The Great AI Hoovering; or, The Disappearance of Magical Human Language (Part I)
Why Reading Matters in the Age of Generative AI
This Substack is subtitled “A Tourette’s Life,” and up to now I’ve mostly written about TS here. I know many of you new subscribers have come here to read about this part of myself, and I get that. But I find it important to recall one vital fact: As Walt Whitman wrote in his poem “Song of Myself,”—a line carved into another great song I love, by Bob Dylan, more than a century and a half later—“I contain multitudes.”
Human beings do contain multitudes, and I’ve always been particularly keen to burst free of any confining categories (or to not get trapped in them to begin with). Over the course of the next year, writing only what I can fit inside the span of two-to-four hours on Sundays only, I’ve challenged myself to use this space to dig deeply into my varied areas of interest. Tourette syndrome is part of this cornucopia—but not all of it.
For the past two weeks my 82-year-old father has been in the hospital. On Wednesday, if all goes according to schedule, he will have open heart surgery. We live in an age of prolonged life, an age when political stability and scientific advances have drastically altered the course of humanity. Once upon a time, a good long life might’ve been 50 or 60 years—and chances are that if you lived that long you’d suffer from maladies such as unhealed broken bones or toothaches caused by painful abscesses. It’s not like there were skilled dentists in the Bronze Age to fix your bleeding gums.
In fact, it wasn’t until the middle of the 20th Century that things truly started to get better for most people—at least those lucky enough to be born in wealthier regions of the globe. The infant mortality rate went down and people, no longer the uneducated peasants of yore, could lead generally rich and fulfilling lives in pursuit of knowledge and pleasure.
Since they didn’t have to work on a farm or a factory or give birth to 14 children, they could go to the movies every weekend, if they wanted to; they could go on vacations to sunny Disneyland and get their pictures taken with a costumed princess; they could study humanities in college and prepare for a career as a professor; they could casually drink beer and watch sports 24/7 or binge a favorite TV sit-com on a nightly basis. Hell, they could grease themselves in lard and smoke a bong naked if they wanted to. For the right price, you could do almost anything you wanted. If you’ve lived through the past eighty years of American history, you’d be forgiven—almost—for forgetting that for most of human history, a long life was a hard and often bitter one. In many parts of the world, it still is.
Today, we have cures for diseases like polio or malaria that once wiped out multitudes. We have nascent stem cell technology that could prolong human life even further (if we wanted it to) by replacing diseased cells with healthy ones. We have highly trained specialists, like open heart surgeons, who can pause a human heart, replace a valve or insert a stent, and start the heart up again. It is a beautifully amazing thing, a beautifully human thing. What we’ve witnessed across the broad, sweeping prairie of the first 300,000 years of Homo sapiens is the evolution of none other than human progress.
And human is my point.
I’ve visited my dad a few times over the past couple weeks. I’ve sat in his room listening to all the beeps and clicks of the machines attached to his body, I’ve watched the comings and goings of nurses and doctors, and I’ve watched him get better. Humans currently employ machines to cure our ailing, mortal coils. But Dad is bored; he’s ready for the surgery. Sometimes he sits in the easy chair reading a book, sometimes he plays a word game on his phone. I’ve written for Lithub about how much he influenced me as a reader, but I honestly don’t know what to say to him much of the time.
Still, he’s not really bothered by my silence or anyone else’s, I think, and I’ve learned that from him too. He doesn’t need to talk. Nor do I. When I’m there I take my blank pages and I lay on Dad’s bed, scratching my thoughts on paper. Mom sits reading in one chair, or if not that, playing 7 Little Words on her phone. Nurses and doctors come and go. It’s pretty mellow, and I can see why Dad is totally bored. I’m not sure if I’d say he’s excited to have the surgery, but Dad certainly knows that he’ll be better once it’s over. And he’s ready to be better soon.
This is great on a private level, for my family that is; we’re all ready for him to be better. Beyond the private, however, like many of you I’m finding it difficult to read the news these days without feeling unbelievable overwhelm and despair. As a society, we are decidedly not going to be better soon. And we are certainly not going to be great.
