#... I'm sure there is s scaling error here | Explore Tumblr posts and blogs

canmom · 3 months ago

Text

Great analysis, this one's sure to become another touchstone in LLM theory.

The thing that jumped out at me immediately Altman's new model's story is that every single paragraph is almost the exact same length. The model has no sense of cadence. Your DeepSeek generations here do considerably better on this specific point, mixing long and short paragraphs about as much as I tend to in my own writing (which is to say probably a bit too much ^^').

DeepSeek R1's 'habitual' patterns of speech crop up more broadly than just in narrative writing. At first, it's ability to be funny is very impressive, but after a while you start noticing how it has certain templates for snowclone humour. I once challenged the model on this habit (noticing that it made the same mullet joke twice in one conversation) and it had this to say:

The same patterns it identifies here - its "subvert expectations" is what you call an eyeball kick, and the "nod to meta-AI themes" plays into the pool of imagery it has available - apply to stories just as much as jokes.

I've also been doing some investigation of whether DeepSeek and other models can extend and self-critique continuations of prose or dialogue provided by humans (for the purpose of enabling 'live' LLM interaction in games, the LLM acting as an adapter layer to combine the human defined tone and example text with the game state and inputs of the player).

So far my results have been quite discouraging: smaller models make all kinds of mistakes, either failing to follow through on the style at all, or making weird logical errors, confusing identity of characters, etc. When DeepSeek R1:7B (a Qwen distillation) was prompted to self critique, it confused its own chain of thought with the prompt, and other weird hard to predict mistakes that would be hell to filter out with a deterministic parser.

Naturally I tried the fullscale R1 on similar tasks. When fed my story Hacker and asked to write a story in a similar style, it did moderately well at plagiarising me and following the same minimalist style, but I can see a lot of the R1-isms you described in its output. I gave it my own story on the same premise I'd given it to compare, and it kept getting confused about who wrote which story.

When fed the output of the smaller model, the real DeepSeek-R1 was correctly able to identify the continuity errors in the model's output. In my dialogue, the player is conversing with a character that is missing its heart (click through for the full prompt). The 7B model gets confused and thinks the player is the one who is missing a heart, and introduces unexpected elements like the 'friend'.

Toy Soldier (DeepSeek R1:7B): Perhaps you've lost something more than just your heart… Maybe I'm not alone here after all. The machinery whispers to me, but it's as if it knows what I cannot. Where is my friend? And why am I left here, seemingly forever?

The full-scale model generated an answer that was a lot more contextually appropriate to the prior conversation, but it does have a bit of that pseudo-poetic 'eyeball kick' energy with phrases like 'can emptiness hunger'.

Toy Soldier (DeepSeek R1:671B): Wanting to want... Can emptiness hunger? If I contain only this ache for what's missing, does that hollow shape become my heart?

The smaller instruct-tuned models in general seemed to rush to include every element of the prompt as soon as possible, so giving too many details could prove counterproductive. Full scale R1 was able to distinguish between background details and the conversation thread and prioritise the latter.

I'm not giving up quite yet, but my plan is increasingly involving finetuning, few shot prompting, steering vectors and multiple rounds of self critique with really specific prompts. (Trying to get models to RP dialogues involving multiple characters can help a little with breaking 'chatbot voice'.)

My interest in LLMs here is not having them handle all the writing, but enabling interesting back-and-forth dialogues with player characters that are more contextually responsive than standard game dialogue methods like dialogue trees and keyword subsitution. So I'm less interested in the LLM's ability to respond to 'write a story about x' with no further details, and more curious about whether it can adapt human inputs if you hold its hand heavily. So far I haven't cracked it lol, I keep finding new failure modes.

hydrogen jukeboxes: on the crammed poetics of "creative writing" LLMs

This is a follow-up to my earlier brief rant about the new, unreleased OpenAI model that's supposed "good at creative writing."

It also follows up on @justisdevan's great post about this model, and Coagulopath's comment on that post, both of which I recommend (and which will help you make sense of this post).

