The Distillery

Introduction: one Emi, many questions

It started, of course, with a comment. Under an AI-generated wave someone confidently announced: this is “not art”. Not because the image changed, not because their feelings changed — but because they suddenly learned which tool was used.

That tiny outrage did something very useful for me: it pushed a very old question back to the surface — who actually counts as an artist? Because if one JPEG can be “art” at first glance and “not art” after reading the bio, maybe the problem is not in the picture at all.

From there, I went backwards. If our contemporary Emis patrol the internet calling themselves “real artists” because they can draw with their hands, what did people historically demand from artists beyond “I can hold a brush”? Spoiler: a lot more than a steady wrist and an expensive set of oils.

Part 1. When the artist stopped being just “the guy with a brush”

For a long time, painters were treated as high-end craftsmen. You paid them like you’d pay a good mason or a goldsmith: for hours, materials, clean execution. Technique was the whole story.

Renaissance theory breaks that. People like Leon Battista Alberti start describing painting as a science and a language, not just a manual craft: the painter should understand geometry, optics, anatomy, proportion, narrative, and human emotion — otherwise the picture may be technically polished and still intellectually empty.

So already in the 15th century, the job description quietly mutates from “someone who applies paint correctly” to “someone who thinks with images”. And that’s the part modern discourse somehow managed to forget.

Part 2. The universal skills that always came with the job

If you strip away specific techniques (oil, red chalk, watercolor, 3D, Midjourney), what did people actually look for when they said “real artist”?

1. The ability to see, not just to make

Artists were always expected to have a trained gaze:

to notice proportions, light, gesture, expressions that a regular viewer simply skims past;

to sense the fine differences in composition: why one scene holds together and another falls apart;

to look at other people’s work not as “like / don’t like”, but as a set of decisions.

Today we call this visual literacy and visual culture. And it is a skill that takes years to build, not a side effect of buying brushes.

2. Building a story, not just a picture

Alberti was already saying that the main thing in painting is istoria — the story it tells. Which means the artist has to be able to:

construct a narrative inside the frame;

place accents so the eye travels along a trajectory instead of bouncing randomly;

choose details that serve the meaning, not just “look pretty”.

Whether you like it or not, someone who builds a series of AI images into a coherent visual narrative is doing the same kind of work fresco painters did, just with different tools.

3. Curation: selection, montage, cutting the trash

No sane artist shows all their sketches and failed canvases. Historically, a brutally important part of the job was:

to throw away 90% of the in‑between junk;

to choose the handful of decisions that actually carry the idea;

and sometimes to assemble found objects, references, quotations into a new statement.

That is curation: selection, montage, composition. In an era where you can generate a thousand images in an evening, this skill becomes the nervous system of creative work.

4. Thinking beyond a single medium

Historically, artists almost never stayed welded to one medium until death. They would:

learn from architects, poets, playwrights;

think in terms of space, body, text, ritual;

transplant techniques from one field to another (composition from theatre, gestures from rhetoric, symbolism from religion).

Today this gets rebranded as “cross‑disciplinary thinking”, “conceptual thinking”, “art direction”. But essentially it’s the same demand: stop being a hand that performs; start being a head that manipulates meaning through whatever tools are available.

5. Social intelligence and reputation

Another entertaining line from old texts: the artist should know how to behave. Not just draw — negotiate, communicate, work with patrons, be a functional human being who doesn’t wreck a commission with their own attitude.

Put bluntly: if you spend your evenings under creative directors’ posts, typing in all caps about how their work is “not real art”, you’re failing a baseline competence that was already considered important five hundred years ago.

Part 3. The modern error: the artist as “hands without a head”

Where did contemporary culture go off the rails? We slid back to the most primitive model: artist = someone who does something with their hands. Mind, taste, visual literacy, curation, narrative, the ability to work with tools — all of this is edited out as “non‑essential”.

That’s exactly where the AI panic comes from:

if your entire concept of an artist is limited to motor skills,

then any tool that shortens the distance between idea and image feels like a personal attack.

People cling to technique as identity: “I am somebody because I can draw like this”. The rest of the profession — seeing, thinking, selecting, connecting, inventing — stays invisible to those who never learned to do it.

And that’s why they can so easily say “not art” about a piece that hit them before they found out which software was in the pipeline. In their worldview, the artist is the hands. Anything that breaks this picture must be kicked out of the category.

Part 4. The artist after AI: back to the original brief

Paradoxically, AI drags us back to an older, more honest version of the artist:

tools become cheap, fast, in some sense interchangeable;

the real battleground shifts to ideas, selection, visual literacy, the ability to carve one clear line out of a million possibilities;

suddenly there is a massive demand for people who can think in terms of art direction, not just “make a pretty line”.

What distinguishes a living artist from endless generations is not that they own a brush, but that they own criteria: why this image exists, what it does to the viewer, where it leads the next one, what it adds to your visual world.

In other words, everything that historically was considered the core of the profession, not a decorative add‑on.

Modern culture has unfortunately raised an army of people who are convinced: “I’m an artist because I can do something with my hands” and have absolutely no idea how absurdly huge a block of work and skills hides behind the word artist.

First they see a wave.

Just a wave — texture, movement, light, that little hit of “oh, damn.” They repost it, call it “real art,” throw hearts around like it’s some considered, eternal verdict.

