Why Ed Zitron is Wrong: An Honest Reflection on AI
Infohazard Alert ⚠ wee-ooo wee-ooo Infohazard Alert ⚠this article contains known infohazards: read at your own peril
In the weeks after my Facial Feminization Surgery, I had what could charitably be described as ‘vivid dreams’, and more accurately as ‘crazy hallucinations’. As my pained, partially nerve-dead facial tissue healed across surgically-altered bones, my fragile mind vanished: vanquished yet vivified into an entropy of pure, unbridled creative energy. I escaped, in other words, from the concerns of our present mortal world and became lost to time and space, my mind playing host to intense, vivid imagery of unimaginable splendor. What I saw and experienced felt as real as reality itself—a vivid dream unlike any other.
Now: the culprit is plain and known. We caught her in the act, dead to rights. You see, I’m a silly gal. I did a goof, an oopsie, a bad time. I won’t bore you with the details, but I had accidentally taken two different NSAID pain-killing medications at the same time, and had been doing so for quite a while. In case you don’t know: if you combine those, you will die. Not quickly and not instantly, but eventually, and slowly, your body will shred itself quite viscerally. My mind was certainly going.
So I erred, so I dreamt, for nearly a week. As soon as I realized my folly, of course I stopped double-fisting NSAID painkillers and just like that: the visions went away. Yet, despite all that I saw, despite knowing it was not real, it still shook me to my core. I’ve been spending the past several weeks piecing it all together, trying to find some deeper meaning to all of it.
Let me be clear. While I do love to say to my friends and any who will listen: “thus the ancients did write: that dreams are prophecies” (and sure enough, yes many, many ancient civilizations did hold that certain dreams could be future portents), my delusions were of far more medical make. Yet what I saw, what I experienced, made me realize that perhaps there was a grain of raw truth in the mounds of existential footage I experienced.
My thesis is such: AI can make such fantastical worlds, such dreams a reality, and we are seeing those roots form in the here and now. This means that AI is not a bubble, which would indicate that there are real economic use cases for this technology, beyond the paucity of chatbot therapists. As I’ll get into: that’s exactly what we are seeing right now.
Let me state this as loudly as I can on a screen:
I despise AI.
It robs creators (artists, writers, directors, actors, et cetera et cetera) of the fruits of their labor; it robs them of their property, of their ability to receive rents from property ownership. While I am not sold on the merits of private property, absent some socio-political revolution, this is not abolition so much as it is fraudulent theft.
I literally worked in an AI team! I know about the theft because I helped do the stealing. For years, I toiled at the [name redacted] Company in the late 10s, where I spent countless hours copying and pasting text into our data library (so: books, articles, research journals, anything anyone ever wrote down) I could get my hands on to build an early sibling of ChatGPT. I could have been a Sam Altman! Evil billionaire woman! But I didn’t have rich parents who could give me an interest free loan, so when the pandemic hit, off I was to unemployment instead of super stardom.
So believe me when I say I get the creator backlash. What I did was evil, immoral, and I regret it deeply. I don’t know if any actual laws were violated as I am no lawyer (#blessed), but I will admit that some moral code was wronged, such as may exist. I stole, and got promoted for it. The company I worked for made billions off my (and my colleague’s) labors. We all juiced and squeezed it out of the hard work of writers everywhere.
Such is the world we live in: trillions spent on acidic death showers that maim and burn babies alive abroad by the millions while hundreds of thousands starve on our own streets every year. That the hard work of artists is sacrificed at the altar of Capital should be no surprise: we are an apocalyptic death cult, less Rome on the Hills and more Thulsa Doom’s tyrannical regime in Conan the Barbarian (1982).
I used to be in Ed Zitron’s AI denialism cohort. Like him, I believed that AI was going to be an Icarus or some other ill-contrived metaphor; basically: that AI was an investor bubble doomed to pop. Microsoft, Amazon, Apple would all fall. Meta was screwed (and so deep in the metaverse they likely would never realize it!). And he’s probably right on this front! These companies are doomed, or approaching it.
