#please stop using ChatGPT for literally everything!!

[fillory-and-furtherr]

“I write fics with AI”

So you didn’t write it actually hope this helps

“I get help for my fics with ChatGPT ”

Okay well I spend four+ hours a night in a writing K-hole as a pure act of passion while I kick my feet and giggle about dialogue because I’m insane and that is always going to be better than whatever the fuck AI generates.

Using AI to create “art” isn’t just unethical but it’s cowardly- it takes the joy out of living and out of humanity.

To be human is to create to have passion to do things solely because we enjoy them stop letting AI take away the joy in being human- please.

Make bad art make any art at all but make it yourself- partake in the human act of creation. (Also it’s killing the environment)

[bloomzone]

Okay, so I just recognized some mistakes I’ve been doing lately literally just saved myself from repeating them again. And honestly, I noticed a lot of other students are making these same mistakes too. So here’s Part 2 of my “Study Mistakes I Did But Saved Myself Before It Was Too Late.” If you relate to any of these, take it as your sign to stop before it’s too late.

part 1

lesson 1 : the “failure” mindset.

Right now, I’m literally in this crazy state. Like the failure mode. It's insane , I do a million past papers, then I tell myself, “Yes, you did well today.” But I wake up the next day with the same mindset: No, girl, you need to do more. Like, what if all those questions don’t show up on the exam?Yes, stress can help it drives you sometimes. For me, it worked most of the time. That fear of failure push me to study harder. But don’t let failure control you. I really wish I could change my mindset, especially before big exams or finals.We always think, what if I fail instead of what if I succeed? So please, try not to stress too much. Even small efforts count. Stay motivated. I know discipline is important, but we’re human. Especially high school students there’s pressure from everywhere. Everyone expects something from you.So try to motivate yourself. Make your breakfast in the morning. Watch something motivational when you wake up. It will remind you why you're doing this.You’re a student. You need to succeed. But sometimes we lose that passion . So keep some backup motivation ready. Think: What if I succeed? Always and Study hard. Do everything you can so you can feel good and proud about urself later.

lesson 2 : using AI for essays.

Please, please, please stop using AI if you’re still doing it especially for essays idk about u use ai or not it's ur life idc but if u are a STUDENT hear me out If your exams are still a bit far and you’re using AI to do your homework, especially your essays, stop now.I swear, some classmates of mine used AI the whole year. Especially for French essays. I know not everyone is great at French even if it’s our mother tongue , some people treat it like a second language tooThere’s this one girl in my class she used ChatGPT for every single argumentative essay. And our teacher praised her: “Oh my god, I love your writing!” and gave her plus points. I got mad. Like… I wrote my essay 100% by hand, all real. And then the teacher accused ME of using AI? I was like Girl wtf 💀💀💀 this is 100% mine.And in the actual exam, that girl got 13/20 and i I got 19/20 and then the teacher believes me after •́⁠ ⁠ ⁠‿⁠ ⁠,⁠•̀.f you use AI all year, you’ll look so stupid in the exam. So please, even if you need help, don’t rely on AI. Just read articles, watch videos on YouTube, act like AI doesn’t exist.Read more, watch teachers online, documentary anything that helps you build real knowledge and vocabulary. You’ll get better and your essays will show it.

lesson 3 : Not organizing your notes.

This one is real. Please don’t do what I did: I lost every single note from first semester. I still don’t know where I put them.So if you’re someone who writes by hand, please always keep a folder next to you. As soon as you finish a paper, put it in the folder. Keep your folders in a place where no one touches them.The only thing that saved me I take pictures of my notes. So please, every time you write something important and feel like you might lose it take a photo. Just in case.

lesson 4 : Not sleeping early.