Then someone scrolls to the bio, sees “AI Explorer” and a Midjourney hashtag.

And suddenly the exact same JPEG stops being art. Not a single pixel changes — the only thing that changes is their willingness to admit their own reaction was real.

This isn’t even a serious question of “what is art?”.

It’s a panicked little rebranding:

“As long as I thought it was oil paint, I was a refined connoisseur.

Now that it’s AI, turns out I just liked a random generator output.”

They literally rewrite their own “wow” after the fact, because it landed on the wrong square of their belief map.

You’re apparently not allowed to enjoy an image if behind it isn’t a properly exhausted human with a brush, but a human+algorithm stack.

Much easier to declare the whole category invalid: not art, and you’re safe again.

My favorite side effect: without that tiny AI trigger, I would’ve completely scrolled past it.

Perfect wave, likes, comments — and I’d have filed it away as yet another pretty picture in the endless feed of pretty pictures.

But the moment someone started shrieking “YOU’RE ART? THIS IS WRONG,” I got interested.

I went to the profile, went through the works, hit follow, because if one frame makes people that unhinged, there’s clearly something worth talking to.

So yes, thank you to everyone typing “not art.”

How many styles can you actually name without googling?

I keep seeing people build their entire “creative identity” on one sentence: I don’t use AI.

Not “I make good work.” Not “I push myself artistically.” Just “I don’t touch the scary machine, therefore I’m clean.”

It’s adorable, in a museum-exhibit way. Like watching someone proudly announce they only use analog film while shooting the same boring picture of a coffee cup for the 500th time.

Let’s talk about slop.

Slop is not AI-specific. Slop is what happens when you take the path of least resistance in your own head and then try to pass it off as “art”.

The 900th copy-pasted horny meme with a different character face slapped on.

The same generic busty blonde in the same three poses, in slightly different “fantasy” outfits.

The fanfic that drops an OC into Hogwarts and then runs them through the exact same three tropes you’ve seen since 2007.

None of that requires a neural network. Most of it is proudly “pencil-only”, “tablet-only”, “traditional”, “pure”.

And you know what? It’s still slop.

If your idea is “Hermione but hot” or “Slytherin OC but even more traumatized” and you just run that through the usual angst/romance pipeline — congratulations. You’ve produced content. You haven’t produced anything culture actually needed.

“Pencilslop” is real. Welcome to the party.

I saw someone meltdown over a Tumblr user who’s pro‑AI and dared to say “pencilslop”. The outrage was almost touching:

“Holy shit I never thought I’d see the day.”

Of course you didn’t. You thought “slop” was reserved for AI prompts and cheap mobile filters. You thought as long as you were grinding away with a pencil, you were automatically absolved of banality.

Newsflash:

You can generate slop with Midjourney.

You can generate slop with Blender.

And you can absolutely generate slop with a pencil.

If you never push your ideas beyond “same hot wizard, different slightly darker trauma”, your medium is irrelevant. The laziness lives in the concept, not in the tool.

“I don’t use AI” has become a kind of moral deodorant.

People spray it all over their work and hope no one will notice:

that their fanart is just the same three expressions on the same three character types;

that their big original project is structurally indistinguishable from a standard Hogwarts self‑insert with the serial numbers filed off;

that the only reason they’re yelling about ethics is because for the first time, their favorite shortcuts aren’t the fastest ones anymore.

“I don’t use AI” doesn’t mean:

“I think deeply about what I’m making.”

“I take risks.”

“I build worlds instead of just repainting existing ones.”

It only means: “I chose a slower tool.”

Slow doesn’t equal profound. Sometimes slow just equals inefficient slop.

The uncomfortable truth is this:

An AI model can produce breathtaking, weird, risky work — in the hands of someone who actually wants that.

The same AI model can happily spit out endless blonde titty‑elves — for someone who thinks that’s the peak of imagination.

A pencil, a tablet, a brush can do the exact same two things.

Your choice of tool says nothing about your courage, your taste, or your willingness to think. It only says something about your relationship to process.

If your whole identity is “I suffer more, therefore my work is better”, I have bad news: Suffering is not a reliable proxy for quality. You can suffer endlessly over something that was doomed to be slop from the moment you chickened out of trying a new idea.

There is one sentence that actually means something:

I try not to make slop.

Not “I don’t use AI.” Not “everything I do is hand-drawn.” Not “I only write original fiction.”

Just: “I take responsibility for whether my work adds anything beyond more warm, recycled mush to the cultural feeding trough.”

That sentence holds you accountable no matter what you use:

If you’re writing Hogwarts fic, you ask: Is this actually a new angle, or just the same trauma fanservice re-skinned?

If you’re drawing fanart, you ask: Does this do more than paste a face onto a meme template?

If you’re using AI, you ask: Am I pushing this beyond ‘sexy elf, again’ into something that would be impossible or impractical otherwise?

Slop is a content problem, not a technology problem. And yes, pencilslop is absolutely a thing.

The original argument is simple, almost disarmingly so:

“I don’t care if you use AI… but you should disclose it. It’s about integrity.”

It leans on an analogy:

Using AI without disclosure is like tracing someone else’s art and pretending it’s original.

And then lands here:

If you’re not chasing credit for work you didn’t do, why not just be honest?

Clean. Polite. Morally framed. Completely detached from reality.

Now let’s talk about the part they quietly deleted: context.

