Here are my quick and unpolished thoughts on the Studio Ghibli AI discourse.
Back story: Recently, OpenAI rolled out an image generator that (I’m assuming accidentally) made pictures that look eerily like Studio Ghibli films. Studio Ghibli is a legendary Japanese animation studio co-founded by Hayao Miyazaki, known for hand-drawn films that pretty much redefined what animated storytelling could feel like.
Here’s the tweet that started it all.
Now, it’s all over your feed. Endless variations of the same whimsical, cozy, slightly nostalgic aesthetic. The internet did what it always does: found something charming and immediately suffocated it to death by oversaturation.
But let’s unpack why this feels especially irritating.
We trained it on ourselves. Oops.
About 75 years ago, Alan Turing came up with the Turing Test—a simple way to figure out if a machine could convincingly mimic human conversation. He also basically invented the modern computer in his spare time, but for now let’s focus on the chatbot thing.
The Turing Test was originally called the imitation game, and then they were like no lets save that name for a hit movie starring Benedict Cumberbatch, arguably the best machine we’ve got at impersonating a human.
The concept was straightforward: if a human couldn’t tell the difference between chatting with a machine and a person, congrats, the machine passed the test.
But what’s become weirdly clear in recent years is how low we’ve set the bar for what “human” even sounds like. It’s easy to blame tech companies for shoving AI into every nook and cranny of our lives, the harder thing is admitting how much AI reflects our own habits.
AI isn’t coming up with anything new. It’s handing back what we’ve been feeding it all along: the same predictable ideas, clichés, and safe formats we’ve repeated endlessly. We like to mock AI-generated content for sounding formulaic, but who gave it the formula? We did. It learned mediocrity from the experts.
I think for me, AI writing feels like single-use plastics. Convenient, disposable, and quietly contributing to a growing digital landfill. We’re polluting the internet, the trashiest place in the world!!!! Why am I sad about this, I don’t know.
How to spot a bot
Disclaimer: I’m NOT a creative purist. I use AI all the time, especially for work. I use it to brainstorm, tighten clunky sentences, pull together quick research, and come up with blog titles.
Because I don’t have an editor for this blog (as you can clearly tell), I use AI to edit and proofread my personal writing. The output is usually quite predictable, it points out all the things that make my writing sound like me—words like sooooo, calling my readers gummy bears, overusing the word ‘stuff’ and ‘thing’, long confusing sentences.
And when I’d read all those suggestions, I’d give my laptop the bird, and calmly hit publish.
But that’s for my blog, where typos and weird ideas run rampant and honestly I’m fine with that.
When it comes to work, however, I rely on AI heavily. It’s fast and efficient. I’m not above it. And I’m not naive about it either. Every time I use AI to write a tech blog, I’m aware that I’m choosing speed over originality. I know I’m adding to the pile of flattened, recycled writing that already floods the internet.
Since I have to write at least two AI-assisted articles a day, I’ve started noticing small tells in machine-written text. Basically, I can spot AI writing so, so, so quick. And it’s not the cringey stuff like overusing jargon-ey words like leverage, utilize, the fast-paced digital world, etc.
It’s also not about the really painful annoying overuse of just the most GARBAGE metaphors and analogies. Orchestra-related metaphors (symphony, harmony) dance metaphors, circus metaphors (juggling act, balancing act).
Here’s how I spot AI generated writing.
- This awful contrast thing
“It isn’t just ____, it’s _____”
Or “this does more than just ____, it’s _____”.
Or “It doesn’t just ___”.
It’s redundant, repetitive, and SO annoying. I hate it. I HATE IT.
- Since Feb 2025, an overuse of this dash: —
Like every writer to ever exist ever, I have a weird relationship with em-dashes (or m-dashes? I don’t know and I’m not gonna google it now).
I love them. They let me interrupt myself, trail off, circle back. It’s the closest I get to writing how I think. But now I hesitate every time I use one. Because I know AI loves them too.
It’s picked up on the fact that m-dashes create that casual, almost chatty tone people like. And once I realized that, they started to feel less like a choice and more like a default setting. Like I was hitting keys in a sequence the algorithm had already figured out. NOOO.
Leave me and my —s alone.
- Moralizing all the time
ChatGPT loves to moralize. You don’t even need to be writing something inspiring, it’ll try to inspire you all on its own. Doesn’t matter what the prompt is, you’ll get a polished little insight about how creativity is a journey or how we ought to embrace something. The imperfections. The journey (that word is ruined for me). The mess. The chaos.
According to ChatGPT there’s just so much to embrace. All the damn time.
Ironically, one of my favorite signs of genuinely human writing now is seeing typos. I feel a small rush when I spot a spelling error because I imagine a distracted silly person behind the keyboard. And that makes me so happy.
What even is a Ghibli anymore
I’m not exactly an anime person, but I get the reverence for Studio Ghibli. I first heard of Hayao Miyazaki’s films from a friend who gave me a list of movies to watch. I only watched one, Spirited Away, and halfway through it, I paused the movie and spiraled into a rabbit hole about Miyazaki.
Once I read about his work, it felt like the movie came alive. Each frame is drawn by hand. It’s the kind of work that doesn’t scale. There are no shortcuts. You have to commit, or it doesn’t happen. And Miyazaki is famously difficult and obsessive. He’s spent decades insisting on fixing imperfections that most audiences wouldn’t even notice.
They require a level of attention (and patience) that’s become increasingly rare. That’s part of what makes his work hit as hard as it does. It’s an insane level of presence.
Sooooo when OpenAI released its new image generator, and the internet began flooding timelines with AI-generated art that looked suspiciously like his work, something about it felt cheap. Not because the images were ugly. They were fine. Some of them were even beautiful in a polished, moody, nostalgic kind of way.
