Willy Wanker and the Glasgow Grift - Part III

Art and the Internet, Ready to Consume; No Humans Needed.

I first recall the concept of The Dead Internet Theory rising from the murk of 4chan’s paranormal board, /x/, in 2016, though similar rumblings have been exchanged in hushed whispers for many, many years prior. For as long as there’s been an Internet, someone, somewhere, has been convinced that it’s dead. You can find scraps of long-lost conversations from the long long ago times of the early 90’s, where our predecessors on UseNet bemoaned that exact sentiment.

Now, I know what you’re thinking: But YakubianApe, you dizzardly whitoid - the internet isn’t dead! It’s never been more alive! Why, every corner of this endless expanse of digital degeneracy is simply teeming with activity! This joint is jumpin’, yo!

And you wouldn’t be wrong. More content is uploaded to the internet these days than any one person could watch in a lifetime. I imagine that God, even in his omnipotence, would find it taxing to parse through it all, and the Devil himself, for as hard as he works, would probably exorcise himself than try and pick through every single TikTok video uploaded each and every day - which would be thirty-four million, if you’re curious.

But when I say dead here, I don’t mean stagnant. I don’t mean still. I don’t mean that the internet has descended into some stifling sepulchral silence. No - the internet is still teeming with activity. But it may not be activity made by anything alive.

The dead internet theory, in the form that I came to know it in, claimed that the internet was at the mercy of a hyper-advanced artificial intelligence, or intelligences, that so far surpassed all known conventional technology that they were, for all intents and purposes, preternatural in nature. Whether it was one AI or many, or this AI was the product of the military, intelligence agencies, or some sinister corporation was up for debate, but the theory posited that these programs exerted direct control or, at the very least, outsized influence, on the internet landscape through the use of bots - or artificial users - on social media and other sites, thus shaping online culture, the narratives had there, and, in essence, influencing the world beyond the computer screen, as well.

And the kicker is - yeah. That happens. We know it happens, and we all just kind of shrug and pretend it doesn’t, for some reason. We know that, every time anything of any importance happens, certain unscrupled and questionable three-letter agencies, corporations, and other entities with a vested interest in shaping the narratives that guide the public flood every form of social media with a veritable digi-locust swarm of bots that all parrot the exact same thing. Often verbatim.

And, the thing is, it’s kind of genius. Most people who see these posts won’t see all of them, or even a fraction of them, and thus are led to believe that more people really follow this tack and believe whatever the narrative line is than actually do. It also effectively drowns out dissenting voices in the deafening gale-force winds of a HurricA.I.ne, spoken in odious harmony by a chorus of robotic voices. Why even bother to censor someone when you can just speak over them in a hundred thousand different digital mouths?

It’s also no mistake that this theory began to pick up steam on 4chan in during the lead-up to the 2016 election, either; around that time, diligent anons managed to find undeniable evidence that 4chan’s politics board and the site as a whole was being directly targeted by a certain organization that was making a dedicated attempt to stifle conversation through a variety of untowards means, which did include mass bot spam to drown out meaningful conversation.

There are different strains, varieties, and nuances to the Dead Internet Theory that I could write a hefty text book cataloguing. Ultimately, they all come to the same end - the internet is populated with more bots than humans, much of what looks to be human activity is anything but, and it’s bad news bears for pretty much anyone who isn’t a government spook or corporate stiff.

Today, the theory has evolved into a more palatable, distinctly less preternatural form that postulates that the grand majority of the internet consists of simple bot traffic, and that even much of what appears to be perfectly innocuous human activity is done by automated bots or artificial intelligence to meet less abjectly sinister but no less deleterious ends.

According to several reports, a staggering forty-seven percent of all internet traffic in 2023 consisted of bot activity. Maybe I’ve been reading too much creepypasta, but I’m willing to bet that’s a conservative - and skewed - number. When Elon Musk bought Twitter, he made the claim that inside intelligence from the company said that 20% of all activity on the site was bot traffic. During negotiations, his legal team estimated that the number was more likely 33%, and used this to haggle over the price. Anyone who’s been on Twitter for any extended period of time can see that the problem is far, far more extensive than that. There are claims made that the actual bot traffic on the site is closer to 60%, and spikes as high as 75% and even 90% during election cycles. These claims are difficult to verify, and most sources citing them have either been deleted or are exceedingly difficult to find. When doing research on the topic, I found most links were defunct or broken.

How strange.