Because this whole “just disclose, have integrity” posture only works in a world where disclosure is neutral. Where it doesn’t immediately tank your reach, your income, your social standing, or your inbox sanity.

We do not live in that world.

We live in a moment where saying “I used AI” functions less like a neutral fact and more like pulling a fire alarm in a crowded building full of people who already decided you’re the fire.

So when someone shows up, gently suggesting “why not be honest,” what they’re actually proposing is reputational self-harm framed as a character test.

Very elegant.

Let’s examine the tracing analogy, because it’s doing a lot of emotional labor here.

Tracing was controversial, yes. But tracing never triggered coordinated harassment campaigns, mass dogpiling, or algorithmic suppression. No one was quietly DMing you, “hey, I agree with you but I can’t publicly like this or I’ll get dragged.”

AI does.

So the comparison fails at the only level that matters: consequences.

When consequences differ this dramatically, “integrity” stops being a moral absolute and starts being a luxury good.

And then there’s this charming implication:

“If you’re not trying to get credit for something you didn’t do, why not disclose?”

Let’s translate that out of moral theater into something usable.

Why wouldn’t someone disclose?

Because disclosure measurably reduces engagement

Because engagement is often tied to income

Because audiences are currently incentivized to punish, not evaluate

Because platforms reward outrage, not nuance

Because people enjoy moral purity tests when they’re not the ones being tested

None of this is abstract. It’s basic incentive design.

You don’t get honesty out of hostile systems. You get concealment, signaling, and selective silence.

There’s also a subtle sleight of hand in the word “integrity.”

Integrity here is framed as consistency between action and disclosure.

But in practice, what’s being demanded is something closer to ritual self-labeling for the comfort of an audience that has already decided what the label means.

Are we really talking about integrity?

No. We’re talking about legibility under threat.

And the funniest part is that the same environment that punishes disclosure will, without hesitation, accuse people of dishonesty for not disclosing.

A closed loop.

Punish → conceal → moralize concealment → punish harder.

Then act surprised that people opt out.

If you strip away the moral framing, what remains is almost embarrassingly pragmatic:

People reveal information when the cost of revealing it is lower than the cost of hiding it.

Right now, for many artists, the math is obvious.

So no, the question isn’t “why aren’t people more honest.”

I’m probably the worst possible person to romanticize fanfiction, simply because I barely read it. In my entire life, I’ve read two fanfics: one based on a universe I didn’t know (so it didn’t register as “fanfic” at all, just as a slightly weird original story), and one based on a world I know inside out.

The second one I didn’t finish. I dropped it halfway through. But the thought it left me with turned out much more interesting than the text itself.

It’s a thought about cheat codes, resource saving, and the very selective morality around “laziness” and “theft” in online creativity.

Fanart: the socially accepted cheat code

With fanart, the pattern was always obvious. You want to get noticed as an artist? You pick a big fandom, draw well-loved characters, and suddenly you’re not screaming into the void — you’re plugging straight into an existing emotional infrastructure.

You don’t have to explain who these people are. You don’t have to build attachment from scratch. You don’t have to prove that this moment is dramatic, or tender, or devastating — the fandom already did that work for you.

You just provide a new visual angle on something people are already obsessed with. And they reward you with likes, reblogs, followers, attention.

It’s not evil. It’s not shameful. It’s a rational growth strategy in an overcrowded attention economy. But let’s not pretend it’s “starting from zero”. It’s a cheat code. A smart one.

Fanfiction: industrial-scale resource saving

Today, after that second fanfic, the second half of the pattern finally clicked into place: fanfiction does exactly the same thing, just on a deeper level.

When you write fanfic in someone else’s universe, you are not just “playing with characters you love”. You are skipping one of the most expensive parts of writing: building the world.

Here’s what you get pre-installed, for free:

The internal logic of the world — how it works without collapsing every three pages.

System rules: magic/technology, social norms, taboos, institutions, politics, history, races.

A global conflict and clear moral axes.

The tone and atmosphere, from color palette to emotional baseline.

The original author spent human-weeks, human-months, sometimes human-years figuring this out. They broke and rebuilt the rules of that world until it held together. They fought with contradictions. They debugged the lore.

You, as a fanfic writer, start from fully built sets and working lighting. No all-nighters trying to reconcile world rules. No existential panic over “if X is true in this world, then Y can’t be, so how do I fix it?”.

The stage is built. The conflict is wired in. The world already knows how to breathe.

You just roll your own actors onto the stage and hit “action”.

In that particular fanfic, those original characters didn’t convince me at all. They felt pasted on. But my problem here isn’t taste — it’s structure. The structural fact that this entire world, including its emotional gravity, belongs to someone else’s labor.

Fanfiction as a growth hack

And it’s not just about saving effort. It’s about tapping into someone else’s audience.

When you write a purely original story, you have to:

Convince a stranger that your unknown world is worth their time.

Build emotional investment in characters from zero.

Drag the reader through pages of “onboarding” into your universe before anything pays off.

When you write fanfiction for a large fandom, all of that is already prepaid:

The reader arrives already loving this world.

They’ve got baseline attachment to the setting, maybe to some characters.

They’re hungry for “more” before you even show up.

You’re not working in a vacuum — you’re entering a pre-heated environment of attention and desire.