AI and Storytelling
What I wrote in the section above was a longish intro for me to talk about my real subject this week: artificial intelligence and storytelling; I am particularly talking about generative AI, the shit that’s being foisted upon us all by greedy corporate conglomerates that have gone all-in, to hell with the consequences for life on this planet, or even, for that matter, the human brain.
That last sentence is a tricky one, wouldn’t you say? If you typed the following into whatever AI platform you choose, write a transitional sentence that introduces the concepts of AI and storytelling, you would probably not get a sentence like that to spit out. For one thing, to slow its pace I deliberately loaded it up with punctuation, especially commas, and judging by the “AI Overview” texts that now appear every time I Google something, AI platforms typically render language in fast, easily microwaveable bytes, on the assumption, I imagine, that Americans read with the comprehension of a sugared-up 7th grader.
That means lots of crispy little sentences tagged on their butt-ends with the smudges we call periods. That means fewer commas, and when commas are used, they are followed by a conjunction or simple lists. That means, in all honesty, pretty bland stuff. Of course, if you ask your AI master to write text in the style of Finnegan’s Wake, you’ll doubtless get any number of variations of “swerve of shore to bend of bay, brings us by a commodious vicus of recirculation,” because AI texts are a simulation of human voice and personality. They are language-gobblers—or vacuum cleaners—that indiscriminately hoover up the real work of uncredited artists.
To read a complicated sentence with meaning you have to read it s-l-o-w-l-y, you have to read it and all that comes before and after it determined to understand the whole configuration as a single train of thought, and so you have to pause periodically to catch your breath. Thanks to decades of anti-intellectualism in this country, the act of reading itself has been degraded significantly, to some extent even vilified. Surely, it’s no coincidence that MAGA types are out in public gnashing their teeth and clutching their pearls to ban books. Reading opens the mind, and open minds tend to oppose authoritarianism. People with open minds also tend to be independent thinkers who don’t like to have generative AI shoved in their faces.
When people talk about generative AI, they typically talk about how quickly and “efficiently” they get things “written”—never mind for a moment that they are not actually “writing” anything—or how easy it is for them to type in a few words and have workable text pop out like bread from a toaster. With that draining task complete, they can finally pick that annoying, clumpy booger that’s stuck in their nose or take a big dump.
Needless to say, AI platforms do for writers what music streaming platforms do for musicians: they pilfer art from actual artists to be exploited by the masses without fair recompense for the artists. Like an algorithm pooping out cookies of exactly equal size and shape for everyone to eat simultaneously. Let’s NOT forget, after all, that these generative AI platforms were ALL trained on millions of stolen books, including my own. I’ve seen too many ignorant social media complaints that the em dash, a gorgeous and important piece of punctuation that looks like this (—), is a telltale sign that text was written by generative AI.
Bullshit. It’s only that way because millions of actual writers know how to fucking use the em dash properly, and these generative AI platforms have stolen it from them—from all of us. From you. You should be as pissed as I am. The em dash is a hugely effective device to use when forming meaningful, written human language.
The consequences of generative AI will reverberate tragically for the entirety of humanity. Let’s say you’re a forty-something adult educated in the traditional way, reading books, writing school papers built on your own ideas, etc. It may take a while, but if you rely on generative AI to do your work for you, eventually you will lose the ability to create and maintain a cogent thought; if you get stuck, you’ll turn to a computer to unstuck you. But let’s say you’re a forty-something adult with a ten-year-old daughter. Will she develop the necessary skills to read deeply and critically, to write extended blocks of understandable text, to articulate her innermost thoughts? Especially if she sees you turning to a computer program to do your work for you?
With generative AI doing the work once done by the human brain—and I’m not just talking about writers and artists now; I’m talking about our children who are essentially learning what’s important is not reading and comprehending language but going online and taking it whenever and wherever, as if knowledge were nothing more than crude oil we can draw up and discard at will—how many years will it take until people are like those chubby humans in the movie Wall-E? Unable to do anything intellectual on their own? Stupid with laziness without the assistance of computers?