But the last few weeks have demonstrated a few things that signify that AI is here to stay, even if every single AI-based company in existence right now were to be shuttered.
First, I wish to challenge Zitron on his base assumption. That there is no possible, profitable business use case for AI. For years, I’ve agreed with him on this. For good reason, too. Shrimp Jesus amused me for 15 seconds but you couldn’t base an entire economy around that.
And then, as if out of nowhere, yet very predictably: a viral phenomenon struck the very heart of the viral economy: AI generated Studio Ghibli art became utilized by a heretofore unseen contingent of the world population. The exact stats are dispersed across various AI models (and who knows if you can trust their data) but the real world evidence is plain: everyone saw these. Some even made their own.
Everyone knows what this is, and what happened, because it was so ubiquitous, but I’m still going to demonstrate my point.
None of this, to be clear, is art. It’s likely it never will. Just look at this.
I shouldn’t need to tell you that this is beautiful art but I will: this is actual beautiful art. It was made with the care and attention to detail only a master in the craft can provide. I think @dieworkwear’s (on X) own commentary on AI said it best.
He goes on to say how much AI resembles trends in the fashion industry. Fashion after all used to be bespoke and unique, with actual guilds and apprenticeships driving production. Yet when the industrial revolution arrived in the late 18th century, the first thing it did was make the act of copying every single piece of clothing trivial and cheap. Of course, bespoke tailors still exist, but this mass copying attribute gave Britain’s burgeoning loom industry a reason to exist, a use case, and thereafter it spawned the industrial titans that dominate today’s landscape (Adidas and Nike for shoes, for instance; or Prada for handbags).
And here we have our first use case in the shoddy copies of better art: it’s more than marketing, it’s propaganda, as I hope the amount of images governments have used and are using illustrates. Of course, there is also a mass marketing appeal to this. Rather than hiring an illustrator, or even teams of illustrators to capture trends and bottle them up into advertisements, you can just spend 5 minutes typing away into a generator. And there you have it: both a government propaganda and a business-to-consumer use case, both, just from one little event. To state the plainly obvious: Studio Ghibli is only the beginning. No franchise is safe, either now or in the future. And because consumers themselves can contribute by spending 5 minutes or less on a generator, we will see true organic viral trends appear, like tech did for video streaming and YouTube in 2006. We did, after all, just see the birth of the first mass viral AI moment with this Ghibli nonsense.
The sycophantic therapist chatbot was never going to be a good profit model. Nor were AI planners or “productivity boosters”. It was always going to be about mass art production, the act of rendering bespoke artists if not obsolete than definitely less immediately useful. But then again, it’s not as if there were a plethora of artists out there to begin with who could’ve handled the demand for Ghibli art, nor were people likely to all of a sudden been willing to pay exorbitant prices for such art generation. It was a moment of lightning in a bottle, in a way that the real world never would have replicated in an AI-free environment.
That being said, I do want to state as clear as possible: I don’t think this is also the death of the artist. Recall when I said that I spent years working in AI? Well, don’t even take my word for it: just look up model degradation or even just use AI yourself. It gets dumber! Always, every time, without fail. This is because without a constant stream of new data: new books, new art, new everything (that must be human-made), the AI will just iterate and copy itself to death.
AI models need ongoing human input to stay effective and profitable. This is not a defense of the system, I assure you: merely a fact. For AI is at the end of the day a printing press that can self-iterate on things it has copied before.
So that’s the first instance I saw of AI having potential actual use. The second is just a far simpler story to tell: Chinese researchers built a cheaper (both cost and energy-wise) open-source model (DeepSeek). They, and others around the world, will continue to do so. Unlike social media, which built closed-off ecosystems of access, the openness of AI source models will mean it will continue to blossom into a truly multiple-source digital ecosystem. It is akin basically to Linux, but there are an infinite number of versions constantly being made by practically everyone. It would likewise take a titan like Microsoft to outspend and outcompete the free models to achieve market dominance (just like they did with the personal computing OS). Yet because the US equities market is self-imploding due to Trump’s economic suicide-by-tariff: that is exactly what won’t happen. Microsoft will instead die a death by a thousand cuts, and a million other AI models will blossom in its wake.