It’s not like AI is out here deciding what matters. We are. AI is just a tool that reflects what we keep asking for, what we train it on. What we reward. The Ghibli aesthetic didn’t become disposable because a machine and a couple of tech bros corrupted it. It became disposable because we treated it like a template: something you can apply to any prompt, over and over, until the original feels like just another version.
Time to bring back the dead Frenchman who was right about everything
When I saw all the discourse about AI art, I thought of Jean Baudrillard again (let’s call him JB for short).
A few months ago, I wrote about JB while unpacking a phase of my life involving slam poetry. JB was a French philosopher who spent his life examining how we experience reality, and how media reshapes that experience. One of his core ideas was something called hyperreality.
According to him, we create simulations (copies or imitations) that eventually feel more familiar, more “real,” than whatever they were based on. Once the simulation becomes dominant, we start forgetting or devaluing the original.
To explain this, JB used examples like artificial fruit flavor. Strawberry candy doesn’t taste like strawberries, but it’s what we’ve all decided “strawberry” tastes like now. The simulation has overwritten the original. It’s not fake, it’s the new standard.
JB wrote a book called Simulacra and Simulation which I haven’t read, like most people who reference it. But it’s famously featured in The Matrix, a popular 90s film that had everyone practicing backbends. In the movie, Neo hides his illegal software inside a hollowed-out copy of this book.
JB later said the filmmakers misunderstood his work, which honestly makes the whole thing even more perfect.
JB warned that modern life was shifting toward a world full of copies. Images of things, references to things, versions of things, with no real anchor. If we continued down this path, culture would start looping in on itself, endlessly recycling aesthetics, phrases, emotions, values, until no one could point to where anything actually came from.
That’s what the Studio-Ghibli-AI-content reminded me of. Not theft, exactly. Not parody, either. Just a slow dilution. I’m not concerned that these AI-generated images are going to replace Miyazaki’s work. They do something worse. They sit next to it. Quietly. Offering a version that’s faster and frictionless.
And when people see them in enough quantity, the original thing, the one that required skill and attention and years of practice, starts to feel like overkill. What’s the point of all his obsession?
Ghibli’s value doesn’t come from the fact that it’s human-made. It comes from how much effort it takes. Even me, someone who isn’t a movie buff and doesn’t know a whole lot—I can feel the tension in the work. The friction of care. His films are slow and strange. They’re full of pauses that don’t advance the plot, characters who barely speak, quiet scenes that linger for no apparent reason.
That’s partly why I struggled to get through some of it. I’m too fidgety, and I’m used to media that’s easier to consume. Studio Ghibli films leave almost too much space for me. And if AYE don’t fully know what to do with that kind of work, I doubt AI does either.
AI can replicate the look, but not the logic. Art like that is a decision. It’s a set of values about time, attention, detail, and discomfort. You don’t get to those choices by generating a hundred variations and picking the prettiest one.
The thing is, work like Miyazaki’s can’t exactly be flattened into a vibe. You can’t reduce it to a visual shorthand without losing what made it interesting and special in the first place.
We did this
And again, just to be clear: I don’t think AI is the problem. We trained the machine to think like us, and it turns out we weren’t thinking all that much to begin with. It’s tempting to draw a clean line.
AI content: bad. Human content: good. But that falls apart v v fast.
Most of the content we look at might have been written by humans, and most of it is still painfully dull, repetitive, or designed to be forgotten in 20 seconds. The problem isn’t the tool, it’s how low we’ve set the bar. (it’s so low guys)
Ghibli’s work feels different because it comes out of friction. Time, labor, stubbornness, care. Basically everything we’ve been trying to optimize out of existence. That kind of process leaves a trace. You can feel it. And you can tell when it’s missing.
And sure, humans have always used tools to make things. Brushes, chisels, cameras, Photoshop. The camera was once framed as a threat to painting. Photoshop was supposed to ruin design. Every new tool has its panic cycle. But most of those tools still relied on people to shape what came out of them.
A camera doesn’t choose its subject. Photoshop doesn’t write your headline. They extend your hand, they don’t replace it. What’s different about generative AI might be that it blurs the line between the tool and the creator. A camera doesn’t decide what to photograph. AI, increasingly, does decide what to generate. And we just select/approve. It’s a weird dynamic.
Which brings me back to the Turing Test. The whole thing was built around a binary: If you can’t tell the machine from the human, then the machine has passed the test. But the more relevant question now might be: what are we even using to judge that difference?
If the benchmark is whether something sounds convincingly like us, then that test doesn’t say much about AI, but it says a lot about how dumb and boring we’ve become. (sorry I love you)
I think JB wouldn’t have been surprised by any of this. He said it was inevitable. He said that when this would happen, we wouldn’t notice it. He said we’d just adjust to it. That we’d stop asking what’s real, not because we had an answer, but because the question would stop feeling relevant.
He didn’t say BURN IT ALL DOWN or BAN THE TOOLS!!!
He said: notice. Notice when something feels familiar in a way you can’t quite place, ask yourself why. Ask where that familiarity comes from, and what you’re trading for it.
And I’ll say this: Noticing is harder than it sounds. We like to say we want originality, but most of the time, we’re engaging with what we’ve already seen a version of before. That’s not an AI thing. That’s a human thing. We’ve trained entire platforms to serve us the safest possible content because that’s what we’ve quietly taught them we’ll engage with.
I really don’t think AI is threatening some golden age of wild, unfiltered creativity. AI didn’t bring hyperreality into our lives. It’s just helping us see how far we’ve already slipped into it.
And as we keep choosing what’s easy over what’s thoughtful, eventually we’re going to have to reckon with the mess we’ve made.
Unfinished Conversations 
Leave a comment