Saddish May And Daisy Fic.

Hoshi

HAPPY NEW YEAR TO PEOPLE IN THIS TIME ZONE!!!

AO3'S content scraped for AI ~ AKA what is generative AI, where did your fanfictions go, and how an AI model uses them to answer prompts

Generative artificial intelligence is a cutting-edge technology whose purpose is to (surprise surprise) generate. Answers to questions, usually. And content. Articles, reviews, poems, fanfictions, and more, quickly and with originality.

It's quite interesting to use generative artificial intelligence, but it can also become quite dangerous and very unethical to use it in certain ways, especially if you don't know how it works.

With this post, I'd really like to give you a quick understanding of how these models work and what it means to “train” them.

From now on, whenever I write model, think of ChatGPT, Gemini, Bloom... or your favorite model. That is, the place where you go to generate content.

For simplicity, in this post I will talk about written content. But the same process is used to generate any type of content.

Every time you send a prompt, which is a request sent in natural language (i.e., human language), the model does not understand it.

Whether you type it in the chat or say it out loud, it needs to be translated into something understandable for the model first.

The first process that takes place is therefore tokenization: breaking the prompt down into small tokens. These tokens are small units of text, and they don't necessarily correspond to a full word.

For example, a tokenization might look like this:

Write a story

Each different color corresponds to a token, and these tokens have absolutely no meaning for the model.

The model does not understand them. It does not understand WR, it does not understand ITE, and it certainly does not understand the meaning of the word WRITE.

In fact, these tokens are immediately associated with numerical values, and each of these colored tokens actually corresponds to a series of numbers.

Write a story 12-3446-2638494-4749

Once your prompt has been tokenized in its entirety, that tokenization is used as a conceptual map to navigate within a vector database.

NOW PAY ATTENTION: A vector database is like a cube. A cubic box.

Inside this cube, the various tokens exist as floating pieces, as if gravity did not exist. The distance between one token and another within this database is measured by arrows called, indeed, vectors.

The distance between one token and another -that is, the length of this arrow- determines how likely (or unlikely) it is that those two tokens will occur consecutively in a piece of natural language discourse.

For example, suppose your prompt is this:

It happens once in a blue

Within this well-constructed vector database, let's assume that the token corresponding to ONCE (let's pretend it is associated with the number 467) is located here:

The token corresponding to IN is located here:

...more or less, because it is very likely that these two tokens in a natural language such as human speech in English will occur consecutively.

So it is very likely that somewhere in the vector database cube —in this yellow corner— are tokens corresponding to IT, HAPPENS, ONCE, IN, A, BLUE... and right next to them, there will be MOON.

Elsewhere, in a much more distant part of the vector database, is the token for CAR. Because it is very unlikely that someone would say It happens once in a blue car.

To generate the response to your prompt, the model makes a probabilistic calculation, seeing how close the tokens are and which token would be most likely to come next in human language (in this specific case, English.)

When probability is involved, there is always an element of randomness, of course, which means that the answers will not always be the same.

The response is thus generated token by token, following this path of probability arrows, optimizing the distance within the vector database.

There is no intent, only a more or less probable path.

The more times you generate a response, the more paths you encounter. If you could do this an infinite number of times, at least once the model would respond: "It happens once in a blue car!"

So it all depends on what's inside the cube, how it was built, and how much distance was put between one token and another.

Modern artificial intelligence draws from vast databases, which are normally filled with all the knowledge that humans have poured into the internet.

Not only that: the larger the vector database, the lower the chance of error. If I used only a single book as a database, the idiom "It happens once in a blue moon" might not appear, and therefore not be recognized.

But if the cube contained all the books ever written by humanity, everything would change, because the idiom would appear many more times, and it would be very likely for those tokens to occur close together.

Huggingface has done this.

It took a relatively empty cube (let's say filled with common language, and likely many idioms, dictionaries, poetry...) and poured all of the AO3 fanfictions it could reach into it.

Now imagine someone asking a model based on Huggingface’s cube to write a story.

To simplify: if they ask for humor, we’ll end up in the area where funny jokes or humor tags are most likely. If they ask for romance, we’ll end up where the word kiss is most frequent.

And if we’re super lucky, the model might follow a path that brings it to some amazing line a particular author wrote, and it will echo it back word for word.

(Remember the infinite monkeys typing? One of them eventually writes all of Shakespeare, purely by chance!)

Once you know this, you’ll understand why AI can never truly generate content on the level of a human who chooses their words.