The third and last thing I saw were, of course, my visions, my hallucinations—my whatever appellation you want to call them. The question it raised was: what if? What if these copiers became advanced enough to generate entire digital worlds—maybe decades from now, maybe never, but still: what if? An infinite creation engine, churning vivid dreams out en masse to a population hungry for some type of escape from the crushing brutalities of reality: a true digital Heaven, made possible on Earth.
And in this Heaven, I saw my parents, still living yet long estranged, standing before me, talking to me. I saw my grandparents, long dead yet still yearned for doing likewise. I saw my mother’s young adulthood, of her rushing to catch the bus before it drove away. I followed her, the familiar streets of my own childhood recalled from memory. All of it: so real, so present, so tasteful and promising.
I saw myself, as my real idealized version of my Girl Boss self, take center stage in several video game experiences. Yes: they were video games but they still felt so real, so enrapturing, so all-encompassing, that the word ‘haptic feedback’ burned into my memory. Total immersion, made possible by vertical technological improvements across the industry including in the VR space. Think those fancy VR headsets but ones which grounded all six (or is it nine?) of your senses into a digital world. The Wachowski Sisters’ world made reality!
In one game, I teamed up with a team of five friends and Mrs Frazzle from the Magic School Bus to go treasure hunting across far away lands from the deserts of Egypt to the jungles of Brazil in the early 20th Century, with combat resembling something like Overwatch or TF2. I got stuck against a miniboss, an evil creature that sniped me and my friends before we could deal damage, yet I was able to summon Johnny Silverhand from the Cyberpunk universe to help me out.
In another, I was in a post apocalyptical, toxic shock swamp world. It was a Playstation Original, part-platformer, part-shooter, part-RPG, with art styles very reminiscent of Netflix’s Arcane. Think: an open world Ratchet and Clank, or a Zelda game, all meshed together. I built my own (sailing and motor) ship, and with companion NPCs, we ventured out into the world filled with toxic rot, to see if we could heal and better the world; a long 100 hour action adventure extravaganza awaited.
I even made and got to play a sequel to Cyberpunk 2077: an elaborate train mission, with hundreds of unique NPCs and objectives, trailing for miles on end, each with unique backstories, backgrounds, and motivations.
My second favorite dream was of an astronaut simulation game with real life gravity effects. The game was whatever but something amazing happened: trans people are nothing if not resourceful, and we figured out that thanks to the gravity simulator, we modded the game to build the world’s first full immersion mass e-sex simulator (if you can simulate zero G physically in the digital world, turns out people will just use that to immediately have sex instead of going out into space and pretending to be an astronaut). I remember having gay, tremendous amazing sex in zero G, floating and giggling away with my friends and strangers in the carefree distances of digital space. Even when the terrors came, they did not keep me away. For the Neo-Nazis found us. They hacked our servers to at first simulate the feeling of dying (I had my own head cut off! Again! And again! And again! It hurt.), then to beam white supremacist propaganda straight to our neuro-visual cortex, with horrid imagery of death and decay on constant blast. It still was not enough to keep us out. We fought back, and won, mostly though in real life (at least: the real life that I thought I was real).
It wasn’t just games. I saw Back to the Future: many, many, many Backs to the Future, with all the same script, lighting, and direction but with different actors each. Val Kilmer plays a killer Marty, perhaps the best of all on par with Michael Fox’s original portrayal, with Doc Brown’s depictions from a wide variety of actors ranging from Whoopi Goldberg to Ian McKellan to Hunter Schafer providing crucial backup. And having seen it: I really do want to see a version of Back to the Future but every actor is Will Smith.
And the dreams increased.
I dreamt of the future, of myself owning casinos and conducting drug deals in them across the globe.
I dreamt I could astral project, a sort of real-time drone-fly across the city.
I dreamt I was Melfi’s actress from The Sopranos and how David Chase had me direct a scene and I did a good job, but the whole crew gave me the cold shoulder despite that.