That’s exactly how fanart works as a growth strategy: you don’t just draw, you tap into an IP that people are already emotionally addicted to. Fanfic is the prose version of that same cheat.

Let’s just say it plainly: fanfiction is a cheat code both in terms of labor saved and in terms of promotion gained.

Tumblr, AI, and extremely selective ethics

Now, enter Tumblr.

You scroll through AI-related tags and get hit with a choir of moral superiority:

“Pick up a pencil, stop being lazy, draw it yourself.”

“Just write it yourself, don’t rely on a machine.”

“AI is theft. AI is cheating. Real art takes effort.”

And then you click on their blogs and see pages of fanart and fanfic based on massive corporate IPs and beloved franchises.

People whose first exposure, first audience, first growth spurts came from drawing or writing on top of someone else’s universe. People whose reach is structurally tied to the fact that they piggybacked on worlds and characters they did not create, because that’s where the attention already was.

These same people are now yelling at others about “laziness” and “cheating”.

Is this ethics? That’s comfort. Ethics would at least require consistency.

“But fanfic is hard work!” — yes, and?

This is where the cognitive gymnastics peak.

When it comes to fanfic or fanart, the defense is:

“But I still work hard! I invent plots, dialogues, emotions, scenes. That’s my effort.”

And that part is true. Fanfic can be technically and emotionally demanding. No one is denying that.

But the moment AI enters the picture, the script flips:

“Using a machine to save effort is cheating.”

“You’re skipping the real work.”

“You’re not a real artist if you rely on tools.”

So let’s demystify this.

Effort-saving is fine as long as it looks familiar and human-shaped:

Building on existing fictional universes? Acceptable.

Using genre conventions and archetypes instead of inventing new narrative structures? Normal.

Riding the popularity of an existing IP to promote your own work? Smart hustle.

The only time effort-saving suddenly becomes immoral is when the tool is new, powerful, and not under their control.

Then, magically, “saving time” turns into “laziness”, and “leveraging existing structures” becomes “theft”.

Everyone uses cheat codes. Some people just lie about it.

Fanfiction is not evil. Fanart is not evil. Neither is using AI.

All of them are ways of not starting from a blank, featureless void.

Fanfic saves you worldbuilding and injects you into a ready-made emotional landscape. Fanart saves you from explaining why these characters matter and uses existing attachment as fuel. AI saves you from parts of the technical grind — the scaffolding, the tedious iterations, the zero-to-one.

If your problem is only with the last one, then it’s not about “ethics of creativity”. It’s about guarding your own shortcut while screaming that everyone else must climb the long way up.

Most beginners would stay invisible forever if they didn’t start on someone else’s universe. That’s reality. That’s how culture actually spreads and mutates. It’s not shameful.

People love the little morality play where there are two kinds of AI.

On one side of the stage you get “serious, analytical AI.” It lives in hospitals, optimizes logistics, assists doctors, politely crunches data and never offends anyone’s aesthetic sensibilities. On the other side lurks “filthy generative AI”: slop, porn, deepfakes, fanfic, “soulless content” and supposedly the end of human creativity.

They beg us: please don’t confuse them, generative AI is the bad one, analytical AI is the good one that will cure cancer.

The problem is very simple: cancer will not be cured by a model that only points at spreadsheets. At some point, something has to be invented.

Looking at the data is not treatment

“Analytical AI” in this discourse is basically a worshipped calculator.

It finds patterns in medical records. It predicts risks. It clusters patients. It nudges doctors toward better decisions. All of that is valuable and genuinely life‑saving at the margins: earlier diagnosis, fewer missed cases, slightly better chosen therapies.

But this entire story lives on top of existing options.

Existing drugs.

Existing treatment protocols.

Existing surgical techniques.

An analytical system can say: “Given what humanity already tried, Option B is better than Option A for this subgroup.” It cannot conjure Option C out of thin air.

When people fantasize about “AI helping cure cancer,” they quietly jump from:

“It analyzes data better”

to

“It proposes new medicines, therapies and biological interventions that never existed before.”

That jump is exactly where they step straight into the territory they claim to despise.

The moment you leave the archive, you’re doing generative work

To move from “helping doctors sort through history” to “changing what is biologically possible,” a system has to produce new candidates:

a new molecule that is not in any current library

a new protein sequence with improved properties

a new combination or protocol that no clinical trial has tested yet

You can rename it all you like, but structurally that is what machine‑learning people call generative behavior: instead of just labeling and ranking what already exists, the model proposes new points in the space.

It is no longer just saying, “this molecule looks promising.” It is saying, “build this one, it fits your protein target like a glove, even though no chemist has drawn it before.”

It is no longer only evaluating protocols. It is suggesting, “combine these steps, in this order, with these constraints; try it, your probability curve just shifted.”

You do not get “AI that actually cures something” without this creative, constructive step. There is no magical third category of system that can design new interventions while somehow staying morally untouched by generation.

The biology they romanticize already runs on generative engines

Most people loudly distinguishing “good analytic AI” from “bad generative AI” have simply never looked at how modern life sciences actually use these systems.

Drug discovery teams use models that design completely new molecular structures given a protein’s 3D surface, desired binding properties and toxicity constraints. The model doesn’t just rate a catalog; it fabricates candidates that labs then synthesize and test.