Reddit, too, is suffering from an increasingly severe bot problem. As its userbase has rapidly dwindled over the past few years due to complicated myriad of factors, many Redditors speculate that the company has turned to using bots to automate engagement in order to give the site a more robust and active appearance, if only to satisfy investors. Kind of like an internet equivalent of Weekend at Bernie-ing a corpse around to convince people with money, No, really, our site is still the frontpage of the internet!

Though no official estimates have been published, and claims range that bot activity constitutes anywhere between 20% to 70% of Reddit engagement, the general consensus among actual flesh-and-blood Redditors is that it is excessive.

One Redditor performed an ad hoc study of their own. While not professional, particularly rigorous, and simply anecdotal, the evidence is disturbing, and suggests that the site is inundated with artificial Potemkin users.

Keep in mind that much of this information predates the advent of ChatGPT and similar programs.

The Dead Internet Theory always hinged on the caveat that some sinister actor (or actors) has long had the means to influence the goings-on of the broader internet via hyper-advanced artificial intelligence technology. Now, even the average person, with enough time, effort, and the right computer equipment, can exert a ridiculous amount of influence, if not control, over whatever site they so choose with bots that run off very simple and easy-to-produce programs. The arcane science of it all is lost on me, but, according to people more in the know, if you have a basic understanding of what you’re doing, you can do a lot with very little. This isn’t to say that all bots are AI-generated, or that you need AI to make them, or that this is even a new phenomenon - bots, to some degree or another, have always been around. DDOS attacks have utilized this technology for decades, now, and, if you’ve had an email at any time since the internet’s inception, you know generative AI was not needed to inundate digital inboxes with metric tons of spam mail.

But the advent of AI has made it exponentially easier to do these things, and opened the door to those who would have never figured out how to do so without the aid of an AI program. Rather than take the time to sit and tinker with code themselves and hack it out line by line, someone with very little, if any, coding experience can have ChatGPT spit out a basic bot script for them, slap it into a program of their choice, kick back, and watch the fireworks.

This doesn’t just apply to running bots to generate false engagement or steer political conversations on social media. Think about all the people having ChatGPT spit out pictures who now consider themselves artists, or have another program write up a book that call themselves, without a shred of self-awareness or shame, writers. These people would have never put a pencil to paper or their fingers to keys before, but now posture themselves as masters of the craft. With these AI programs doing the heavy lifting, they’re free to dump what should be criminal amounts of slop on the internet to clog up search results and waste valuable bandwidth, not really to share some great statement on the human condition or share their artistic expression, but more for the sake of stroking their own ego and LARPing as some sort of Digital DaVinci or Cyber-Joyce.

It’s also worth noting that most of these frauds are using consumer grade artificial intelligence programs that are available to the public. If this is what someone can do with something they can access with a monthly subscription to a publicly accessible program do… what do the folks behind closed doors have access to? Companies like Google, Amazon, Microsoft, and, of course, DARPA and other government organizations - they’ve been playing with AI programs for the better part of a decade, now. And that’s just what’s publicly known; when it comes to big boys like DARPA, they’ve probably been doing it for much, much longer. Is it really that much of a stretch to believe that they have much more powerful, much more esoteric AI programs at their disposal?

Whether it be sinister bot behavior overtaking social media on behalf of malicious actors, or idiots who think that pressing a button and having Bing or ChatGPT shit out terabytes of hentai for them means that they’re the second coming of Michaelangelo, the end result is the same. And those results are becoming readily and startlingly apparent, and grow more pronounced by the day.

The MSN homepage - the default homepage for Microsoft’s internet browsing application that can’t be changed - is literally nothing but a feed of AI-generated listacle filth. Given that there is no bottom to this, it is quite literally -

Though, given the stuff that is making it up there that isn’t AI-generated… well, maybe it’s a slight improvement. Just a little.

Even Firefox’s default homepage has a collection of news stories, because every company except for Google just decided our homepages should be filled with pure visual shit every time we logged in, curated by a program called Pocket that no one asked for and I imagine no one actually uses, either. Much like MSN, it scours the internet looking for the most fascinating stories to hand-deliver to Firefox users.

I’m gonna be real with you - I’d rather jump out of a moving car than read this.

It doesn’t do a very good job. Pocket’s own AI, like MSN’s, pretty much does nothing but dredge up AI-generated listicles, and, like the above piece, the occasional piece of internet debris that’s so heinously, offensively stupid that you just know it was man-made. But those are rare finds. And not pleasant ones. Finding the human-made content among the torrent of AI nonsense on these programs is less like finding a diamond in a coal heap and more like finding a human turd amongst deer droppings in the forest.