I’ve been doing this mistake recently, during prep week when we’re home from school. The stress is too much. Not because I can’t sleep I actually fall asleep in minutes. The problem is I don’t sleep early. I have no fixed schedule.If you study in the morning, you need to sleep early. I wake up around 5 or 6 a.m., but I sleep at midnight. That’s just 4–5 hours of sleep. And it’s not enough. I get headaches. I burn out.Today, for example, I woke up at 6 after sleeping at midnight. I had the worst headache. I made coffee but didn’t even drink it.Also, if you get headaches easily, be careful with coffee. Drink water first. Move around a bit before having it.So then I tried studying but couldn’t. I opened a YouTube video to study, but I couldn’t focus. I ended up lying down, playing the video on my phone, and falling asleep again with the teacher’s voice still playing I felt like a loser fr it made me so sad. So please, if you work best in the morning, be in bed at LEAAAST by 10 p.m.And if you’re a night owl and study until 2–3 a.m., please sleep long enough after that. Your brain needs it.Just be organize your sleep schedule. Everyone has their best time to focus just build your routine around it. The time will pass anyway, so spend it right.

stay healthy everyone 💭

@bloomzone

[adaobiiii]

Felinette February

So i thought there was already one shots for Felinette in February (cause they start with the same letter...) but apparently there isn't. So! Instead of making some up, I made ChatGPT come up with prompts for me!!!

Here They Are !!!

Please tell me what you think...

Week 1: Valentine’s Season Begins

February First Sight – Felix and Marinette meet unexpectedly at a flower shop, both buying flowers for themselves.

Library Whispers – Felix finds Marinette asleep in the school library and debates whether to wake her up or let her rest.

Skeptical Suitors – Marinette’s friends assume Felix is up to something when he offers to walk her home.

Fake Love, Real Feelings – Marinette needs a date to get her friends off her back, and Felix volunteers, but neither expect to enjoy it.

Roses Are Overrated – Felix insists red roses are unoriginal, so Marinette challenges him to find a better Valentine’s gift.

Not Your Average Love Letter – Marinette receives an anonymous love letter, and Felix is suspiciously quiet about it.

Stray Cat Serenade – Felix catches Marinette singing on her balcony, and she catches him listening.

Week 2: Sweet & Spiteful Confessions

Rainy Day Rescue – Felix begrudgingly shares his umbrella with Marinette when they get stuck in a downpour.

Betting on Love – Adrien bets Felix he won’t be able to make Marinette blush, and Felix takes the challenge way too seriously.

Stitching Up a Heart – Marinette helps Felix fix a torn jacket, and he finds himself entranced by her concentration.

Jealous? Who, Me? – Felix and Marinette keep getting mistaken for a couple, and Marinette teases him about how flustered he gets.

Valentine’s Disaster – Marinette’s Valentine’s cookies get ruined, and Felix secretly makes a new batch for her.

A Dance with Fate – A masquerade ball leads to an unexpected and heart-stopping moment between them.

Last-Minute Valentine – Felix panics when he realizes he actually wants to give Marinette a Valentine’s gift but has no idea what to get her.

Week 3: Tension & Softness

Say That Again? – Felix accidentally calls Marinette by a pet name, and neither of them knows what to do about it.

Through the Camera Lens – Marinette catches Felix taking candid photos of her and demands an explanation.

Your Smile, My Weakness – Felix realizes that Marinette’s smile has way too much power over him.

Accidental Date – A misunderstanding leads to them spending the entire day together, much to Felix’s horror and Marinette’s amusement.

Touch-Starved – Felix is not used to casual affection, and Marinette is determined to change that.

Midnight Calls – Felix gets an unexpected late-night phone call from Marinette, and he doesn’t have the heart to hang up.

Falling for You (Literally) – Felix tries to be smooth, but he ends up embarrassing himself instead.

Week 4: Serious & Soft Moments

The Cat’s Out of the Bag – Felix accidentally reveals something personal, and Marinette reassures him in a way no one else has before.

Stargazing Confessions – They find themselves alone under the stars, and Marinette asks Felix what he truly wants.

Soft as Silk, Sharp as Thorns – Felix realizes Marinette is stronger than she looks, and he respects her even more for it.

Comfort in the Chaos – After a rough day, Marinette finds unexpected comfort in Felix’s presence.