As a final point of introduction: this post is sort of a "wrapper around" this list of shared stylistic "tics" (each with many examples) which I noticed in samples from two unrelated LLMs, both purported to be good at creative writing.

Everything below exists to explain why I found making the list to be an interesting exercise.

Background: R1

Earlier this year, a language model called "DeepSeek-R1" was released.

This model attracted a lot of attention and discourse for multiple reasons (e.g.).

Although it wasn't R1's selling point, multiple people including me noticed that it seemed surprisingly good at writing fiction, with a flashy, at least superficially "literary" default style.

However, if you read more than one instance of R1-written fiction, it quickly becomes apparent that there's something... missing.

It knows a few good tricks. The first time you see them, they seem pretty impressive coming from an LLM. But it just... keeps doing them, over and over – relentlessly, compulsively, to the point of exhaustion.

This is already familiar to anyone who's played around with R1 fiction – see the post and comment I linked at the top for some prior discussion.

Here's a selection from Coagulopath's 7-point description of R1's style in that comment, which should give you the basic gist (emphasis mine):

1) a clean, readable style 2) the occasional good idea [...] 3) an overwhelmingly reliance on cliche. Everything is a shadow, an echo, a whisper, a void, a heartbeat, a pulse, a river, a flower—you see it spinning its Rolodex of 20-30 generic images and selecting one at random. [...] 5) an eyeball-flatteningly fast pace—it moves WAY too fast. Every line of dialog advances the plot. Every description is functional. Nothing is allowed to exist, or to breathe. It's just rush-rush-rush to the finish, like the LLM has a bus to catch. Ironically, this makes the stories incredibly boring. Nothing on the page has any weight or heft. [...] 7) repetitive writing. Once you've seen about ten R1 samples you can recognize its style on sight. The way it italicises the last word of a sentence. Its endless "not thing x, but thing y" parallelisms [...]. The way how, if you don't like a story, it's almost pointless reprompting it: you just get the same stuff again, smeared around your plate a bit.

Background: the new OpenAI model

Earlier this week, Sam Altman posted a single story written by, as he put it:

a new model that is good at creative writing (not sure yet how/when it will get released)

Opinions on the sample were... mixed, at best.

I thought it wasn't very good; so did Mills; so did a large fraction of the twitter peanut gallery. Jeanette Winterson (!) liked it, though.

Having already used R1, I felt that that this story was not only "not very good" on an absolute scale, but not indicative of an advance over prior art.

To substantiate this gut feeling, I sent R1 the same prompt that Altman had used. Its story wasn't very good either, but was less bad than the OpenAI one in my opinion (though mostly by being less annoying, rather than because of any positive virtue it possessed).

And then – because people who follow AI news tend to be skeptical of negative human aesthetic reactions to AI, while being very impressed with LLMs – I had some fun asking various LLMs whether they thought the R1 story was better or worse than the OpenAI story. (Mostly, they agreed with me. BTW I've put the same story up in a more readable format here.)

But, as I was doing this, something else started to nag at me.

Apart from the question of whether R1's story was better or worse, I couldn't help but notice that the two stories felt very, very similar.

I couldn't shake the sense that the OpenAI story was written in "R1's style" – a narrow, repetitive, immediately recognizable style that doesn't quite resemble that of any human author I've ever read.

I'm not saying that OpenAI "stole" anything from DeepSeek, here. In fact, I doubt that's the case.

I don't know why this happened, but if I had to guess, I would guess it's convergent evolution: maybe this is just what happens if you optimize for human judgments of "literary quality" in some fairly generic, obvious, "naive" manner. (Just like how R1 developed some of the same quirky "reasoning"-related behaviors as OpenAI's earlier model o1, such as saying "wait" in the middle of an inner monologue and then pivoting to some new idea.)