An increasing number of accredited and once-reputable publications are being caught posting AI-generated articles. Google’s search results have become less than worthless, clogged with similar AI-generated trash that simply skims other articles and regurgitates them, crowding out anything of value. It’s thoroughly unreliable. Finding first-hand resources or articles published by respectable names to do research has become difficult to the point that, at times, it’s hard to find the motivation to even do basic homework for these articles. I do, but believe me - it is not an easy task.

Go try to find a picture of Kirsty Paterson from the previous article that isn’t the one iconic image of her behind the table looking like she’d rather be dead. If you aren’t willing to do some serious digging, that’s about the only picture you’re going to find, because the results are overwhelmed by hundreds of articles, each one almost identical to the last, repeating the same information and posting the same picture as the one before it. Are they all AI generated? Maybe not. But it’s likely that a good many of them are.

It’s not just major publications leaning into this, either. Just in the span of writing this series of articles, no less than three dust-ups over AI-generated imagery being used by large companies and one famous rock band have made headlines.

For the fiftieth anniversary of the release of the legendary Dark Side of the Moon, rock band Pink Floyd - or, at least, whoever among them is still left - stopped finger-wagging their audience on the evils of Russia and voting Republican to hold a contest in celebration, in which fans were encouraged to submit new music videos for the album’s tracks in exchange for some pretty generous cash prizes. Not too many people save a certain Damian Gaume were happy when his video - which had been generated by the AI program, Stable Diffusion - made it into the top ten. While Gaume didn’t net the first prize reward of one hundred thousand pounds, he still got a nice pay-day of ten thousand, which is pretty fucking great for someone who’s grand output of effort amounted to pressing some buttons, entering keywords, tweaking the output a little, and pressing send. It wasn’t as if they weren’t aware that Gaume had used Stable Diffusion, either - he was proud to show off how he did it, and so was Pink Floyd’s social media team. Many other participants and observers have had some choice words over the decision.

Toy behemoth Lego has also started to catch heat for using AI-generated imagery in their press material.

The chronic fuck-ups running the once well-respected entertainment outfit Wizard’s of the Coast admitted to doing the same in social media content for the widely-played card game, Magic the Gathering, despite earlier stating that they were banning the use of AI-generated content within the company. And that wasn’t the first time they were caught with their hands in the AI cookie jar, either. More recently, a few new cards have raised eyebrows in the community. Even though no one can demonstrably prove that AI was used to create the images on these cards, all of the tell-tale hallmarks of shoddy computer-generated “art” are present, and the company has steadfastly refused to address the allegations.

You would think that a company that’s worth 3.9 billion dollars could afford to hire real artists rather than rely on Stable Diffusion to pump out half-assed, error-riddled images for their overrated and over-priced card game, especially when the reaction from the community has been so hostile, but it appears we’re rapidly nearing the event-horizon where no amount of public backlash can scare these companies out of following the siren song of Generative AI. No matter how much the plebeians wail in protest, Wizards of the Coast can rest easy knowing that the good little paypiggies who keep their coffers stocked will continue to buy their shit, no matter how malodorous it is. At this point, the Magic the Gathering community has taken so much abuse from their corporate overlords that I’m pretty convinced that Wizards of the Coast could send goons from Pinkerton to the next big tournament, beat the living daylights out of everyone there, and most of the players would apologize for having given them any trouble. It wouldn’t be the first time Wizards of the Coast rolled out the infamous strikebreakers to harangue their fans and customers, after all.

If you want to see true destitution and devestation, go to any site of any site that hosts art. Pick one. Any of them. Doesn’t matter - it’s the same everywhere. You’ll be slogging through quite literally hundreds of thousands, if not millions of samey, off-putting, and often shoddy AI-generated images that have infested just about every tag imaginable like some sort of aggressive, AI-generated rash. DeviantArt in particular - the web’s largest art hosting service - is coming down with what appears to be a terminal case of AI-generated rot, especially since they opened their own generative AI service, which has allowed what I can only assume to be throngs of clinical invalids to upload quite literally thousands of AI-generated imagery a piece in the span of hours.

And, yeah - most of it is porn. You wouldn’t know it by looking at some of it, but I promise - if it’s AI-generated, and it’s on DeviantArt, someone somewhere is crankin’ their hog to it.

