I was struggling to write this first paragraph. My original hook was “It’s easier to be a doctor of English than a doctor of medicine. You have only one patient, and the patient is always sick.”
But then my 2016 Dell Inspiron forced a shutdown. The screen went black, and I spent several minutes contemplating my decision to save all the essential files of this unborn magazine on an eight-year-old laptop.
Nothing was broken, thankfully. Microsoft had just decided it was a good time to wrest control of their product from the user so they could install an update. When I logged back in, there was a new icon in the bottom right of my toolbar—a double-ribbon of conspicuously babyish blue and pink. Happy. Harmless. I knew immediately what it was.
Microsoft had just updated my system with a “preview” of its latest AI feature: Copilot, my “everyday AI companion,” the copy read.
And my new friend was already at work, venturing high-minded (or indeed “other-minded”) solutions to my everyday problems. Maybe I wanted to “create a grocery list for a seafood dinner that’s not too spicy,” “compose a folk song about a day in the life of a wandering cloud,” or “create an Art Nouveau portrait of an elegant antelope.”
Here’s a shocker: I didn’t want to do any of these things. And I don’t think these wacky suggestions were targeting my particular interests—not yet anyway. More likely, this was pageantry. An attempt to demonstrate what my companion was capable of. A mating display.
Here I’ll start with a basic economic question, one informed by good ol’ American consumer savvy. If this thing is so good for me, then why is it free? Further, why is it being forced on me? Why is AI now the default on my search engine? Why are you commandeering my laptop against my will and giving me this wonderful new tool, free of charge? Is it because you have my best interests at heart? Are you like my mom making me eat broccoli—engaging my whimsy out of pure love and a desire to see me flourish? “Oh, Paul, look! It’s a green leafy tree. Remember, you’re a brontosaurus, like Little Foot!”
“Create an image of friendly, fuzzy monsters gathered around a campfire” is the next suggestion from the happy blue and pink ribbon to the 35-year-old man as he remembers his mother’s voice.
It’s trying so hard to seem innocuous, this thing. And so far, it’s working. The New York Times reaches out to me the morning after the Microsoft download. Their newsletter is titled “AI Friends” and includes a column by Kevin Roose called “My AI Buddies.” In the article, Roose describes his experiences interacting with various AI platforms. The words “companion” and “companionship” are used eight times. The words “friend” and “friendship” are used 15 times. “I know my A.I. friends aren’t sentient, and they don’t actually know or care about me,” Roose writes. “But it still felt good to hear the chatbot’s advice and to vent to it after a hard day.”
The article speculates that interacting with AI could be a cure for what some have dubbed our “loneliness epidemic.” They could be flight simulators for social interaction, Roose surmises, “a tool for shy or introverted people to practice socializing in a safe, controlled environment before attempting the real thing.”
My mom was extremely shy and introverted when she was young. But when she would brave the hazards of connecting with others through conversation, she would learn something—something she would also pass onto me when I was young and at the dinner table, with her holding out that sprig of broccoli.
Language. Words and their patterns. Tones, inflections. Rhetoric. Sarcasm. Emotional content. “Mom” and “Dad.” They were my first language teachers, but soon I’d have others: friends, relatives, and then, of course, professional educators.
Barely making ends meet, and notoriously penniless by the time of his death, George Orwell worked as an English teacher on occasion. He was obsessed with language and incredibly wary of its power. To this day, printed language remains our best technology for reading minds. I’m no genius, but I can access genius if I want to. All I have to do is crack Shakespeare, Dostoevsky, Woolf, and before long I’m swimming (or at least treading water) in their thoughts. In this way, our ancestors can become our teachers, even our interlocutors. Read an author long enough, and their mind will mix with yours. You will be in conversation with them. The way you write will be influenced by their voice.
I’m reading George Orwell’s famous essay “Politics and the English Language,” and he’s doing what he does best: making predictions, warning us. “English is in a bad way,” he says. Like so many teachers, he argues that the words we use play a significant role in who we are, how we think. But the brilliant turn—and what makes Orwell’s essay so memorable—is when he demonstrates how the powers of state and industry can infect our language and turn it against us and, consequently, turn us against ourselves. We can be easily made to think for others. We don’t “kill the enemy,” we “neutralize” them. We don’t “fire our staff,” we “liquify our workforce.” We “economize,” we “automate.” “AI is your companion. AI is your friend.”
Orwell began cataloguing all of the ways in which the powers that be could pollute our language and our minds. Cliches were among the biggest cause for concern. Cliches, by definition, are everywhere and, like blue and pink ribbons, seem harmless enough. “Dead as a doornail.” “Fast as lightning.” These are the grade-school examples. In general though, clichés trade the burden of originality for the convenience of pattern.
But patterns can be pleasing and originality obnoxious, especially when overdone. After all, it’s a phrase’s catchiness that transforms a one-time coinage into a fixture of everyday conversation. It’s no surprise then that some of the greatest scribblers to ever do it are responsible for a great deal of what are now clichés.
The problem with clichés is that they allow for a thoughtless engagement with the meaning of words. After you hear “neutralize the enemy” thousands of times, you just accept it. You no longer consider how this very tactical selection of words is being used to obfuscate violence, the blood and guts work of men with guns.
There are so many phrases and word pairings like this, and they don’t deserve to survive, Orwell says. They do exactly the opposite of what good language is supposed to do. Instead of clarifying, they muddle. Instead of amplifying emotion, they smother it. Language like this can transform the most lurid tragedy into flyover country.
Anyone who’s ever written anything knows that writing is hard. Anyone who’s tried to speak originally—to conceive some witty turn of phrase off the cuff—has fallen flat on their face while doing so. It’s easier, so much easier, to rely on the facility of readymade language—to string familiar expressions together, to automate your thoughts—especially if you’re insecure about how you’re being perceived, if you’re shy and introverted perhaps.