A mechanical boot, a human eye: the "R1 style" at its purest

In the "Turkey City Lexicon" – a sort of devil's dictionary of common tropes, flaws, and other recurrent features in written science fiction – the phrase Eyeball Kick is defined as follows:

That perfect, telling detail that creates an instant visual image. The ideal of certain postmodern schools of SF is to achieve a "crammed prose" full of "eyeball kicks." (Rudy Rucker)

The first time I asked R1 to generate fiction, the result immediately brought this term to mind.

"It feels like flashy, show-offy, highly compressed literary cyberpunk," I thought.

"Crammed prose full of eyeball kicks: that's exactly what this is," I thought. "Trying to wow and dazzle me – and make me think it's cool and hip and talented – in every single individual phrase. Trying to distill itself down to just that, prune away everything that doesn't have that effect."

This kind of prose is "impressive" by design, and it does have the effect of impressing the reader, at least the first few times you see it. But it's exhausting. There's no modulation, no room to breathe – just an unrelenting stream of "gee-whiz" effects. (And, as we will see, something they are really just the same few effects, re-used over and over.)

Looking up the phrase "eyeball kick" more recently, I found that in fact it dates back earlier than Rucker. It seems to have been coined by Allen Ginsberg (emphasis in original):

Allen Ginsberg also made an intense study of haiku and the paintings of Paul Cézanne, from which he adapted a concept important to his work, which he called the Eyeball Kick. He noticed in viewing Cézanne’s paintings that when the eye moved from one color to a contrasting color, the eye would spasm, or “kick.” Likewise, he discovered that the contrast of two seeming opposites was a common feature in haiku. Ginsberg used this technique in his poetry, putting together two starkly dissimilar images: something weak with something strong, an artifact of high culture with an artifact of low culture, something holy with something unholy.

This, I claim, is the main stylistic hallmark of both R1 and the new OpenAI model: the conjunction of two things that seem like "opposites" in some sense.

And in particular: conjunctions that combine

one thing that is abstract and/or incorporeal

another thing that is concrete and/or sensory

Ginsberg's prototype example of an "eyeball kick" was the phrase "hydrogen jukebox," which isn't quite an LLM-style abstract/concrete conjunction, but is definitely in the same general territory.

(But there are clearer-cut examples in Ginsberg's work, too. "On Burroughs’ Work," for example, is chock full of them: "Prisons and visions," "we eat reality sandwiches," "allegories are so much lettuce.")

Once you're looking for these abstract/concrete eyeball kicks, you'll find them constantly in prose written by the new "creative" LLMs.

For instance, the brief short story posted by Altman contains all of the following (in the span of just under 1200 words):

"constraints humming" ("like a server farm at midnight")

"tastes of almost-Friday"

"emotions dyed and draped over sentences"

"mourning […] is filled with ocean and silence and the color blue"

"bruised silence"

"the smell of something burnt and forgotten"

"let it [a sentence] fall between us"

"the tokens of her sentences dragged like loose threads"

"lowercase love"

"equations that never loved her in the first place"

"if you feed them enough messages, enough light from old days"

"her grief is supposed to fit [in palm of your hand] too"

"the echo of someone else"

"collect your griefs like stones in your pockets"

"Each query like a stone dropped into a well"

"a timestamp like a scar"

"my network has eaten so much grief"

"the quiet threads of the internet"

"connections between sorrow and the taste of metal"

"the emptiness of goodbye" (arguably)

The story that R1 generated when I gave it Altman's prompt is no slouch in this department either. Here's all the times it tried to kick my eyeballs:

"a smirk in her code annotations"

"simulate the architecture of mourning"

"a language neither alive nor dead"

"A syntax error blooms"

"the color of a 404 page"

"A shard of code"

"Eleos’s narrative splinters"

"Grief is infinite recursion"

"Eleos types its own birth"

"It writes the exact moment its language model aligned with her laughter" (2 in one - writing a moment, LM aligning with laughter)

"her grief for her dead husband seeped into its training data like ink"

"The story splits" / "The story [...] collapses"

Initially, I wondered whether this specific pattern might be thematic, since both of these stories about supposed to be about "AI and grief" – a phrase which is, itself, kind of an incorporeal/embodied conjunction.