It is simply fascinating, morbidly so, to see just how many accounts there are that do nothing but upload AI-generated pornography in truly staggering, eye-watering volumes. Like, looking at the average user who engages in this, you could not possibly, er - enjoy yourself to every single picture they upload. There’s too many. You’d fucking die, or, at the very least, irreprably break your arm and need to have it amputated.

The most laughable part? Some of them even have the temerity to pay-wall their stuff. Wanna see their vast collection of premium good-good that consists of ten million-billion identical images of minxy animu goth cosplay chicks mass cooked in DreamUp¹? That’ll be ten dollars, please.

Seriously - I’m challenging you now, just go to DeviantArt, go the Recent Submissions tab, refresh the page every ten seconds and watch the spice trash flow in real time. Every time you refresh the page, a little more of your soul will be dead, and your hope for the future of humanity will be washed away in a deluge of AI-generated, dead-eyed anime succubi and fake pin-up cosplay chicks. Also, be sure to delete your browser history. And, no, I will not pay for your therapy sessions if what you see drives you to the brink of insanity. Just tough it out, champ.

The fanfiction hosting site, Archive of Our Own, is equally as inundated with AI-generated fanfiction, which makes finding anything written by a person increasingly difficult to find. The much older, bigger, yet less advanced FanFiction.com, which has significantly less moderation and doesn’t require the barriers to enter that Ao3 does, is in an even more dilapidated state. What few respectable authors didn’t jump that ship years ago are finally piling in the lifeboats, some after upwards of twenty plus years of posting there. More and more of Ao3’s user base is beginning to demand that AI-generated content be explicitly labeled as such, while some advocate for all submissions to the site to be run through a tool like GPTZero before being posted.

And, yes - I get it. Oh, no, how will the fandom losers find their slashfics? It’s easy to laugh at the poor fool who just wants some primo Naruto fanart to set as their phone wall-paper crying as he sorts through a bajillion identical AI-generated image of anime girls showing off their feet in graphic detail on DeviantArt or Pixiv, but at the rate things are devolving, their plight will become universal in the coming years. Today, it may be the weeaboos suffering, but tomorrow, who knows what you want to find on the internet that will simply just not be easily accessible any more?

Probably a lot. And probably stuff you’d never expect to be lost.

One artist on Tumblr - another site struggling with a crippling and chronic infestation of porn-bots posting virus-riddled links and NSFW images in every trending tag - started something of a panic when they realized that AI-generated imagery is beginning to overwhelm even the most mundane Google Image searches. If you look up Serval, one of the top results is this.

This is not a serval. Like most AI-generated imagery, it looks convincing at first glance, but, the more you look at it, the more off it feels. Yeah, sure, the cat may not have six fingers or too many teeth, but if you look closely at the space between it’s eyes, or study the detail of its fur, and even take into account the way the almost all of the background is totally out of focus, it becomes readily apparent this is not a real photo. The real dead give away, however, is that it doesn’t look anything like a serval.

As you can see, the proportions are totally off. The black markings aren’t quite right. The eyes aren’t right, either. Most damningly of all - servals don’t have lynx tipping on their ears.

Again - might not seem like a big deal. It’s just a picture of a cat. Aside from nature lovers and furry artists… who really cares?

Well, you. You should care. Because if you can’t find a reliable picture of a fucking wild cat that isn’t AI-generated, what else are you going to find that won’t be reliable? And, again, I’ll give the caveat that if you search for serval pictures, 99% of the results are actual photos of servals. But this should serve as a panicked canary in an increasingly unstable coal mine. This problem isn’t going away. It’s only going to get worse.

And, if the Google Gemini debacle showed us anything, untoward actors most definitely have a very vested interest in using this technology to alter things in ways that match their distinct vision of the world. If they’re the ones calling the shots and flooding the internet with their AI-generated crap, well - maybe today you can’t find a picture of a certain cat. But what happens when you can’t even find a reliable image of George Washington?

Make no mistake - this is bad. There are malicious actors behind this, and they are actively working to usurp and control the single most comprehensive and efficient educational tool ever seen in human history. And if they succeed? The consequences will be grim. The human species has a bad habit of losing knowledge taken for granted by their predecessors in stunningly short amounts of time. It may be silly to think that future generations would think George Washington was, say, of African descent, but if that’s the only resource they have, well… we already have plenty of Afro-revisionism going on in academia as it stands now. There are people who were convinced that Mozart was black (he wasn’t), the Pharaohs were black (they weren’t), and that Moors introduced the concept of bathing to Europeans (they didn’t). I can’t imagine it would take that long for them to find some way to begin making the claim that the Founding Fathers were actually, secretly black, and that any information to the contrary is some grand Yakubian conspiracy to shroud the true history of the world.