Protein designers feed sequences and structural data into models that output redesigned proteins: more stable, more specific, more potent than the natural versions. Some of these are already being pushed as next‑generation biologic drugs.

Regenerative medicine groups use large models trained on biological sequences and literature to propose new variants of reprogramming factors — from the same family as the famous Yamanaka factors — that flip adult cells back into more plastic, stem‑like states with far higher efficiency.

All of that relies on generative behavior. The system is not just answering “Which of these older options should we try?” It is generating new options for reality to test.

The “AI in healthcare” these posts worship is already crawling with generative models; they’re just hidden behind unfriendly interfaces and paywalled journals, so no one writes moral threads about them.

What they actually hate is the interface

Notice what triggers the outrage.

Nobody is flooding tags with essays about how “protein‑design AI is killing the soul of structural biologists.” Nobody writes three‑screen laments about molecular generators “replacing real chemists.”

Why? Because there is no shiny public button to press. The generative behavior is wrapped in opaque tools, dense papers and corporate pipelines. It feels like distant, respectable magic.

The same architectural ideas, the same family of models, put behind a chat box and an image canvas, instantly become a cultural scandal.

The tantrum is not about:

probabilistic modeling,

latent spaces,

sequence generation.

It is about who gets to access that power.

As long as generative capabilities stayed locked in the hands of pharma, big tech and specialized labs, they were “serious AI.” The moment the same kind of models started writing, drawing and coding for anyone with a laptop, they became “slop” and “soulless trash.”

They’re not guarding science. They’re guarding a hierarchy.

“We love AI in medicine, just not that kind” really means “we didn’t read the manual”

When someone writes:

“Don’t confuse AI in general with generative AI. One is great for healthcare and navigation, the other just produces garbage and porn.”

they are not voicing a subtle technical distinction. They’re unconsciously saying:

“We want the benefits of generative modeling in biology and medicine, but we refuse to admit it’s the same species of system that draws fanart.”

They want:

AI that proposes new drugs,

AI that designs new proteins,

AI that finds novel treatment strategies,

AI that reprograms cells more safely and effectively,

and at the same time they want to keep hating “generative AI” as a category.

To make that feel consistent, they have to imagine a non‑existent third kind of AI that somehow invents new things without ever doing anything they would recognize as “generation.” That creature does not exist.

If your system:

changes what is available to try in the lab,

alters the design space of biology,

invents candidates beyond the existing catalog,

you are living in the house that generative modeling built, whether you spell it out or not.

The comfort of “smart but harmless” machines

There is another, more psychological angle.

An “analytical AI” that just classifies and predicts is easy to domesticate in the imagination. It is basically a better statistician, frozen at the level of “give me insight, I, the human, will decide what is real and what gets built.”

A system that proposes new molecules, new texts, new images, new mechanisms trespasses on territory many people consider exclusively human: design.

Suddenly the machine isn’t just helping you look, it is helping you invent. That is where panic starts.

So the distinction hardens into a moral fable:

“Analytical AI” — clever, obedient, under control, helping experts.

“Generative AI” — unruly, vulgar, leaking power to the masses, vandalizing sacred domains like art and writing.

Strip away the storytelling and the pattern is the same:

Learn a distribution from existing data.

Sample from it in new ways.

Evaluate and iterate.

In art, the samples look like illustrations and essays. In science, the samples look like novel molecules and protein sequences.

The difference is not purity. The difference is comfort.

Stop preaching categories you don’t understand

The next time someone earnestly insists:

that generative systems “only produce garbage,”

that “real” AI in medicine is purely analytical,

that we must “protect healthcare from generative slop,”

the honest translation is:

“I am talking about interfaces and vibes, not about how these systems actually work.”

They want the future drugs, the anti‑cancer therapies, the rejuvenated cells, the protein redesigns. All of that already leans on generative modeling under the hood.

And at the same time, they want to preserve the right to despise any public implementation of the same principles — especially the ones that hand creative leverage to people they are not used to seeing at the controls.

Everyone thinks their problem with AI art is “the theft.” The dataset, the scraping, the broken lore detail, the wrong horn, the slightly-off colors. That’s the official story. That’s the polite story.

And then someone does what this author did: they take their own piece, feed it into a model with a prompt like “turn her around,” and expect a funny little abomination. Something easy to hate. Something that will confirm: see, I’m still special.

Instead, the model answers: “Sure. I can do what you do.”

And that’s where the stomach drops.

What actually happens in this post

Let’s lay it out.

Original: fanart of Karlach, a character owned by Larian, drawn in a very recognizable fandom style: saturated colors, fire background, big axe, demon girl, dramatic pose.

Prompt: “turn her around” in an image model.

Output: a front-facing Karlach, clearly based on the same image: same palette, same lighting logic, same design (with errors), same mood.

And, crucially, a watermark that looks enough like the artist’s signature to hit a nerve.

The artist describes a visceral reaction, calls it a “skinwalker,” says it twisted their stomach. But notice what they fixate on:

Not “wow, it gave her a wrong horn, that’s wrong lore.” Not “the anatomy is broken”. What really disgusts them is: “MY FUCKING WATERMARK”.

The monster, for them, isn’t the wrong horn. The monster is the realization: this thing can imitate the part I thought was uniquely mine.

Accuracy? Replaceability!

The usual anti‑AI script goes like this: “I hate it because it’s bad.”

But if you look at this case honestly, the emotional arc is the opposite.