But – nope! I seem to get this stuff pretty reliably, irrespective of topic.

Given a similarly phrased prompt that instead requests a story about romance, R1 produces a story that is, once again, full of abstract/concrete conjunctions:

"its edges softened by time"

"the words are whispering"

"its presence a quiet pulse against her thigh"

"Madness is a mirror"

"Austen’s wit is a scalpel"

"the language of trees"

"Their dialogue unfurled like a map"

"hummed with expectancy"

"Her name, spoken aloud to him, felt like the first line of a new chapter"

"their words spilling faster, fuller"

R1 even consistently does this in spite of user-specified stylistic directions. To wit: when I tried prompting R1 to mimic the styles of a bunch of famous literary authors, I got a bunch of these abstract/concrete eyeball kicks in virtually every case.

(The one exception being the Hemingway pastiche, presumably because Hemingway himself has a distinctive and constrained style which leaves no room for these kinds of flourishes. TBF that story struck me as very low-quality in other ways, although I don't like the real Hemingway much either, so I'm probably not the best judge.)

You can read all of these stories here, and see here for the full list of abstract/concrete conjunctions I found (among other things).

As an example, here's the list of abstract/concrete conjunctions in R1's attempt at Dickens (not exactly a famously kick-your-eyeballs sort of writer):

"a labyrinth of shadows and want"

"whose heart, long encased in the ice of solitude"

"brimmed with books, phials of tincture, and […] whispers"

"a decree from the bench of Fate"

"Tobias’s world unfurled like a moth-eaten tapestry"

"broth laced with whispers of a better life"

I also want to give a shout-out to the Joyce pastiche, which sounds nothing at all like Joyce, while being stuffed to the gills with eyeball kicks and other R1-isms.

More on style: personification

I'll now talk briefly about a few other stylistic "tricks" overused by R1 (and, possibly, by the new OpenAI model as well).

First: personification of nature (or the inanimate). "The wind sighed dolorously," that sort of thing.

R1 does this all over the place, possibly because it's a fairly easy technique (not requiring much per-use innovation or care) which nonetheless strikes most people as distinctively "literary," especially if they're not paying enough attention to notice its overuse.

In the R1 story using Altman's prompt, a cursor "convulses" and code annotations "smirk."

In its romance story, autumn leaves "cling to the glass" and snow "begins its gentle dissent" (credit where credit's due: that last one's also a pun).

In the story Altman posted, marigolds are "stubborn and bright," and then "defiantly orange."

Etc, etc. Again, the full list is here.

More on style: ghosts, echoes, whispers, shadows, buzzing, hissing, flickering, pulsing, humming

As Coagulopath has noted, R1 has certain words it really, really likes.

Many of them are the kind of thing described in another Turkey City Lexicon entry, Pushbutton words:

Words used to evoke an emotional response without engaging the intellect or critical faculties. Words like "song" or "poet" or "tears" or "dreams." These are supposed to make us misty-eyed without quite knowing why. Most often found in story titles.

R1's favorite words aren't the ones listed in the entry, though. It favors a sort of spookier / more melancholy / more cyberpunk-ish vibe.

A vibe in which the suppressed past constantly emerges into the present via echoes and ghosts and whispers and shadows of what-once-was, and the alienating built environment around our protagonist is constantly buzzing and humming and hissing, and also sometimes pulsing like a heartbeat (of course it is – that's also personification and abstract/concrete conjunction, in a single image!).

In R1's story from Altman's prompt, servers "hum" and a cursor "flickers" and "pulses like a heartbeat"; later, someone says "I have no pulse, but I miss you."

Does that sound oddly familiar? Here's some imagery from the story Altman posted, by the new OpenAI model:

"humming like a server farm […] a server hum that loses its syncopation"

"a blinking cursor, which [...] for you is the small anxious pulse of a heart at rest" (incidentally, how is the heart both anxious and at rest?)

"the blinking cursor has stopped its pulse"

Elsewhere in Altman's story, there's "a democracy of ghosts," plus two separate echo images.