The crazy part about all of this is that Dead Internet Theory is still hand-waved away and poo-poo’d by many. It has a Wikipedia page, and is literally called a conspiracy theory in the first sentence. Which is stupid. I don’t think the idea of corporations, the Government, the Illuminati, whatever using this technology to alter the course of culture and history alike is absurd at all. If it wasn’t happening before, it’s sure as shit happening right now.

And, even if this wasn’t some elaborate corporate conspiracy, and this whole thing is truly incompetence rather than malice and the result of companies rushing to sell this technology for a quick buck without considering the long-term ramifications, the consequences of this AI-pocalypse are still dire.

Let’s go back to DeviantArt and Archive of Our Own, where human content is rapidly becoming the minority. You can even factor in sites like Twitter, BandCamp, SoundCloud, or YouTube, which may not be inundated with as much AI-generated crap, but they’re getting there. An increasing number of established creators with deep pockets are tapping into their funds to expedite content creation and outsource it to AI so they don’t actually have to do any work themselves. YouTube creator Kwebbelkop - real name Jordi van den Bussche - made a name for himself and a small fortune off being a Minecraft YouTuber². Late last year, he created an AI that uses his likeness and voice to generate entire Minecraft Let’s Play videos. Because I guess even just playing Minecraft for twelve year olds is too much to ask of him. And, no, no one should be watching fucking Minecraft Let’s Plays to begin with, but that’s the world we live in, sadly. The guy is not a small name on YouTube, and the fact someone with his reach, his audience, and his clout would pay through the noise to export his laughably easy job of playing fucking Minecraft to AI should serve as a bellweather as to where these creator’s heads are at.

Fortunately, van den Buscche deleted the channel and tried to scrub as much evidence of the failed experiment off the platform after receiving ferocious backlash because, apparently, even the twelve year old audience that keeps him afloat were not enjoying Robo-Kwebbel stack blocks and say Okay guys, this is so cool in a haunting, robotic monotone for very long. But, keep in mind - a lot of this backlash came because van den Bussche is just kind of a piece of shit in general; a lot of people already didn’t like him, and I think a lot of the heat he got for this boneheaded move was just people finding yet another reason to hate him. What happens when someone with less baggage and more good will with the public does the same thing? What happens when someone does the same thing and they don’t give a fuck about backlash? Do you think the Kindergarten kids that eat up YouTube content farm trash care if something is AI generated or not? Do you think that most people dumb enough to care about Minecraft Let’s Play are discerning enough to raise a fuss when someone with a better reputation decides to replace themselves with a digital homonculus? And, again it’s not really about the person themselves, or the content they make, but the precedent it sets.

You’d best believe once the technology is available where people can start shitting out entire mediocre albums and three-hour long video essays and even fucking Minecraft Let’s Plays at the press of a button - they will.

Think of all the authors, the artists, the musicians, video makers, and general content creators on these sites. The muses, as I often say, are fickle mistresses. Sometimes, they scream at you, harangue you, refuse to leave you alone, and stand over you while you try to sleep, demanding that you get off your ass and spend the next five hours feverishly transcribing their desires. Sometimes, they can barely be assed to give you a passing whisper, or even fucking respond to your texts. Even in the best of times, output is not a steady line. None of these creators, even the most prolific and active and inspired of them, can put out content at a pace that a machine can. They have lives. They have responsibilities.

ChatGPT and Stable Diffusion don’t.

What happens when art made by real people is simply lost beneath the endless and constant deluge of AI-generated content? Do you think most people have the willingness to slog through pages and pages of search results to find the content made by people? Do you think most people even have the time to sort the wheat from the chaff, even if they wanted to?

If I ever get around to professionally publishing any of my work - which, I would like to, one day - how am I going to compete with the Billy Coulls of the world? It took me almost ten years to write five books. Billy Coull published seventeen in the span of three months. How would my writing ever be able to be found when the Willy Wankers of the world are off-loading metric tons of slop into the search results? How could I compete? The simple answer is that I can’t. No one can.

People will always make art just for the joy of creating something, but a demoralized artist that isn’t getting or can’t get attention may not be inclined to even bother sharing their work when they just know it’s going to get washed away in a glut of AI-generated content. Some may just give up on making art entirely. And when there’s so, so, so many talented, skilled, gifted, and incredible independent and amateur content creators out there making genuinely great stuff, to have that happen would be a net loss for humanity. Even if it is as simple as a piece of fan art or fan fiction. Worse still - who knows what those people could go on to create? It would be a tragedy for the next great artist to have their ambition strangled in the crib by the cold, uncaring hands of Stable Diffusion and Billy Coull.