They went in expecting bad.

They wanted to feel superior to the machine.

They wanted confirmation that their individuality is safe.

Instead, they got a piece that:

Carries over their composition logic.

Reproduces their color language.

Mirrors their weapon design.

And then casually imitates their signature gesture of authorship.

Of course the reaction is “skinwalker.” Not because the AI “stole their soul,” but because it showed that the “soul” they were banking on expresses itself through patterns — and patterns are, by definition, replicable.

The horror is not: “this is nothing like me.” The horror is: “this is much too close, and it did not need me.”

The fetish of the watermark

The funniest part is the outrage around the watermark.

Color palette? Shared. Character? Owned by a studio. Visual tropes? Genre boilerplate.

But the watermark — the little mark that says “me, the irreplaceable human, was here” — that’s sacred. That’s supposed to be untouchable.

So when the model generates something that looks like their watermark, a few things collide at once:

The fantasy that “at least my signature is mine” gets shattered.

The idea of authorship as a magic ritual suddenly looks like what it always was: a convention.

The last line of defense — “you may copy my style, but you can’t be me” — stops feeling secure when a model casually fakes the gesture of being “me”.

That’s why they say it “disgusts” them. It’s not disgust at the machine. It’s disgust at the implication: even your little “human-only” talisman is just another learnable texture.

Karlach, fanart, and the illusion of “my” character

There’s an extra layer of irony here: this isn’t even their OC. It’s Karlach. Larian’s Karlach. A character that already exists as IP, already mass‑produced, already drawn by thousands of other people.

So the artist’s position is implicitly:

I drew a popular corporate character in a common fandom style, but this particular representation is uniquely mine and the universe must acknowledge that uniqueness.

AI has absolutely no reason to respect that illusion. It takes the picture as pixels, not as a fragile ego structure. It doesn’t care which brushstrokes were “personal.” It sees distributions: of light, of form, of contrast, of signature shape.

And in doing so, it shows the artist a version of themselves that fandom culture usually helps them avoid: not a singular genius, but one node in a dense web of aesthetic repetition.

That’s where the “skinwalker” dread grows from. Not from violation. From recognition.

“I refuse to learn it” = I refuse to know

The tags finish the story beautifully:

“fuck AI so so much”

“anti generative ai”

“avoiding learning how to use ai tooth and nail and will continue”

“will go to my grave without it”

That’s a defense mechanism. It’s basically: “I do not want more evidence of how replaceable I am.”

If they learned the tools, they’d see, very concretely, how easy it is to remix their own style, push it, iterate, scale it. They’d also have to admit that some chunk of what they call “my art” is in fact… quite legible to a machine.

Much safer to call the machine a “skinwalker,” frame it as a horror entity, and refuse interaction. If the thing that threatens you is a monster, you don’t have to admit that what really scares you is the mirror it’s holding.

So what’s actually going on with these “AI skinwalker” stories?

Strip the drama, and you get something like this:

Generative models are very good at exposing the patterned, repeatable part of human creativity.

Artists are raised on the story that their value lies in being uniquely irreplaceable.

When those two collide — especially on their own piece — the reaction is not “hm, interesting,” it’s visceral panic.

“Skinwalker” is just the aestheticized name for a very mundane fear:

“I thought my work proved I’m one-of-a-kind. Now a model shows me I’m a style that can be approximated. Where does that leave my sense of self?”

And that’s why the wrong horn is tolerable, but the fake watermark is unbearable. The horn says, “the AI doesn’t fully get the lore.” The watermark says, “the AI doesn’t have to fully get the lore to step into your silhouette anyway.”

“My life is ruined and AI did it.”

Didn’t get into grad school?

Not because you coasted on “gifted kid” vibes for a decade — no, obviously it’s because some sophomore used ChatGPT on their sociology essay.

Can’t see the mountains because of air pollution?

That’s not fifty years of industrial policy, that’s… students using an autocomplete to fix their grammar.

People dying at the beach because the ocean levels are higher?

Clearly not fossil fuels, not politics, not our collective addiction to comfort.

No, babes. It’s the girl who asked an LLM to outline her homework.

There’s something almost adorable in that kind of posts: the way it glues together climate change, GPA anxiety and “kids these days don’t use their brains” into one big hysterical AI effigy and sets it on fire. “Generations are being sacrificed because you didn’t hand‑write your essay like a real martyr.”

The “using AI instead of using your brain is disgusting” line is… cute.

Do these people also scream at calculators? Spellcheck? Google Maps? Or is “using tools” only a crime when the tool is new enough to make them feel replaceable?

You don’t stop thinking when you use AI — unless you come to it already determined not to think.

And funny thing: those people existed long before large language models. They just plagiarized from their friend’s essay instead of from a stochastic parrot.

If your political imagination jumps from “LLMs bad” straight to “people are dying at the beach”, the problem isn’t technology. It’s that you have no model of power except yelling at undergrads in your head.

So yes, let’s blame AI for everything:

– your burnout

– your student debt

– your unread books

– your ex’s emotional immaturity

– your cat refusing to have kittens on schedule

Same anti‑AI crowd after the court ruling: “Oh my god, AI companies are destroying books, red flag, Fahrenheit 451!!”

Guys, did you even glance at the paperwork between screams? This is a procedure blessed by a judge — and, by the way, that little blessing is exactly what turned it into a standard pipeline for the whole industry.

