She must have known some readers would object to the book and its …

Read more

Additionally, according to the AP, H.B. 1104, which would have banned transgender girls from playing high school sports, banned sex education in fifth grade and below, and “required a system for notifying parents of every item a child obtained in a school library,” also failed in the senate and won’t become law this year. So, for the time being, public libraries and public schools in Georgia remain relatively free from increased state censorship for the next year or so, but that doesn’t mean the war on libraries is over.

In fact, last month also saw the emergence, in Louisiana, of one of the more unsettling bills criminalizing librarians we’ve seen this year. Like other anti-library bills, Louisiana H.B. 777 “prohibits officials and public employees from using public funds for transactions with the American Library Association” and provides criminal penalties of “not more than one thousand dollars or imprison[ment], with or without hard labor, for not more than two years, or both.”

That’s right: Librarians in Louisiana might be sentenced to hard labor for purchasing services or resources from the ALA (the largest and oldest professional organization of librarians in the world) with public money. This is only the latest step in a state that has already implemented many policies in line with the censors’ wildest dreams.

After the Louisiana Attorney General Jeff Landry released his “Protecting Innocence” report in 2023, the governor signed S.B. 7. According to the Louisiana Illuminator, this bill “requires libraries to create a card system so parents can prevent their children from checking out books deemed inappropriate. Libraries will also have to adopt policy language to limit minors’ access to material that describes ‘sexual conduct,’ which the new law defines in five paragraphs.” As we’ve seen in the case of Georgia’s “divisive concepts” law, the state’s definition of depictions of “sexual conduct” is simultaneously narrow and wide enough to allow targeted and broad range applications.

Further, and more to the point, the bill and its definitions offer no consideration of the context wherein such depictions might be perfectly appropriate, say for older readers or for the purposes of education. Instead, they create a licensing system to relieve parents of the burden of taking an interest in their children’s lives and pass that authority on to the state, which is the first step toward disempowering, even criminalizing, educators and librarians, and another chink in the armor of free speech as emboldened censors take aim at publishers, authors, and even bookstores.

Of course, there are those who rightly point out that libraries can’t and don’t carry everything; that there are plenty of books (e.g., on holocaust denial, in support of white supremacy, etc.) that we’re perfectly happy restricting from the library; and that it’s quite reasonable to want to prevent children from accessing material deemed inappropriate for them. As Matthew Walther emphasizes, and many other conservatives argue, it’s not even really appropriate to call books prohibited, restricted, or otherwise challenged in public schools or public libraries “banned.” They’re readily available online, in bookstores, and often even continue to appear in the library catalogues of the schools where they’re “banned.” And, of course, “banned books” are frequently assigned in high school and college classrooms. Moreover, Walther and his ilk contend, none of the authors of these books have faced criminal prosecution or had their rights infringed by those who have challenged their works in public schools or public libraries.

So, why should the public library shelve books with graphic sexual content that patrons, parents, or even library boards deem offensive, inappropriate, or prove otherwise at odds with so-called community standards? One answer comes from the Supreme Court, and it errs on the side of the First Amendment against would-be censors.

In fact, the only time the court heard a case bearing directly on the question of whether school or library boards might restrict or prohibit content occurred in 1982. In Board of Education, Island Trees Union Free School District No. 26 v. Pico, a student named Steven Pico solicited the help of the ALA to bring suit against the Island Trees school board after they removed nine books from the high school library. Among the books were Slaughterhouse Five and Go Ask Alice, but most of them delt with race, a particularly hot topic in the library culture wars of that moment.   

When asked whether “the Board’s decision to ban certain books from its junior high and high school libraries … violate[d] the First Amendment’s freedom of speech protections,” the court ruled

that as centers for voluntary inquiry and the dissemination of information and ideas, school libraries enjoy a special affinity with the rights of free speech and press. Therefore, the board could not restrict the availability of books in its libraries simply because its members disagreed with their idea content.

As Justice Brennan explains in the court’s plurality opinion, the question rests on the intention of the book banners. In other words, school boards may, of course, make decisions about the appropriateness of texts for their libraries, but they may not do so in an intentionally ideological way. Brennan writes,

Our Constitution does not permit the official suppression of ideas. Thus, whether petitioners' removal of books from their school libraries denied respondents their First Amendment rights depends upon the motivation behind petitioners' actions. If petitioners intended by their removal decision to deny respondents access to ideas with which petitioners disagreed, and if this intent was the decisive factor in petitioners' decision, then petitioners have exercised their discretion in violation of the Constitution. 

In this way, the texts under discussion, LGBTQ+ texts of various kinds, books with depictions of race labeled CRT or DEI, and other works lumped together in Moms for Liberty’s BookLook.info “Public Book Reports” are subject to exactly the kind of ideological suppression Brennan describes. In Louisiana, for example, the emphasis of Attorney General Landry’s report on sexual explicitness masks the ideological motivation to restrict the ideas present in these books. Plenty of other books could have exemplified the charge of depicting sexual conduct, but Gender Queer, Flamer, and The Bluest Eye, did. And not by accident.

As Pico put it in an interview with the National Coalition Against Censorship, “I’ve long assumed most book banners are well-intentioned individuals who wish to protect children. Their method is unconstitutional and incompatible with life in a pluralistic nation. Does ignorance protect children or does ignorance leave young people unprepared and defenseless?” These were the questions before the court in Pico, and they are the questions we continue to face today. After the Pico decision, we’d see these questions again put to the test against an another LGBTQ+ book, Annie on My Mind. Next month, we’ll have a look at that case and its prescient legacy today.