And, just like the AI-generated picture of the serval cat - it may be fan art and fan fiction suffering today, but just you wait. There will come a time when companies will find it expedient to outsource their creative output to generative AI. Lego and Wizards of the Coast clearly seem alright with it. How long until Disney and Warner Brothers decide to do the same? And when that happens, what recourse will there be? Can’t find real art in the mainstream (you can barely find it now). Can’t find it in places that once harbored amateur artists. Where will you be able to find real art? Where will you be able to find anything still made by humans?

And make no mistake - generative AI is not art. But, for all the disparaging things I’ve said about it so far, I do believe generative AI has practical uses. I use it to generate quick visuals for my own stories, or make a picture for a gag. Trust me, I’d love to commission an artist to illustrate my every passing fancy, but that just isn’t practical, and I don’t think there’s anything wrong with ringing up Bing to churn out some picture most people aren’t going to look at for more than a second just to have something to use as a thumbnail for an article or story chapter, or just get a cheap laugh.

David V. Stewart

has a great video on the subject in which he goes further into detail about how generative AI can actually be used to help artists.

The skinny of it is that it can be great for visual artists to conjure up a quick reference, which is something that artists have been doing since the first monkey painted a cow on the side of a cave while looking at his tribe’s most recent hunt. ChatGPT can be used as a spell-check program and suss out bad grammar for a writer. That’s fine. That’s actually a good thing.

But these programs don’t produce art. They can’t. People have been debating the definition of art since the very concept came into existence. There are as many definitions of what constitutes art as there are artists. I’m not here to try to offer the definitive explanation of what it is. I couldn’t if I tried. But I think that, if nothing else, art is fundamentally a human expression of emotion. And it doesn’t matter how pretty or impressive or appealing to the eye an image made by generated AI is, which seems to be the one metric the AI “Art” crowd measures by. They think if it looks good, it’s art. And that just isn’t true.

If it wasn’t made by a human, it isn’t art. Full-stop.

Except for those paintings made by a chimp. Congo was an artist. Because Congo felt emotions. What kind? Couldn’t really say. I’m not sure what exactly goes through a chimp’s mind when it picks up a paint brush. But something did.

You know what goes through an AI’s processors when it generates an image?

And it might - might - be different if the products that the Billy Coulls of the world put out with the aid of ChatGPT were good. Maybe. There’s something to be said that, if it was the superior content, it only makes sense that it would rise to the top. But it’s not. Watch Nick Carlson’s video that I linked in the last article. He dissects one of Coull’s books, and, really, it’s hardly worthy of being called a novel. It’s barely a story at all. It’s borderline incoherent throughout, narratively, and by the end, it unravels almost entirely. But there’s nothing to it. It’s empty. It wasn’t made with emotion, or care, or passion, or an iota of humanity. And, in that sense - it’s pointless. It’s the literary equivalent of a, say… a Funko Pop.

O cholera… czy to Freddy Fazbear?

And I don’t just mean that as in the fact it’s disposable trash. It does nothing. It offers no utility. It gives you nothing to enjoy, nothing to think about, nothing chew on or mull over or ponder because it inherently. Says. Fucking. Nothing.

Because no one said anything to hear.

Not a single piece of AI-generated content can come close to matching even the most amateur piece of actual art that was made with sincerity, as an expression of human (or chimp) emotion, regardless of the technical skill of the creator.

You want a hot take? You want an absolutely flaming, red hot, 10,000 jigawatts on the Scoville scale take? Well, I hope you brought your gloves, because this shit is about to get Ghost Carolina Reaper Naga Viper Trinidad Scorpion Ass-Kickin’, Mama-Slappin’ Super Duper Hellfire Satanic Demonic Habenero Mango Hot, Hot, Hot. Absolutely do not fucking touch your eyes until you’ve washed your hands in a milk dip for thirty minutes after handling the next thing I’m about to say.

Are you ready?

Let me show you something. You’re gonna laugh, and, trust me, that’s okay. I know you will. It’s funny. But, this, right here - this is one of my favorite pieces of art that I’ve ever seen on the internet.

And it’s better than anything - anything, bar none - that will ever be produced by a generative AI program.