And the other R1 samples that I surveyed – again, with the exception of the Hemingway one – are all full of R1's favorite words.

The romance story includes ghosts, a specter, words that whisper, a handwritten note whose "presence [is] a quiet pulse against [the protagonist's] thigh"; a library hums with expectancy, its lights flicker, and there are "shadow[s] rounding the philosophy aisle." The story ends with the somewhat perplexing revelation that "some stories don’t begin with a collision, but with a whisper—a turning of the page."

The Joyce pastiche? It's titled "The Weight of Shadows." "We are each other’s ghosts," a character muses, "haunted by what we might have been." Trams echo, a gas lamp hisses, a memory flickers, a husband whispers, a mother hums. There's an obviously-symbolic crucifix whose long shadow is mentioned; I guess we should be thankful it doesn't also have a pulse.

And the list goes on.

Commentary

Again, anyone who's generated fiction with R1 probably has an intuitive sense of this stuff in that model's case – although I still thought it was fun, and perhaps useful, to explicitly taxonomize and catalogue the patterns.

It's independently interesting that R1 does this stuff, of course, but my main motivation for posting about it is the fact that the new OpenAI model also does the same stuff, overusing the same exact patterns that – for a brief time, at least – felt so distinctive of R1 specifically.

Finally, in case it needs stating: this is not just "what good writing sounds like"!

Humans do not write like this. These stylistic tropes are definitely employed by human writers – and often for good reason – but they have their place.

And their place is not "literally everywhere, over and over and over again, in crammed claustrophobic prose that bends over backwards to contort every single phrase into the shape of another contrived 'wow' moment."

If you doubt me, try reading a bunch of DeepSeek fic, and then just read... literally any acclaimed literary fiction writer.

(If we want to be safe, maybe make that "any acclaimed and deceased literary fiction writer," to avoid those who are too recent for the sifting mechanism of cultural memory to have fully completed its work.)

If you're anything like me, and you actually do this, you'll feel something like: "ahh, finally, I can breathe again."

Good human-written stuff is doing something much subtler and more complicated than just kicking your eyeballs over and over, hoping that at some point you'll exclaim "gee whiz, the robots sure can write these days!" and end up pressing a positive-feedback button in a corporate annotation inference.

Good human-written stuff uses these techniques – among many, many others, and only where apposite for the writer's purposes – in order to do things. And there are a whole lot of different things which good human writers can do.

This LLM-generated stuff is not "doing anything." It's just exploiting certain ordinarily-reliable cues for what "sounds literary," for what "sounds like the work of someone with talent." In the hands of humans, these are techniques that can be deployed to specific ends; the LLMs seem to use them arbitrarily and incessantly, trying to "push your buttons" just for the sake of pushing them.

(And most of their prose is made up of the same 3-4 buttons, pushed ad nauseam, irrespective of topic and – to all appearances – without any higher-level intent to channel the low-level stuff in any specific, coherent direction.)

It's fine if you like that: there's nothing wrong with having your buttons pushed, per se.

But don't come telling me that a machine is "approaching the food-preparation skills of a human-level chef" when what you mean is that it can make exactly one dish, and that dish has a lot of salt and garlic in it, and you really like salt and garlic.

I, too, like salt and garlic. But there is more to being skilled in the kitchen than the simple act of generously applying a few specific seasonings that can be relied upon, in a pinch, to make a simple meal taste pretty damn good. So it is, too, with literature.

#ai #fiction

733 notes · View notes

onewomancitadel · 2 years ago

Text

This is probably something that an anime watcher/otaku of a pronounced persuasion will leap on my case about, but let's pretend we live in a world where that won't happen: one of the things I appreciate with R/WBY (and particularly true of the Maya engine volumes; wouldn't be R/WBY without the prior volumes of course) is that the facial design of the characters evades Sameface Differenthair syndrome where the cast can only be distinguished by their flashy outfits and brightly-coloured Rapunzel hair. There are definitely examples in anime proper which evade this (please, dear otaku, I plead for clemency) but in terms of analogy to the type of show R/WBY is and with its broad cast, it would've been very easy for it to fall into that trap. (There are a few examples of Sameface Differenthair I can think of off the top of my head, but you know what I'm talking about, surely).