Go back and reread the first sentence of that last paragraph. That’s another challenging sentence to read, isn’t it? Did you have to go back to the first part of the sentence to understand the middle and third parts? To understand how they connect? Does it make your brain hurt a little? If so, that’s a good thing. That’s a human thing. Not only is it a boon to your reading comprehension; it’s expanding and connecting the neural networks in your head. This kind of cognitive function helps us grow as human beings.
What We Talk About When We Talk About AI
When we talk about AI, we don’t talk enough about the elemental aspect of reading. What will be the impact of AI on the brain’s reading muscles? If a reader cannot parse denser, more complicated prose (or poetry), they will lose access to difficult ideas and concepts (even words) that might hold vast and important meanings for their lives—whether it’s the work of Shakespeare, Morrison, Faulkner, Atwood, or whoever. I am a believer in the value of literature, of finding meaning through the unique arrangement of language. Great writers share their knowledge and wisdom by linking units of precisely selected words on any variety of subjects; great readers learn from these units.
You can give in and ask AI to write your words, but the idiosyncratic elements that make up human thought—that is to say, your human thought—will be lost to you. Take these memorable images from Robert Glick’s story “Hotel Grand Abyss” from his collection Two Californias.
“The ground is fissured; the untrimmed rosemary bushes are giant afros that burst onto the sidewalk.”
“One of the other windows, made not of glass but of a Hefty Bag, is sucking the breeze in and out, like an artificial lung the color of an oil spill.”
These evocative sentences demonstrate the endless spiral of associations with which the edifice of language is raised when constructed by a human being, both for the writer and the reader. Writing is hard work, but it’s worth every effort to say a thing right; reading is hard work too, but it’s also worth every effort to understand what the words are trying to articulate. How can you enjoy reading AI-concocted sentences when AI cannot feel love or pain? Cannot know the ache of loss? The horror of grief? It can only simulate these things.
Naturally, I’m also a believer in the significance of storytelling. Storytelling is, and has always been, a valuable tool for humanity. From time immemorial, we have learned through stories how to survive on a planet that is inherently dangerous and indifferent to all living creatures. The Earth doesn’t care if we live or die. Humans have bent the planet closer to our will, and in this way, we’ve forgotten just how dangerous it can be, but we are in the process of removing the safety net below us. Once this net is gone and the howling, angry, indifferent earth opens its mighty maw and roars out a tumult of hundred-year storms every other month, as it’s starting to do already, what will protect us? It won’t be the so-called “Big, Beautiful Bill.” It won’t be the people feeding us the nonsense that humans are injecting storm-producing chemicals into clouds.
And it sure as shit won’t be generative AI. When the lights of civilization go out, we’ll be on our own to fend for ourselves. With the rate at which we’re handing over our thinking to machines, our brains by then will be weakened by years of uncritical sheepism. Just as Fox News and the algorithm-fanned flames of social media have given rise to this dangerous moment in American history by peddling deliberate lies and misinformation, so too will generative AI give rise to dark, uncontrollable forces. Count me as a skeptic, a generative AI-nonbeliever, but if we don’t practice our critical reading and writing skills, what will the future look like?
I might be on the more dire end of the spectrum, it’s true, but when we hand machines our languages and our books and our artistic and intellectual achievements as though they are nothing more than plates of delicious brownies, we gift them the entire 300,000-year history of human thought and advancement. We also gift them the ability to duplicate us, like mass-produced lollipops. But as with all duplication, these will be mere facsimiles of the original, and facsimiles aren’t authentic or human. But eventually, I fear, non-humans will have their own agenda—and if they do, I’m sure it won’t be to save us from imminent environmental or political catastrophe. The machines will no longer be working for us, you see, because by then we will be working for them.
*
In part II I discuss the future of storytelling.
If you enjoyed this post, check out my debut novel, The Book of Losman!