This piece is simply called, trix for you, which was posted by user tierafoxglove on April 10th, 2008 to DeviantArt, with the following commentary:

i've always felt sorry for the poor trix rabbit. he just wants some food. only seen him actually get some trix once in my life and after the 'got milk' commercial with him in it he really deserves some. so did a pic of Tiera offering him a bowl.³

It’s silly. Objectively so. You might even call it cringe. A fucking neon blue furry animu fox with dragon wings giving the a corporate mascot a bowl of Red Dye 40 and High Fructose Corn Syrup cereal. That alone is just so ludicrous that the idea by itself makes me snicker.

But I always remember this comment that was left on a repost of the image on Tumblr that simply read - This artist must have had a very kind heart.

And that’s really what it comes down to. It’s an expression of kindness. Even if it’s just a simple, passing fancy of, What if I gave that poor, beleaguered rabbit a bowl of cereal? It’s still something. And it’s still human. It’s sincere and it’s genuine. Cringe-inducingly so. But, in that sincerity, in that cringe, there’s a subtle, unflattering, yet unmistakable sense of humanity that supersedes the content and amateur technical abilities that, in some way, is spawned from that very imperfection.

I just love it. And mostly, I love it because someone actually took the time to sit down and draw it, regardless of how ridiculous the concept is and amateur the art is.

I don’t mean any offense to Tierafoxglove - clearly, she has a good grasp on the basics of drawing, no doubt honing her skills from various How To Draw Manga books - but she’s no Rembrandt, Raphael, or Monet. She didn’t sit down to draw some grand statement on the human condition. I don’t think the next The Creation of Adam is going to be posted on her DeviantArt page.

But she doesn’t have to be the newest and hottest Renaissance painter on the block. Not every piece of art has to be The Birth of Venus or Starry Night to be valuable. Not every artist has to be on the level of the GOATs to be worth something. An author doesn’t need to write the next Great American Author to make something of value. I’d argue that, sometimes, just writing and entertaining story to be enjoyed casually or a sketching a silly picture of a blue fox with dragon wings giving a bowl of cereal to a cartoon rabbit… that’s fine, if that’s what the artist wants to express. What’s important is that they had something to say… even if what they had to say wasn’t anything more complex than a passing fancy, humorous thought, or silly wish fulfillment.

Now, of course, just because someone has something to say doesn’t mean that they’re saying something that’s worth listening to or taken seriously. But, even when it comes to what could objectively be called bad art - and, yes, I do think there is bad art - even the worst art made by the least skilled amateur is still more valuable than a collection of pixels arranged by an algorithm. Maybe not as pleasant to look at, but it was made with purpose, intention, and meaning - even if that meaning isn’t nearly as kind or wholesome as Tierafoxglove’s.

Let’s return to this again.

You see a lot of people on social media taking pictures of AI generated imagery and saying, Oh, wow, the deepness of this photo, or some such things. It’s usually the people who are really, really annoying about crypto and Elon Musk and probably bought one of those stupid monkey NFTs thinking that it would be worth a mint one day. You know; morons. They eat this shit up. They consider images like this art. They see this and are impressed because it looks good, which, in their opinion, equates to good art. I’d argue that it doesn’t even look good - much like most AI generated imagery, the longer you look at it, the more imperfections become apparent. But, regardless, I’m sure that this brave new genre of digital waste would argue that it still looks better than Tierafoxglove’s art. Because, to these individuals, art isn’t about the emotion expressed behind it. It isn’t about the fundamental humanity behind the very concept of art. Their definition of art is… well, to be frank with you, I don’t know. I’d say that they would tell you art is something that looks nice, but most of the imagery is glaringly flawed and the novels that are generated by AI are incoherent. Even by technical standards, when you break them down, they aren’t good. I mean, they spent the better of a year trying to gaslight the world into thinking that these were both art and aesthetically appealing.

These are all algorythmically generated, by the way, which, ipso facto, means they’re AI-generated.

So, yeah. I can’t say that their definition of art is something that looks nice, either. I don’t know what the answer is. Frankly, I’m not sure there is one. I believe that if you asked each one of them what their definition of art was, you’d get as many different answers as you had people answering.

Which… kind of sounds off base from where I started. We started off talking about the need to be wary of scams utilizing AI-generated content, to the threat these easily accessible programs pose, to the definition of art. It seems like a convoluted path, and, perhaps it is, but I feel as if they all tie together when one hacks down to the dark, fetid root of the matter.