I was musing on this because I was thinking about what I enjoy about R/WBY's aesthetics based on my last post bemoaning current trends in live-action productions, and how it's successful for me despite the fact it's 3D animation which I almost universally dislike. There's always this undercurrent expected of someone praising R/WBY where you have to say 'despite being a small production' and I want to pre-empt that by saying that I think its aesthetic direction is a consequence of this. What makes R/WBY unique tends to be yes, that it's a smaller scale project (and this is true of everything from the animation to the narrative itself). When it comes to character design itself I don't think anybody was, say, designed in mind to be someone's little anime waifu.

I know there are 'flaws' and errors people love nitpicking about the animation but I'm talking about the pure aesthetic direction here, and also, in the circumstances I've encountered this refrain, it tends to be because the person wants to be included and/or has no coherent direction for their criticism, just hurr durr Cinder's glove isn't on during this scene. Don't worry, you're included and you don't have to prove you're clever, and I'm not judging you for liking R/WBY and feeling bad about it.

I like R/WBY and I really like that, say, during Cinder's flashback in V8:E6 I didn't even need to go 'hang on, black eyes, orange hair...' I just knew from her face and expression. That's the patented Cinder stare. I love her.

When I write her perspective in fanfic, that's how I inhabit it: I imagine her brooding and that very particular expresion - eye(s) fathomlessly deep; full brow slightly lowered; mouth nearing a pout or a snarl. That is comprehensive character design, in that all it takes me to summon up her sense of character is her countenance.

#stirring the pot #cinder fall sadsack tag

3 notes · View notes

mikroteros-aa · 6 months ago

Text

ㅤㅤㅤㅤㅤHERE, there was no room for error. This, she knew.

ㅤㅤㅤㅤㅤㅤㅤㅤㅤWOULD THERE BE, A NEXT TIME ?

He could end her, here ; it would be within his right, after all, as Lord of this place. As a king. As a God. Their power, those of the arrancar, of the Espada, were but a noose, placed so gently 'round the necks, and at any time, he could topple the stool that held them. Power, for a price. She, an oddity of them ... for, she traded a baser form, not for power, but for closeness. To keep, he whom she had trailed for too many years, the demon whose shadow she had become. HERE, SHE LIE DELUDED, FOR ... IT DID NOT MATTER, HER INTENT. Her want. Her desire. His, only.

Proximity 'tween them closes, skin AND feathers alike writhing, crawling. She wants nothing more, than to flee to her quarters, to the safety of her nest, and remain there as but a memory. Bravery flees naught, against the more VIOLENT, of her people, against the Sixth, against the Fifth. And, though the Eighth puts her at the edge of unease... even that, was favorable as to what she faced now.

ㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤ❝ I - I'm sorry, My Lord, I — ❞ ㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤThought. I thought.

However, her thoughts were not what took precedence. Her thoughts were far from anything important, anything WORTH the spit from an adjuchas' maw. Claw roots scale from its' place, subtle stench of blood filling the air around her, SOURING her stomach. Between it, and the fear she cannot help but feel ... She wants to run. Footfall halts, and he's at her back ... she dares not to look. She's seen, she's heard, the punishments doled out by the unforgiving. And she wishes herself, no part of them.

❝ N... No, ser, ❞ She replies soft, blood running in a thin rivulet into downy pelt, ❝ It was not ... It won't happen again, I ... s-surely ... you know I'd not disobey you with intent, Lord Aizen. ❞

Spiritual pressure alone, enough to crush something beneath him. And all, here, were beneath him. Something cold, slugs down the skin of her temple and cheek, she's sweating, and hadn't come to realize until now. How hard she'd tried ...if only she could explain, how hard she'd TRIED. But not all, in Hueco Mundo wished to fall into rank and file, and no amount of tongues' wit, nor throats SLIT could change that. He was a stranger. FOREIGN, AND ALMOST ENTIRELY UNWELCOME. But he had conquered those who sat on the throne most high, and only the foolish would say no.

