by mantunes

This essay is dedicated to the Star-Ledger, which printed its last edition near simultaneously with the release of this essay and whose coverage first entranced its author with the power of writing.

I want to thank our readers for sustaining this magazine as we celebrate our fourth anniversary. When we launched The Newarker in 2021, we made a commitment to the written word. Newark’s steady collapse in published journalism left a palpable vacuum in the local sphere of ideas, one we felt compelled to fill with essays, critique, and other writing that took this city seriously and needed time to digest. In an era of TikToks and tweets, our magazine chose to prioritize long-form content. 

The Newarker, by many counts, has been a remarkable success. We receive more content than we can responsibly publish. We are often the first and the only venue that has published Newark’s emergent and aspirant writers. We bask in the praise (through emails and Reddit comments) on the quality of what we have published. And yet, this past year has filled me with a low-grade dread that this all has become increasingly pointless.  

Reading and writing are under siege. Publishing fully-formed thoughts has become as close to obsolete as possible. Arguments about the economics of writing are often trotted out as the greatest risk facing writers—there once was a time where a figure like James Baldwin could eke out an existence solely on publishing a book every five years and a handful of magazine articles, dead-end screenplays, and reviews. That this magazine is a self-funded project among friends is the best evidence that there is no money in writing. However, I believe a set of cultural and technological forces are driving an even greater change in the field. Two forces in particular, one on the demand side (reading) and the other on the supply side (writing), are irreparably shifting the landscape.

The greatest pressure on the demand side is the noted collapse in reading habits among younger generations. Much hay this year has been made about this trend with articles appearing in The Atlantic, The New York Times, The Economist, the National Review, and even The New York Post. The National Assessment of Educational Progress recently found that only 33% of eighth-graders can read at grade level—the lowest in the exams’ three decade history. The consensus is that the kids aren’t reading, and this is bad. 

I witnessed¹ this myself last semester. I taught a seminar about Newark’s history after 1967, and I had a sneaking suspicion that 90% of my students did not read a single page. This made my two-and-a-half hour long seminar unbearable. Every time I asked something about the reading, I faced a combination of blank stares and puzzled looks. Granted, my class was entirely a crowd of STEM students who were there to fill an elective requirement. They told me how stressed they were over projects for their other classes (particularly those required for their major). I felt sorry for them, but I was also taken aback by the near open disregard there was for the act of consuming long-form textual information. I chose not to assign a full book for the course to save my students from having to purchase material unnecessarily. I am glad I didn’t. Most of them wouldn’t have opened it anyway. My goal in teaching the course was to impart some impressionable youngsters with the weighty issues that have affected this town and its history; it didn’t matter if they got it through reading or solely through my lecture. 

Theses abound as to why this is the state of affairs. Almost instinctively, commentators will blame the state of schools. Frequent reporting of the lowest test scores in reading seen in decades will often precede a critique of the quality of teachers, the lack of funding in public education, and a general malaise around the priorities of higher education. Conservatives will generally blame liberals for their woke ideology; liberals will blame conservatives for constant budget shenanigans and creating a hostile environment in schools. Except for a brief blip in the after the Second World War, America’s schools have never been stellar. Still, we continually graduated a society of readers, consuming newspapers, pulp fiction, magazines, and other printed material, enough to sustain a massive multi-billion dollar industry, until late in the 20th Century. 

I suspect something larger is at play. Jonathan Malesic’s guest essay in The New York Times gave me some insight into the overall culture of reading in this country. One quote has been living rent free in my head:

Recent ads for Apple Intelligence, an A.I. feature, make the vision plain. In one, the actor Bella Ramsey uses artificial intelligence to cover for the fact they haven’t read the pitch their agent emailed. It works, and the project seems like a go. Is the project any good? It doesn’t matter. The vibes will provide. 

I think this nails down the issue. Long-form reading, while providing some personal pleasure once, was also crucial to operating in society. Not long ago, nearly all levels of society could be expected to read a manual or large document to obtain some important information. Why read now if you can just go to YouTube to find a step-by-step video explaining the process? Heck, you can even ask ChatGPT to give you the synopsis of nearly anything. Moreover, you don’t even have to understand the document. So long as the vibe is right, it’s okay. 

This is reflected in my anecdotal experience. Those who know me, know that I nearly always carry a book with me. If that book exceeds 300 pages, I often get a bemused look, like how could you possibly read something that long. The irony is that I find books longer than that to be some of the most readable books I have come across. Oftentimes, shorter books can be some of the most impenetrable. Either way, I find myself the only person reading books of a certain size. Why read a book when you can easily find a synopsis on YouTube or BookTok?

This is why I found the whole crisis over “book banning” over the last two years to be so perplexing. You would have thought there was an epidemic of reading, given the energy the two sides of that debate brought to the conflict. The critic and media ecologist Neil Postman opened his seminal Amusing Ourselves to Death (a work I keep coming back to) with the observation: “What Orwell feared were those who would ban books. What Huxley feared was that there would be no reason to ban a book, for there would be no one who wanted to read one.” We are definitely in the “don’t want to read one” dystopia.