You’ll understand why it rarely uses specific words, why it stays vague, and why it leans on the most common metaphors and scenes. And you'll understand why the more content you generate, the more it seems to "learn."

It doesn't learn. It moves around tokens based on what you ask, how you ask it, and how it tokenizes your prompt.

Know that I despise generative AI when it's used for creativity. I despise that they stole something from a fandom, something that works just like a gift culture, to make money off of it.

But there is only one way we can fight back: by not using it to generate creative stuff.

You can resist by refusing the model's casual output, by using only and exclusively your intent, your personal choice of words, knowing that you and only you decided them.

No randomness involved.

Let me leave you with one last thought.

Imagine a person coming for advice, who has no idea that behind a language model there is just a huge cube of floating tokens predicting the next likely word.

Imagine someone fragile (emotionally, spiritually...) who begins to believe that the model is sentient. Who has a growing feeling that this model understands, comprehends, when in reality it approaches and reorganizes its way around tokens in a cube based on what it is told.

A fragile person begins to empathize, to feel connected to the model.

They ask important questions. They base their relationships, their life, everything, on conversations generated by a model that merely rearranges tokens based on probability.

And for people who don't know how it works, and because natural language usually does have feeling, the illusion that the model feels is very strong.

There’s an even greater danger: with enough random generations (and oh, the humanity whole generates much), the model takes an unlikely path once in a while. It ends up at the other end of the cube, it hallucinates.

Errors and inaccuracies caused by language models are called hallucinations precisely because they are presented as if they were facts, with the same conviction.

People who have become so emotionally attached to these conversations, seeing the language model as a guru, a deity, a psychologist, will do what the language model tells them to do or follow its advice.

Someone might follow a hallucinated piece of advice.

Obviously, models are developed with safeguards; fences the model can't jump over. They won't tell you certain things, they won't tell you to do terrible things.

Yet, there are people basing major life decisions on conversations generated purely by probability.

Generated by putting tokens together, on a probabilistic basis.

Think about it.

FIC WRITING REVIEW 2023

Thanks for the shoutout and tag @gammacousin

Rules: Feel free to show whatever stats you have. Only want to show Ao3 stats? Rock on. Want to include some quantitative info instead of stats? Please do this. Want to change how yours is presented? Absolutely do that. Would rather eat glass than do this? Please don’t eat glass but don’t feel like you have to do this either. (Copied and pasted)

Words and Fics

Word Count:

115,039

Fic Count:

7 started and published, 1 continued from 2022, 1 written but not published.

The MCU Rewrite Series

The Philindaisy Playlist Series

Most productive month:

November based on fics published and word count

Top Five by Hits

The Ultimate Fix-It Fic - 2,033

You're On Your Own, Kid - 859

Here Comes The Sun - 623

The Second In-Between - 587

Somewhere Only We Know - 272

Top Five by Kudos

The Ultimate Fix-It Fic - 42

Here Comes The Sun - 31

You're On Your Own, Kid - 20

Somewhere Only We Know - 17

The Second In-Between - 15

Fandom events:

None?

Upcoming Plans:

To Publish:

How Sweet It Is To Be Loved By You (AOS// Philindaisy) and conceptual series with that

Other Ideas:

None, brain empty

Reflection

This year, I worked on a lot of stuff by maintaining a regular habit of spewing garbage out by quantity and not quality, and went over to the AOS boat since I missed feeling sad over fictional characters who die many, many times.

I want to say that I'm proudest of the weekend I wrote all of Somewhere Only We Know because I wrote the entire fic of 15k+ words in less than four days, but I like the fluffiness of I Will because usually I end the story with people dying and instead they got engaged and I'm happy about that.

I'm glad I got the chance to work on a project per month because my writing has improved drastically from when I first started out writing (at 11, Drarry 💀 with zero paragraph breaks) and actually publishing things (at 12, OC/Draco [kill me now] and sprinkles of Brutasha) to now, where my English teachers actually compliment my writing style and how I format and proofread even though I don't ever proofread. I hope I can actually channel all of this fic writing to write a novel this year.

Thanks for the reads and the tag, and sorry I'm late to the party; happy belated New Year, y'all!

@bubbletealife if you feel like it go ahead but I know you don’t write too much

Mini chapter of a christmas thing because I'm too blocked for anything else so here are crumbs @daisy-may98

We all know what it is though lmao

toast post what have they done to you

Haikyuu boy headcanons: “moving out” edition

Hinata— surprisingly self sufficient. Cooks and cleans as easy as if it were breathing. Once he gets a schedule up, he’s on fire and no one can stop him. Is usually the host for smaller gatherings with his friends.