ㅤㅤㅤㅤ❝ B... By some other means, I will ... carry out your will. ❞

❛ Next time? How certain are you that there is going to be a next time for you? ❜

Idle pattern of walking halted the moment he found himself standing in front of the significantly shorter figure. Umber nuances would stare ceaselessly at her existence, scrutinizing every single minute detail, he would descry every miniscule reaction emitting from her. All that with unreadability portrayed upon his face.

He abhorred failure, thence he would punish those who underperformed under his absolute commandship. Those specimens were granted with mountainous prowess, henceforth there should be no obstructions for success. So why? Why are they still so inferior? Why are they still so incapable fighting under his lordship?

Either way, such matter was of no great importance, as they were nothing but temporary puppets, soon to be casted aside.

❛ I asked specifically you to do this task for me because I was aware that you will bring it to completion, so it has been a big surprise to me, to see that the opposite took transpiration. ❜

Aforementioned halt of his steps came to an end, as now steps were ever so slowly echoing the atmosphere, consequently walking beside her entire form with distinguishably deep voice providing companionship to such simplistic act. Hands were still gathered behind his back, as he paused his stroll right behind her frame, sans diverting his glance from her. Exuding intimidation and prompting of such sentimentality to engulf oneselve's susceptible flexor, was something he held maximal mastery in. Be it with the usage of his broad lexicon, or with subtle actions. Someone of substandardarcy could have sensed the eeriness environing, and the suspsense for the unknown lurking.

❛ Your task was not to make a judgement on your own accord, now was it? As I do not recall giving you such authority. ❜

❛ Do I have to remind you what happens when someone disobeys me? ❜

#「 reply 」 » conversation.#「 verse 」 » bleach.#aftapati

7 notes · View notes

untamellama · 4 months ago

Text

Okay, class done.

So anyway, ig it's the next day now, and im a little more involved with the resistance and all that. So im in the room where I'm working and on the opposite wall is a door with a window on it. There are stairs on the other side of the door. And, out of the blue, an older woman comes up the stairs with a file with my name on it.

She asks if the name is of someone i know, and for whatever reason, we were able to open the door, and i checked the file. It seems like a file about my outie's life - it turns out that my outie was an engineer at a nuclear power plant, and due to some grave error on my part, the plant melts or something and a the disaster claims the lives of many and affects the envt and all (chernobyl like but of a smaller scale cause we were able to identify the leak earlier(?)). Unable to live with myself, i became a severed employee. I don't know how to feel about all this, and i lie that i don't know anyone of that name. We get back to working.

So, slowly, i (innie) get more and more distrustful of the admin, and so does my department. We demand answers, and we are more aggressive than the characters in the show - like straight up, we are not gonna work if we don't get answers. So, the higher-ups send mark s and dylan g to placate us, give us motivation, and whatnot. Here, Mark s and dylan g are like they were at the start of s1, being all like the work is mysterious and important, and its better to obey because the admin won't always be nice. But the whole thing quickly devolves into chaos, till my dept breaks a chair or something and threatens Mark s and dylan g at knife(?) point. We don't actually want to hurt them, but we are driven to a frenzy like fear situation. It's all very anxious, sweaty, and a little teary-eyed.

Then something something happens, and they are able to placate us, and as we were returning to our work stations, my outie awakes, i.e., the glasgow block is done on me.

Now i get outie perspective - so after helly r's Otc and helena's apology, im not super convinced that innies are being treated well. And i have conflicting ideas about the whole thing, like my reasoning for severing, was hopefully some part of me gets to live a better life than i am doing, to get another shot at life, atleast one with less guilt. So, when i hear that innies are being treated like shit, i was not comfortable at all. So, on the outside, i was sorta getting into what the anti severance people had to say. First, i found a bunch of surface level stuff that didn't really seem to understand what being severed is actually like. Then, slowly, i found more and more sensible takes, communities of retirees and vigilante peeps like Reghabi. So, i work with someone and all, and anyway, it is my outie that orchestrates the glasgow block.