One physical copy → scanned → the physical book gets destroyed because the court said that’s how you keep it “legal.” That’s a court‑approved workflow.

If anyone in this story is cosplaying the firemen from Fahrenheit 451, it’s the legal system that basically decided, “eh, we’ll survive losing a few rare books.” And in the actual book, burning wasn’t about recycling paper — it was a weapon of state censorship, a way to erase ideas from public space, not a side‑effect of some clumsy training pipeline.

I’ve started to notice a very specific ritual in my notes and reblogs. It goes like this:

Anti‑AI paladin discovers my post.

Writes a dramatic “gotcha” reblog, full of moral fireworks.

The moment I try to answer, Tumblr spits out an error: surprise, baby, you’ve been blocked.

At this point I don’t even bother typing the reply before I test it. The tone is usually such a dead giveaway that I’ve developed a whole little protocol: first I try to reblog their reblog with nothing in it and save it to drafts. If Tumblr screams at me — that’s the classic “comment‑then‑block” maneuver. I grab a screenshot or a link, and then I just reblog my own post with their sermon attached. Who exactly is going to stop me.

This pattern amused me for a while. Then I realized it’s not just funny, it’s textbook.

They’re not talking to you, they’re performing at you

These people are not coming to have a conversation. They’re coming to stage a scene.

The point of their reblog is not: “let’s test our ideas against each other.” The point is: “I need a public artifact that proves I am on the Correct Moral Side.”

You can see it in the tone: no questions, no curiosity, just a packaged mini‑lecture that reads like it was written in advance and waiting for any pro‑AI post to land on. It’s not engagement, it’s deployment.

That’s why the block happens instantly. The minute they hit “post,” the useful part of the interaction (for them) is already over. They got the screenshot, the feeling, the imaginary applause from their mental audience. Your reply would only introduce risk: you might respond clearly, sharply, and now their little victory monologue looks less like “righteous truth” and more like “underbaked talking points getting dismantled by someone who reads.”

So they cut the wire before the bomb can be inspected.

Comment + block = control freak theater

The combo of “lecture” + “instant block” is basically emotional risk management for cowards.

The comment gives them the fantasy of dominance: “I came into your space and told you what’s what.”

The block guarantees they never have to deal with the follow‑up: no counter‑arguments, no nuance, no evidence that their framework leaks like a sieve.

They get the power rush of “having the last word” without doing the work of actually winning an argument. It’s like walking into a debate, yelling your opening statement into the mic, then sprinting out of the building and calling that a “victory.”

They don’t want a conversation. They want a screenshot with your icon above their speech bubble so they can feel like they “owned” the Enemy. You’re a prop. An extra. A necessary villain in their little morality play.

Echo‑chamber hygiene: keep the heretic out

Blocking you instantly also serves another function: hygiene for their echo‑chamber.

If they keep you unblocked, three dangerous things might happen:

Their followers might see your reply and realize you’re… coherent.

Someone from their side might start questioning the “every anti‑AI is a noble victim, every pro‑AI is a cartoon villain” script.

They themselves might feel the itch of cognitive dissonance: “why does her argument sound less hysterical and more grounded than the fear‑bait I’ve been reposting?”

We can’t have that.

So the ritual is:

Mark the outgroup as immoral and stupid.

Deliver emotional verdict.

Purify your environment by deleting the possibility of counter‑speech.

This isn’t ethics. It’s superstition with better wifi.

Anti‑AI as emotional outlet, not moral principle

They will tell you they are defending Art, Humanity, Workers, Insert Capitalized Abstraction Here. But look at the behavior instead of the slogans.

If your goal is to defend anything in good faith, you want debate, precision, actual information. You don’t run from reply chains like a vampire from sunlight. You don’t block the first person who might say, “okay, but that’s not how law / technology / authorship actually work, let’s unpack it.”

What they’re doing is using “anti‑AI” as a socially acceptable container for their anxiety and status panic. The tech is moving. The tools are changing who gets to do what. Their internal narrative “I am special because my medium is gatekept” is evaporating. That hurts.

It’s much easier to scream “thief”, “slop”, “plagiarism machine”, post the incantation, and slam the door than to sit with the far less glamorous question: “what does my value look like in a world where powerful tools are widely available?”

The first scenario gives you instant moral superiority. The second requires introspection. Guess which one performs better on a dashboard.

Why I find it funny instead of threatening

By now, the pattern is so predictable I can smell a comment‑then‑block from the opening sentence. That’s why I test the reblog before I type.

And when Tumblr says “nope, you can’t reply to this person,” that doesn’t feel like censorship. It feels like a diagnosis. The Brave Anti‑AI Paladin rode into my blog, shouted their script at the dragon, then teleported out before the dragon could say: “Sweetheart, your sword is made of cardboard.”

So I screenshot. I link. I respond anyway. Not for them — they ran. For the people who can actually read, think, and stay in the room when someone talks back.

1. What do you think of people saying ai is increasing utility or ram prices?

2. Is using ai for roleplay one of the wrong reasons for using it? Should people just stick to writing instead since it’s rumored that using chatbots can be addictive?

Hi there!

