by TJ Price
1.
It’s no secret that every society thinks its days are the last days. “But there’s something different about these days,” each cycle of humanity insists, and circumstances provide the breeding ground for our justifications in believing such. In these days, there’s unprecedented amounts of strife and calamity, as evidenced by the ever-more-definite probability of climatological oblivion and global political unrest with the rank scent of war in the wind—not to mention that the lower classes don’t trust the upper classes, but now the upper classes have learned how to hide better from the guillotine.
We are all distracted from what matters, if we’re to believe the headlines and the studies, by inessential, vain pursuits. We spend hours of our lives moving our thumbs up and down over screens, or else sitting in front of one kind or another. Now, there’s the threat of Large Language Models—so-called “AI”—to contend with, too, which routinely and with ever more speed replaces us on a functional level. Though it is fairly clear this new technology is not universally reliable, it perhaps says something about us that we are so eager to embrace even a flawed substitute for human work.
I’m with Wordsworth when he says, in his now-famous sonnet,
“The world is too much with us; late and soon,
Getting and spending, we lay waste our powers;—
Little we see in Nature that is ours:
We have given our hearts away, a sordid boon!”
Now we do not only give our hearts away, but also our brains, perhaps even (to wax dramatic) our souls. We seem determined to shed ourselves in favor of a desiderative mode. I think that we are tired, all of us—exhausted. We have sliced ourselves to ribbons on the cutting edge. I see the signs and symptoms all around, from a slackening of empathy and a rise in frustration with the systems in place to outright violence and sanguine disregard for the lives or wellbeing of others. Society has never been a place where one can call oneself totally safe, but now it has become necessary to protect oneself in addition.
We shoulder carapace on top of exoskeleton. We develop weapons fashioned from whatever sharp edges we can. Knives. Keys spiking out between fist-clenched fingers. Words. Anything to threaten, to ward off.
To empathize is to reveal, and to reveal is to be made vulnerable. So down comes the portcullis. Up goes the bridge. The alligators snap in the moat, their tails frothing the water.
2.
I read an article yesterday concerning recent tectonic developments in the Neapolitan region of Italy. This area, known as the Phlegraean Fields, is a large caldera with a rating of seven on the eight-level volcanic explosivity index (VEI). This article was fed to me by the algorithm which populates my Google Chrome app’s homepage, and which I spend time scrolling through idly. The article’s title was as follows:
“It Could Plunge the Planet Into Chaos”
The Phlegraean Fields Supervolcano Is Deeply Worrying Scientists.
Below this dramatic heading: a small box, indicating a tag—ostensibly, for grouping purposes. The tag: “HOME IMPROVEMENT.”
This article was written for farmingdale-observer.com, by someone named Bob Rubila, who, as far as I can tell, is not a real person. At the bottom of the page, there is a link to his LinkedIn page, but the link is broken and does not exist. The “About Us” page of this website lists three names as staff: Bob Rubila, Dave G. Rub, and Rose Dixon. The entire site appears to be nothing but a platform for advertising: both sides of the page are narrowed by vertical ads, every other paragraph is interrupted by an ad—most of which are distracting videos—and the article itself is written in a flat, cheerless syntax. A few links exist within the article to give the illusion of depth, but when one is clicked, it only leads to another page of similarly-tagged articles under the incongruous headings of “TECHNOLOGY” or “GARDEN.”
farmingdale-observer.com’s About Us page reveals little about its scope or provenance. The first sentence says “Welcome to the Farmingdale Observer, your premier local journal dedicated to providing reliable and unbiased information on home improvement, gardening, and now, local news.” It goes on to say that it was founded in 1963, and is committed to serving the Farmingdale community. It does not reveal in which state, county, or even country, this mysterious “Farmingdale” is located. Further research indicates that there is, indeed, an actual Observer that was founded in 1963 (and which dutifully serves the greater Farmingdale, Bethpage, and Melville areas of New York state) but farmingdale-observer.com is not that. Further, it appears that there is an actual journalist out there named Dave Gil de Rubio, who does have a working LinkedIn profile, but nowhere therein is it listed that he is the Senior Director of Farmingdale-observer.com. To the best of my knowledge, there is no Rose Dixon.
My point is this: the algorithm fed me an article that it knew I would have interest in reading. I have read articles before on similarly worrying catastrophes lurking in the wings: tsunami, earthquakes off the Oregon coast, Yellowstone, the rapid desalinization and subsequent collapse of the AMOC, et cetera. But this is quite clearly an AI-generated site, a “lorem ipsum” of aggregated information patchworked together as filler, the binding agent for a farrago of advertising. Whether or not there is human intent behind this littering of the so-called “information superhighway,” it’s a kind of evil, in my opinion. The dissemination of unreliable information under the pretense of “news” is not only insulting, but dangerous.