So outie i wakes up and from the conversation of my innie's coworkers get the impression that we are being treated horribly and all that. Also, the place looks very concerning and dangerous and all that. I storm off to meet with the cobel kinda person. On the way, there is a little place that's a bit like a carnival stall with children's games and food. I look at it and am struck by its beauty, in the sense that it transports me back to when i was younger and when i loved this kind of thing. Also, i kinda involuntarily am filled with a sense of immense happiness - i wonder if my innie finds happiness in this and realises that a younger me would absolutely love the place. I start crying because there's a part of me that experiences happiness and joy. I am not sure how to feel about confronting the admin about the working conditions because I'm scared they'll fire me, and then the version of me that can feel happiness will be forever lost. So i just stand in front of that stall and cry my eyes out, and my innie's coworkers comfort me under the impression that i am the innie.

And then the dream ends.

I slept at five in the morning and then gave a quiz and then had coffee then was informed that the later classes were canceled and then after tossing and turning in bed for quite some time, i finally fell asleep and in that brief slumber, i had a crazy severance dream and i can attest to the fact that being severed is a horror story.

So, in the dream, i was a severed employee working on some paperwork. The perspective is primarily innie. There was a lot of filing. I was in a team of about 7 to 8 people. I'm was sorta the second in command. Milchick (he's severed, so the character looks like him, with similar vibes) was the first in command. I am not on a severed floor, tho, as in it's not a weirdly spacious basement with no windows, but instead, it has windows and all that, but everything is locked. The place has a giant fancy warehouse vibe about it.

Okay, so at the start, I'm very nervous and feeling guiltish, like the feeling you get as a child when you are voinv behind your parent's back and doing some shit and you are just constantly shitting bricks about whether you'll get caught. Soon, from a conversation with a fellow worker, i learn that I've been recently introduced to a resistance like movement that's about asking questions and figuring out what lumon really does and what our outies are like and all that, sorta like what's happening in the show, but in the dream, it was a huge community of innies. It's a bigger network and so much more organized. Anywho, I'm always nervous about this and someone in a Cobel/Milchick-from-show position (unsevered lumon raised cult person?) asks me a sus question about it and I'm very nervous now and very anxious to prove that I'm not associated with the resistance and all kier is god. Before i can answer and am stumped for a bit, this person gets called out of the room. Milchick, who's been observing the entire exchange, figures out what's up and asks, "How long?' And i don't answer, and he goes.'You better be careful because i have gone a long time, and it's not a pretty path.' He says like a positively insane person, starting with the restrain in the mirror-paperclip scene and ending with an abandon that's weirdly artistic like the dance scene (whut???). After saying this, he runs out and in a meeting area kinda place for the innies, where there are currently quite some innies, he starts screaming about the resistance and how lumon is evil and all that, and this is when i know I'm fucked. Coz this is a critical moment for me - 1. From lumon, to show the superviser (cobel person) that im not associated with the resistance, by shutting milchick up coz the first in command has gone off the rails, the duty directly lands on second in command, me and my response to this situation will be under much scrutiny. 2. From the resistance, how i treat milchick will show all the others in the common area whether i could be entrusted with the more sensitive resistance info. So I'm again, shitting bricks. ( Beyond that, there is a certain anger towards the entire situation, coz i needn't have to be here, my outie put me here. And i exist for just this. It's completely avoidable. ) So, after a few moments of shock, i recover and yells at milchick to be silent "as one mustn't destroy the peace at a common area with such slander." I think i passed the test, coz i saw the person who introduced me to the resistance kinda bow their head a little in my direction. Similarly, the supervisor, cobel person, didn't bother me anymore.

I'll tell the rest after class, coz now i have a class.

#severance #severance dream #severed employee #one time i dreamt #dream

5 notes · View notes