About RAM prices

On my side of the world I don’t see “AI makes your home electricity bill explode”. What I do see is AI eating the hardware market. Memory and server components have been pushed into a hype‑bull phase because fabs are prioritising high‑bandwidth DRAM and SSDs for AI data centres, and everyone else is left paying scarcity prices. So yes, when people complain that AI is making RAM more expensive — that part tracks. But that’s an infrastructure and supply‑chain story, not a moral failing of people who use chatbots.

About AI role play

The short version: humans will turn anything that responds to them into a playground, a therapist’s couch, or a drug. Books, MMOs, fanfic, forums, Tinder, Discord – everything carries addiction potential if it gives you predictable dopamine and a sense of control.

Long version here

From anti-ai tags: “I’m anti-using AI for ‘creative purposes’ because it displaces real artists and overtakes their livelihoods.”

Right. And yet somehow Photoshop didn’t trigger a moral uprising about oil painters. Graphic tablets didn’t summon candlelight vigils for charcoal sketch artists. Entire industries quietly absorbed new tools, workflows shifted, aesthetics evolved — and no one performed this much public grief.

Curious.

Even more curious: the same people clutching pearls over AI seem perfectly comfortable vacuuming their floors, running a washing machine, or cooking their own meals. By their logic, that’s three jobs casually erased before breakfast — cleaner, laundress, cook. No thinkpieces. No guilt. Just… convenience.

So let’s be honest about what’s happening here.

Is this really about protecting artists? No, it’s about selectively assigning moral panic to a technology you personally dislike.

Because if this were about artists, the conversation would look very different. It would sound less like abstract declarations and more like receipts.

When was the last time you commissioned original work? Not liked, not shared, not “boosted.” Paid. How much? How often? What percentage of your monthly budget actually goes to the creative labor you claim to defend?

Silence gets very loud at this point.

“Support artists” has become a kind of aesthetic performance — a slogan you can wear without ever opening your wallet. A moral stance that conveniently costs nothing while policing what others are allowed to use, build, or explore.

“I wish all you AI artists and AI writers believed in yourselves half as much as you believe in AI.”

Cute sentiment.

But honestly, I want the opposite for them: I want them to finally stop obsessing over things that are none of their business.

A self-sufficient person is painfully boring in practice.

They do what they enjoy, with the tools they actually like.

If they’re not into AI, they simply… don’t open the tab. Wild concept.

Instead we get theatre: people who could be happily doing their “authentic” painting spend their time camping under someone else’s prompts, writing essays about how real creativity requires suffering in the correct, properly analog way.

The image I can’t unsee:

someone standing there, holding their own genitals in their own hand. Everything’s fine. Their body, their process, their pleasure.

And instead of enjoying any of that, they crane their neck to stare at the neighbour — how he is holding his, with what, how often, and in which software.

Then they write a post:

“If you truly believed in yourself, you wouldn’t use that thing.”

No, darling.

If you believed in yourself, you wouldn’t short‑circuit every time strangers use tools you don’t control in projects you’ll never see.

You wouldn’t need to declare every process outside your comfort zone “barf slurry” just to feel taller.

Real centering on yourself is not a moral audit of everyone else’s brushes, apps, and models.

It’s banal focus: these are my hands, my text, my tablet, my prompt, my ritual — and that’s enough. I don’t need to go sniffing around in someone else’s pants to feel real.

You want to talk about “authenticity”?

I keep coming back to the same boring, unsexy axis under all this AI panic: a lot of people were never told, explicitly, that nothing in this world is ideal. Nothing. No one. Everyone and everything is flawed as hell — the readers, the artists, the systems, the economy. The only thing you can call “perfect” is whatever god you personally subscribe to, and that’s already out of scope for this discussion.

We somehow let a whole generation grow up with the implicit promise that if you “follow your passion” hard enough, there will be a clean, linear, morally satisfying arc: study the thing you love, become The Professional, stay that person forever, die validated. Meanwhile, real life has always worked more like marriage: people fall in love, marry, and… divorce. No one walks down the aisle planning the divorce papers, but reality doesn’t care about the script in your head.

Same with careers. When you pick a profession at 18 or 20, you’re not thinking “I’ll have to rethink this, maybe several times, and reinvent myself along the way.” You imagine one big, coherent identity. But the world we actually live in is one where it’s becoming normal to go through three or four professions over a lifetime instead of getting stuck in a single one. That’s not a glitch; that’s the environment. The sooner we normalize that, the easier it becomes to normalize the idea that you have to keep learning and keeping your skills alive instead of embalming them and praying the market never changes.

I honestly don’t remember anyone “preparing” me for the complexity of life in any systematic way. It felt more like life itself doing the onboarding: it keeps turning unexpected sides toward you and seeing how you cope. By design. That’s the tutorial. I can imagine there are people who were cushioned more — whose immediate environment, family, country, whatever, built them a softer, more stable reality. Good for them.

But when those people collide with what’s happening now — with AI, with shifting markets, with the erosion of “one job for life” — of course they experience it as an existential betrayal. They thought there was an implicit guarantee: if you pick the right dream, reality will respect it. And then someone like me walks in and says, “No one promised you that. Not to me, not to you. The only constant you actually have is your ability to think and learn.”

And maybe that’s exactly why what I’m saying lands as “disappointing”: there is no silver bullet hiding behind the corner. No secret little trick, no clever handwave that makes everything safe again. There is no move that magically keeps your dream job intact while the whole landscape shifts.