And here we are, happily plugging ourselves into a machine that slowly sips us dry. Generating images that edge closer and closer to not just a simulation of reality, but which are indistinguishable from reality. Hell, most of the time those who use AI for imagery or art or text are happy to settle for what is clearly machine-generated, even proudly labeling it as such with captions like “Image generated by ChatGPT.”
3.
It’s easy to use AI. The interface is deceptive, and seemingly boundless. A few keystrokes to compose a prompt or ask the machine a question, and text or images flow onto the screen as if conjured by magic. Renowned science fiction author Arthur C. Clarke wrote extensively on possible futures, including a well-quoted adage which states that “Any sufficiently advanced technology is indistinguishable from magic.” It’s important to note here that the two components of this axiom are “technology” and “magic.” Whereas the former is based in empirical data and invention, the latter is based in mysticism and awe. An easier way to state Clarke’s third law, though, might be to say that any technology, no matter how simple, is magic to those who do not understand how it works.
Spinoza, the famous 17th century Portuguese philosopher, spoke of the concept of “wonder,” or “admiratio,” as a kind of stalling of the mind, creating an absence of emotion when we perceive the singularly unfamiliar or inexplicable, and it is this sense of wonder that I would argue plagues our increasingly fraught relationship with AI. It is a kind of blindness to its working that allows us to continue using it—whether to supplant function, analysis, or even the generation of art (such as it is)—despite being largely unfamiliar with its process.
Like many others, I have heard about the outrageous price tag associated with this new technology. I’m not talking about the monthly subscription fee just to be able to use it, though—I’m speaking of the energy levels it demands in order to be operable.
In a disturbing article at the MIT Technology Review, it is revealed that it’s actually impossible to know exactly how much energy AI needs to consume in order to spit out its half-baked prompt responses (or even cover art for highly-regarded publishing houses.) This is because leading AI companies and data center operators simply refuse to disclose the information necessary to determine usage—most are known as “closed” models, which guard their statistics closely, claiming they fall under the aegis of “trade secrets.” Despite all this, the MIT Technology Review article was able to make projections to estimate the impact AI has on emissions and energy, and the answers are frightening even at their lowest.
4.
When I was a kid, I used to fantasize that I would be a survivor when the world came to calamity; that I would live to be the scribe of the ragged group of a blessed few charged with continuing past the collapse of civilization. My unique task, of course, was chronicle this new era. When we are that age, we all think we are the golden child. The hero of our own story. After all, without us, what story would there be to tell? This solipsistic view of the world extends sometimes beyond the egocentrism of youth. When we leave the room, those remaining within wink out of existence. They are not matter, they do not matter. That which is not there cannot be taken into consideration—we don’t think of our children when they have yet to blink at our furious sun. We defer, we let our “next of kin” be saddled with the increasingly high electric bill. (I’m afraid my generation’s bill may have been stamped with the ominous phrase “FINAL NOTICE.”)
In grade school, they taught us catchy songs to help make us aware of the crisis our planet—and, by association, our species, perhaps even all life we know of—faces. I still remember singing “R-E-C-Y-C-L-E!” while clapping my hands and swaying back and forth. Decades later, a particularly cynical roommate of mine explained why she did not engage in the practice—“It’s all bullshit,” she said, confidently. “If it was a good solution to the problem, someone somewhere would be making money off of it.” At the time, I felt like I believed her, but then I was only casually invested in making sure I separated the plastics and aluminum and paper from the rest of the detritus doomed to pile up in a landfill somewhere, another festering mountain accreting on the surface of the Earth.
What is there to do? The efforts of one seem paltry when compared to the indifference of so many others. The frustration is rampant. The streets are overflowing with garbage. There are more unhoused now than there have ever been. I can’t say for sure if this is true, though, mostly because I’m afraid if I look it up, its provenance will also be of dubious authorship, and the statistics given unsupported by citation.
Just yesterday, I asked a coworker a question, and in response, they automatically (as if programmed to do so) whipped out their phone from a pocket and typed the query into Google. Confidently, they read off what appeared on the screen. It didn’t sound right to me, so I leaned in to check. They had, in fact, been reeling off the information from the “AI Overview,” (marked only as such by a tiny title and a blue star) and the facts they’d relayed to me were incorrect, gathered from a website link that didn’t work when we clicked on the source.
Is this the same star that once winked in the sky over Bethlehem? Is this the dropped pin on the map, toward which the rough beast slouches, its hour now come at last—or is this just another turn of Yeats’ ever-widening gyre?
Maybe I’ll ask ChatGPT. Surely, it can tell me what blood-dimmed tide is loosed, what ceremony of innocence has been drowned, right? Surely it can see its own face, even without the use of a mirror. Another trick, to stun and amaze.
Such magic, that leaves its audience gasping for breath! Not for wonder alone, but— cruelly and inevitably instead—lack of breathable oxygen.
***
Enjoying the content on 3QD? Help keep us going by donating now.