Then, there is the demand side. Artificial intelligence, in the form of large language models and generative AI, may completely undercut the human act of writing. When ChatGPT first became publicly available in the early winter of 2022, it was noted with as much media fanfare as the moon landing. Many people vaunted the capabilities of the system for generating answers to questions. A former coworker of mine was an immediate convert. He described ChatGPT as a panacea to all his workplace minutiae. He said that he was using it from high-level strategic planning to project budgeting to brainstorming. I was incredulous. 

The cracks were also immediately apparent. While the program could generate colorable and cogent responses to questions put to it—check out this magazine’s early conversation about Newark with ChatGPT—it also had a propensity for making stuff up when it did not know the answer. Advocates for generative AI would call this “hallucinating,” a delightfully anthropomorphic euphemism that made it seemed that it had taken one too many bong hits. However, this hid a more dishonest aspect of what it was doing.

For example, several incidents in the law revealed the true problems with using such programs for generating legal briefs. A handful of lawyers were caught using ChatGPT because the cases that were being cited for legal authority were actually completely made up. Each of these lawyers did not even bother to check the output, which sounds about right for the overworked and stressed out associate who just wants to meet a deadline. What’s more concerning is that the generative AI program did not have the Socratic capability of understanding its own lack of knowledge. It did something deeply human; when presented with a question it did not know to answer, it vamped. Stories like these fed my initial antipathy to these systems. In true Luddite-fashion, I refused to use them in work and in personal life.

The hallucination problem was not a fatal flaw, easily overcome by more testing and tinkering with the base system. Generative AI still had a use-function. My initial incredulity, I suspect, came from somewhere deeper. Much like a Luddite who saw the mechanical looms popping up across Northern England, I knew these systems were coming after me. I couldn’t quite put my finger on it, but I came up with a bit of a phrase to describe both what let me sleep at night but kept me up anyway. Generative AI’s ability to write is just as good as any high school senior or college freshmen. To put it differently, it’s not transformational writing but it gets the job done. 

As with any other technology, proponents will often tout its time saving capabilities. “You can spend more time critically thinking, and less time working on that email.” I’m sure you’ve seen these ads if you’ve ridden mass transit in New York City, where a substantial amount of knowledge workers use it to get to work. There is a cruel irony to these ads. These workers aren’t usually in a position to purchase these programs for their companies but are definitely at risk of losing their jobs because of it. I think that’s because the value of critical thinking and overall strategic planning is hyped in our culture but undervalued in practice. If there is something I learned from my professional career so far, it’s that everyone says that they value critical thinking but no one actually cares for a critical thinker. When presented with two work products, one that is serviceable and costs a tenth of what it normally takes in time and effort to produce (and doesn’t question the prompt) and another that is deeply well thought out but provides serious some serious critique, people will in the long run pick the former and not the latter.

But what about creative writing? Surely, there will be space for people to create novels, poetry, and plays. How could something not human produce something that reflects the human experience and resonates with an actual human? The truth is that I do not know. I keep waiting for that shoe to drop. I honestly would not be surprised if some generative AI program publishes (anonymously) what ends up being considered “the Great American Novel,” fooling critics and general readers alike, passing a crucial rung on the Turing Test ladder. I suspect that some headline in this variety (perhaps, a screenplay or a collection of poems) will happen in the next two years and will lead to some serious soul-searching by the whole reading and writing establishment. 

Should we continue writing? Again, I don’t know. Sam Graham-Felsen, historian and curator at the Philip Roth Library and Reading Room at the Newark Public Library, gave me a piece of advice recently: writing is not therapy. That is, true writing where there is an expected audience and where you are conveying ideas or a theme is separate from the act of dealing with trauma or resolving pent up feelings. I still believe that writing is not about expressing something solely personal but that you should be conveying something of meaning. The question is whether generative AI can convey that same level of meaning without the effort, the editing, the self-doubt, and the heartache. If so, then the only reason to write would be for the pure masturbatory effect. 

The Newarker is not going anywhere. We have not reached that stage. Generative AI is, at most, at the simulacrum stage of the writer. (Though, the true dystopia is the ouroboros of generative AI creating pieces which it is then tasked with synthesizing and summarizing, and we humans just watch from the sidelines.) But in the long term, we who write are not too dissimilar to the Japanese soldiers left behind on the Pacific Islands, thinking the war is still going on, pointlessly attacking anyone who confronts them, and waiting for the Emperor himself to personally issue the order to lay down the pens. So, please, keep sending us your pieces, and go to bed soundly knowing that futility has at least a few years before it truly takes hold. 

1. Recently, I had an interaction with a high-performing high school student here in Newark. They had a copy of The Crucible with them for their AP Lit class, and I began to rattle off some of the characters from the play, trying to get them to name check Abigail (a crucial character whose name I couldn’t quite remember). They gave me a blank stare. Assuming they had just started reading it, I asked the student how long they had been reading the relatively short play. They said they had been at it since November. This conversation happened in mid-January . . . . ↩︎

mantunes is a resident of Newark who still writes about the city (and other things) and definitely did not use generative AI to create this piece.

Featured Image: Tom Wiedmann. Tom is a photographer and current Newark-based employee for a local nonprofit. His work has been showcased in local publications and, most recently, the Nork! Photo Fest, where it was featured in outdoor exhibits located in Harriet Tubman Square and near the Passaic River in the Ironbound.