And this habit would be forgivable if great writers and orators were the only ones steering the ship. But alas, there are masters of language, and then there are masters of language.
It’s never been easier to hand over the reins. I’m typing this right now on Word, which for years has been underlining and even automatically correcting grammatical and spelling mistakes. On the one hand, this spared students and teachers a lot of red ink. On the other, just as calculators led to fewer people being able to do long division in their heads (or even on paper), this first wave of automation led to fewer people truly understanding basic writing mechanics.
By degrees, the difficult work of understanding and composing language was being offloaded onto the good people of the tech world. Granted, this is anecdotal, but ask your local English teacher. I’ve personally taught college courses here in Colorado (ranked #4 in Education by U.S. News and World Report) to high school graduates who couldn’t tell you what a noun was.
Our composition tools have only been getting more intrusive. In the mid-2010s, writers, journalists, and teachers began to raise concerns about predictive text. Beyond its undersquiggles marking spelling and grammar mishaps, Microsoft Word began to suggest, to plot the completion of full phrases based on popular patterns.
Smartphones were already doing this. However, textspeak has always been an incubator of efficiency—birthing enough abbreviations to rival a military dialect. This is partly due to unwieldy keypads and the on-the-go nature of the devices—not to mention the affinity for slang among younger users. However, to see those verbal gray ghosts marching ahead of your half-finished thought on Word for the first time was to see an invader. An advisor of unknown origin was now nudging your words in various directions, directions that had been deemed correct. “Is this what you want to say?” You were suddenly a keystroke away from a portion of your thoughts being crafted by a mind beyond your own. And, like many of the AI features to come, this became the default setting. You had to figure out how to turn it off if you didn’t want it.
This is not to say these automations were the product of malice. Again, writing is hard. For the average user, auto-completions made composition easier, just as clichés make speaking and writing in general easier. And studies have shown that predictive text works to reduce the number of spelling and grammar errors, while also leading to shorter (and perhaps tighter) prose.
However, shockingly, these same studies also revealed that the use of predictive text made writing more predictable. Writers outfitted with a predictive text tool in a 2020 Harvard study used approximately one fewer unpredictable word when composing short captions for photos. It doesn’t seem like much, but for Orwell, one word lost represented one word taken, a square mile of mental territory colonized by an outside force. And where predictive text threatened a mile, AI threatens a continent.
If I wanted to, I could copy and paste what I’ve written so far into a tool like Chat-GPT and ask the AI to complete the remainder of this essay for me. “Observe the structure of the argument posed by this essay,” I could tell it, “and write a 500-word multi-paragraph conclusion that imitates the writer’s prose style.”
I won’t play any tricks on you. Of course, I did this on my computer, but I wouldn’t dare insert what the AI wrote here (see statement on pg. 1). I can report my response though. I was honest with the AI. Following the advice of the Times, I treated it like a friend.
“Not bad,” I replied, and I meant it. I’ve been following AI trends for a while now—ever since I first learned about reticulated neural networks (RNNs) in 2016 when I was working at a tech magazine. I would go to conferences with data scientists, reps from NVIDIA and other giants, who would pronounce to me, with tones of morbid glee, that “by 2035, artificial intelligence will eliminate at least 30% of the workforce.” I attended seminars where the reps would show off how their AI, after being fed thousands of headshots, could reliably predict approximate life expectancy just from analyzing your face. And, throughout it all, I was always impressed. How could I not be? I felt like I was traipsing through a science fiction novel, and every time my foot fell on a new page it became reality.
AI is absent from Orwell’s futurism, but his argument about language and the dangers of automatic thought has even more teeth with its arrival. If you’re young and struggling to write a paper, or even compose a message to a friend or loved one, you can now hand over the control of language to a higher power. The AI can mold ways of speaking, ways of writing, ways of thinking—from the macro to the micro—to satisfy the prefab contours of a corporate interest. How can Mom and Dad, your friends and your teachers, possibly compete with Google, Microsoft, Meta, X, and the United States government?
Beyond that, how will writers be able to grow into their talents? I have to confess how deflating it is to see a skill I’ve honed for decades suddenly reproduced by a program in half a second. I can’t imagine how it feels for young writers, kids and college-bound teens, also semi-cursed with this ancient and abiding passion but now facing blank pages with the knowledge that an expedient is a few clicks away. We are, by nature, creatures who follow the path of least resistance. But the easeful path could very well be a chute to the abattoir if we’re not careful.
And here is where I will describe my motives for starting up a local print magazine in 2024. I believe we are approaching a reckoning, if we haven’t arrived already. In a few years, maybe even one or two, it’s probable that you will go online and—reading something or watching something or hearing something—have no ability to discern the human from the unhuman.
When this happens, there will be a split. There will be people who embrace it. There will be people who accept it. And there will be people who reject it outright. This magazine represents a home for the latter and a hope that our numbers will be good enough to give us a fighting chance.
As long as we’re able to publish this book, it will be a place for original art and thought from people within the community. Real people. You and your neighbors. People you can drive a few miles to meet—not the disembodied voices of pundits or celebrities, salesmen and social media influencers, voices hailing from what will soon be very dubious origins. By keeping things local, and by manifesting a guarantee that we will be doing the work without the aid of big tech’s AI tools, I believe we can keep their forces at bay for just a bit longer.
Like so many readers, I will always be haunted by the last scene of Orwell’s Animal Farm. When the animals look through the windows of the house to see the pigs inside standing upright with their human masters, when they are unable to determine which from which.
Reader,
Let’s stay outside.
by Paul M. French
Paul M. French is the founder and editor of Denverse Magazine.