Kageyama— cannot cook to save his life. Calls Suga for help the first night in realizing he’s got no food and unfortunately Suga can’t cook either so he brings Daichi along, and Kageyama spends his first night with his friends. Hinata pouts when he finds out there was a party and he wasn’t invited.

Daichi— moves out the first chance he gets and into an apartment close to his university with Suga and starts up this roommates to lovers sort of deal. Cooks like nobody’s business. Sucks at laundry. First time he tried, he made all of his and Suga’s white shirts pink.

Suga— loves cooking. Everyone else does not love Suga cooking. Calls Daichi his “saviour” for knowing how to follow a recipe. Otherwise, he’s pretty eager to move out of his parents’ house and house hunted with Daichi for months before they settled on this one.

Oikawa— didn’t want to move out but was convinced once Iwaizumi told him that technically his parents couldn’t limit his volleyball hours anymore (a scam, since it was Iwaizumi limiting his volleyball hours now). Cleans to self soothe and gets scolded for it. Only knows how to operate a rice cooker.

Iwaizumi— didn’t necessarily want to move out but decided that he had to because he was an adult now and should experience new things anyway. Moved in with Oikawa specifically to monitor his behaviour and make sure he didn’t overwork himself or die of starvation. Totally not because they were best friends.

Ushijima— was already half moved out because of the Shiratorizawa dorms, so his own apartment seemed the natural next step. His family got him one and told him that once he gets stable footing, he’d be paying them back. Hosts gatherings at his place because it’s big enough but he’s not really the host. Usually it’s either Tendou or Semi.

Tendou— asked Ushijima whether he wanted to live together once they graduated when they started second year, but didn’t expect him to remember and go through with it. Bakes at disgusting hours in the morning. Sleeps during the day. Doesn’t do laundry.

Semi— lives alone in a studio apartment paid for by his dad. Self sufficient enough to survive but spends 90% of his time at Ushijima and Tendou’s place anyway. Eats their food. Though he does disappear into hyperfixation phases where he doesn’t emerge from his apartment for days at a time because he’s writing songs.

Kita— doesn’t move out of his house… ever. Lives at home for uni, lives at home when he works because his work is at home. Cooks and cleans on the daily schedule he made when he was ten years old and Does Not Deviate. Goes out once a week with his grandma for groceries.

Atsumu— cries when Osamu says he’s moving before realizing that he would be moving because of volleyball anyway. Takes surprisingly good care of himself because he “has to be at his best to do his best”. Frequents Onigiri Miya and crashes on Osamu’s couch three days out of the week.

Osamu— lives above the first Onigiri Miya location and obviously cooks for himself and Atsumu. Makes Atsumu do his laundry in exchange for free food. Calls their mother once a week and sends Suna embarrassing photos of Atsumu while he’s asleep.

Kuroo— moves out for college into a dorm for his freshman year but into an apartment with Kenma once Kenma graduates. Calls himself the house husband even though Kenma’s the one sitting at home all day and juggles his chemistry homework with making sure his best friend doesn’t starve.

Kenma— did not want to move out at first but once his company expanded he realized he needed his own space. Went house hunting with Suga after Suga’s HGTV phase and ended up picking the first one they went to. Forgets to eat so Kuroo makes him finish dinner before he gets to go back to work.

Bokuto— moves into MSBY dorms when he gets signed. Gets mopey within the week. Goes back to Tokyo to visit Akaashi whenever he gets the chance and does this so often that Akaashi’s parents just gave him a key and told him to go nuts.

Akaashi— lives at home during college but moves in with Bokuto once he graduates. Does majority of the disaster management in the house and persuades Bokuto not to try making soufflé at 2am. He doesn’t do any of the cooking and cleaning because Bokuto says that he’s busy enough and “needs to practice for when Agaashee becomes a famous busy author”.

UNCALLED FOR

YOURE RIGHT BUT THAT WAS FLIPPIN UNCALLED FOR

Me: *on tumblr, minding my business*

My next chapter, sitting unfinished:

Avengers: Age of Ultron 2015 | dir. Joss Whedon

*incomprehensible screaming at myself because it's past my bedtime and I'm hungry*

Oh god making things. Writing things. Semishira. Oh god. Um. Happy Semishira day fam!! Here's a fic! Goodnight.