I even ventured beyond movies, beyond games, beyond reality, really. I crafted my own world, where I was a secret UN government agent designed to keep the spacetime continuum from collapsing in on itself. In that role, I battled against a metafictional Jeremy Irons, the publishing industry, and my own eternal demons to rise to—well, I won’t spoil it.
It would be so easy to dismiss this all as delusions of a fanciful mind. That: yes, I poisoned myself, I had delusions, sometimes of grandeur yet often of mediocre vulgarity. And for the most part, I did. I don’t honestly think we can or should build a Back to the Future As A Service movie industry. At best, I thought about writing a (non-AI) book about the grand old tale about me hunting the metafictional Jeremy Irons, for the story was extremely unique and hilarious when I awoke late one morning to write down all I had seen. Thank you me, for poisoning me: and I was well enough to leave it at that. Yet the world kept turning, and what I had seen stayed with me, long enough to see it all change.
I just kept thinking: what if? What if that type of world: where whole universes could be created, whole genres, whole games, whole entire modes of existing online could be willed into existence with the snap of a finger, on a mere whim? An infinite probability generator, in other words: a personal TARDIS for every person who could afford it? It tantalizes, I will admit.
I remain convinced: I don’t think AI is profitable, at least in its current iteration. I also want to stress that I don’t think AI is good, and I don’t think it ever will be. I don’t even think we should proceed in an attempt to find profit or virtue; I am instead reminded of the first episode of The Last of US on HBO Max, where the doctor character says simply in response to the release of the deadly cordyceps parasite: “Bomb. Bomb the city.” I agree!
Bomb.
Bomb the data centers.
And everyone in them.
Kidding. Of course. I’d prefer to legislate but in lieu of legislation, one can’t help but feel that a quick and dirty solution is preferable.
I don’t think AI will kill us all, at least not in the traditional sense, with Terminator robots and nuclear war. But as the chatbot therapist scandal has shown (where shoddy AIs prop up & expand people’s delusions and mental health issues), people are vastly too stupid to use this “intelligent” system intelligently. “@gork is this true” has become a stale joke by this point but even so it demonstrates how reliant we are on AI as dispensers of truth or some intelligence. One thing I want to state absolutely clearly: Artificial Intelligence is a misnomer. It is a lie, an attempt to do some less-than-subtle marketing by tapping into viral media trends about AI. Everything we can call AI can instead best be summed up as “Machine Learned Copier”. That’s all this is.
In fact, there should be a law that says that you cannot call something, or really anything AI, that you must call it a Machine Learned Copier or some other queer name. It is artificial but it has zero intelligence, for it is not alive. It’s a virus: it is capable of replicating itself and that is it. Even if does get smarter, we should not turn off our ability to critically reason. As psychologists the world over will tell you: it is about trust. If you put your trust in a system, and that system hallucinates, or breaks down (as software is wont to do often): then you could end up doing things that you, a thinking person, would never otherwise do. Like murder your entire family. Or stop taking your meds. Or launching your nuclear arsenal at Russia and China. “Gronk told you to do it” is less a defense and more a condemnation of you.
And that’s it. Just this one simple change would save humanity. But that’s thinking small, isn’t it? So let’s not stop there!
Pass more laws:
Pay the artist class what they are owed for helping build data models.
Keep them employed, and employ more of them to keep the data models fresh and ever growing.
Learn to Code should be replaced with Learn to Draw. Learn to Write. Learn to Speak. Learn the non-STEMS, for our economy needs millions more of those.
Be open to using cheaper, better models from overseas (removing trade barriers on chips and other tech sectors: we sprint together or not at all as a species).
Keep transitioning to a non-fossil fuel economy.
But if none of these laws are made, if people do keep calling it AI, do keep relying on it as a soothsayer then yes: let loose a thousand bombs—or in fact don’t even bother. By then it will have been too late.
I don’t know if I’ll ever see the worlds I saw in my dreams again. I would like to, at least once before I die. And if I do see them, I want to see it in a just and humane world, not one teetering on the brink of extinction. At the end of the day: that’s all we can ever hope for.