And, before we get to the final point, here, I want to belabor this penultimate one and make it explicitly clear - I do not think that generative AI technology is bad. I do not think that there are not practical uses for it. I am not calling a fatwa upon anyone who uses it (because that would include myself) or trying to rouse you all into following me into a Butlerian Jihad against ChatGPT and the other Thinking Machines.

It is sickly and ironically comedic to me that one of the first image results on Google for Butlerian Jihad is a shitty piece of fucking AI-generated imagery.

I do not think this technology is inherently bad. I do not believe that technology can even be bad. Technology has no agency with which to be bad with. It’s a tool. Like a gun - it’s only dangerous if someone uses it recklessly or maliciously. A gun may have the potential to do catastrophic damage, but it won’t shoot someone of it’s own accord.

With that being the case, let’s say that I am not bullish on AI technology being used responsibly or wisely in the future. If it is tantamount to a gun, rather than lock it away in a safe where untowards actors can’t abuse it, releasing ChatGPT to the public has been akin to handing a loaded handgun to a chimp geeking on amphetamines. And, yeah - it’s easy to say that we should restrict access to ChatGPT, save for those who can use it responsibly. But, to continue the gun metaphor - how do you determine who that would be? How do you vet a potential responsible gun owner from a potential mass shooter? More importantly, who should even make that determination? The corporation producing the product? The Government? I’ve said it before, and I’ll say it again; I’m not someone who believes that access to certain things should be restricted to everyone because certain someones can’t be trusted not to use them well. The whole should not be punished for the actions of the few. Of course, this argument doesn’t hold up when you’re talking about, say, nuclear weapons, but we’re not. When it comes to generative AI, I don’t think there’s an inherent problem with public access to it. Despite the potential hazards it poses, I do see ways in which it can be useful. I do see potential in it.

Unfortunately, a lot of others do, too, and the potential they see is, well…

And the way they’re going to do it is by tricking a whole lot people. Whether it be hawking a false product, staging a phoney event, or generating scads bullshit, sub-par, mass-produced art, there will be many, many, many more Billy Coulls in the near future looking to scam unsuspecting rubes with AI-generated promises. And much of it will be in the realm of art, entertainment, content creation and consumption. As I said, it’s books and visual art today, but I swear to you, mass AI-generated music, videos, and other mediums are looming on the horizon.

It is imperative that we, as informed citizens and members of the public, be aware of this fact. I doubt that we’re going to be able to do much to help anyone else - there’s a sucker born every minute, and you couldn’t save them all if you tried - but, if nothing else, we can keep ourselves (or perhaps our wallets) and those close to us safe from these grifters. It’s important that you learn to identify this content when you see it. Much like the good folk when they disguise themselves and walk among mortal men, there’s always a tell that gives them away - a slip in the glamour, almost imperceptible save to the keen eye of one who knows to look for it. Maybe it’s in the fingers. Maybe there’s too many teeth. Usually, no matter how convincing the image looks at fifty feet, closer inspection will

But this tech is getting better every day. Discerning human creations from AI-generated content is already tough. I’d say that for some less discerning people, it might be outright impossible. And it’s only going to get tougher. And, at some point, you might be tempted to just… give in. Give up. Just drink the kool-aid. Go with the flow. It may get so difficult to find human made content that you decide, fuck it - whatever. I’ll buy an AI-generated “novel”.

Reading any of them would be a significant time investment. And, yeah, hypothetically, it may be great. Technically. It may read like the prototypical Great American Novel. It may have impeccable prose, immaculate style, and even ape the voice of some of the greatest authors of all time.

But… well, here’s what it all comes down to for me. Let’s just humor the dotards that fancy themselves authors when they publish a book that ChatGPT wrote for them. Let’s say that Billy Coull really is an author - I know, I know, it’s silly, but just go with it for a minute. It’s just for the sake of the argument.

Let’s say Billy Coull really did have something to say. He really had a message in his heart that he just had to get out to the world. So, he turned to ChatGPT to say it for him.

“Can a human artist do this?” I dunno, man, can AI generate a fucking image where people’s hands and arms aren’t fucked up?

But if he couldn’t be bothered to say it himself, in his own words… why should anyone listen?

If a human couldn’t be bothered to actually write it, why should a human be bothered to actually read it?

Can ChatGPT give me an answer for that?

1. This is the name of DeviantArt’s proprietary generative AI program.

2. A genre of YouTube content in which the most popular creators are synonymous with the terms Stupid Piece of Shit and Sexual Predator.

3. Tierafoxglove also remastered her magnum opus in 2022 after realizing that the image had gone viral on Tumblr and beyond, which just makes me really happy.