Lost & Found – Marinette loses her sketchbook, and Felix goes out of his way to return it.

Handwritten Honesty – Marinette finds notes from Felix tucked into her things, and she’s not sure if he meant for her to see them.

The Moment We Knew – One simple, seemingly insignificant moment makes them both realize that something has changed between them.

February 29 (Leap Day Special!)

Once Every Four Years – Marinette asks Felix if he would take a leap of faith for her, and his answer changes everything.

I just copied and pasted them and haven't actually read over them yet... I'll start from D1 (even though it's the 9th already :P)

[shadowofthesun123]

How I use AI in my artworks and what I think about AI in general

I think it’s time to have this conversation because questions keep coming up more and more often.

I know that many people here don’t like AI or completely ignore it. In this post, I’d like to share my personal view on the topic and also answer the question of whether I use AI in my artworks.

Let me start with the simpler question. The answer is “yes,” I do use AI tools in some of my artworks. However, none of them can be considered fully AI-generated.

Sometimes, I use AI to create backgrounds or silhouettes, but I completely change the faces, like here: https://pin.it/2RUhZKz4S.

Other times, almost everything is AI-generated, for example: https://pin.it/N8apDmpfz.

Sometimes I use face swap technology, faceapp for hairs, etc.

I also have arts created fully without AI, but the faces there are based on real photos of Hunter and Jenna and edited with filters.

For example:

https://pin.it/1DZvBcKVw, https://pin.it/5wAdQvIFe, https://pin.it/19G1mffb6.

In general, I use anywhere from 3 to 5 or more tools for my artworks, including AI.

Some cases wouldn’t be possible for me to create without AI because I simply don’t have the technical skills yet. Others, on the contrary, are impossible to get from AI, no matter how hard you try, so I go back to good old collage techniques and hand-drawing.

Now, about my attitude towards AI overall.

You may like it or not, but it’s now part of our lives, and ignoring it is no longer an option. No AI boycott will stop its development. Almost any app nowadays has inbuilt AI tools.

I’m currently on maternity leave, but even in my professional field, AI is already widely used, and when I return to work, I’ll either need to know how to use it or fall behind the industry. That’s why I’m studying AI, and I do enjoy the possibilities it offers.

That said, I don’t deny there are issues - copyright concerns, as well as serious problems with fraud and fakes it can enable. As a mother, I’m terrified by the thought that someone could use my voice or image against my children. Now we have to be much more careful with what information about ourselves we put online - photos, videos, audio.

AI also creates a lot of low-quality content. But there was also low-quality content before AI.

As for writing, I don’t use AI there. It’s terrible at creative writing, plus that’s a space I want to keep entirely my own.

However, I do use ChatGPT to help write and translate tons of bureaucratic emails in a foreign language (which is, unfortunately, necessary when living abroad), and this saves me time to do what I actually enjoy, including writing.

One of my favorite YouTube videos from last year is about scientific breakthroughs of 2024. It’s in Russian, so I won’t link it here, but it mentions how AI, thanks to its ability to process massive amounts of data and detect patterns, has already helped scientists with several discoveries. For example, it helped identify a substance that might be effective against antibiotic-resistant bacteria. It also contributed to a breakthrough in forensics related to fingerprint analysis, among other things.

Of course, AI makes mistakes, it’s often inaccurate, and everything it produces needs double-checking. But it can also be useful. That’s why I disagree with people who view AI as pure evil.

I use it for art, video projects, web design, and official correspondence, languages learning. Even though I try to keep up with AI trends, it evolves so fast that it’s impossible to follow everything - it updates literally every day.

It can be used for bad things (fraud), but also for good (scientific research). I’d say how I use it is somewhere in the neutral zone. I hope it doesn’t harm anyone - and maybe even makes someone happy.

So, I will continue creating art, including with the help of AI, and in my next post, I’ll tell you about my experiments with AI videos.

I’m open to discussing this topic - but please, be polite.

[lixies-favorite-cookie]

okay so lately I've been thinking to post like fluffs or something and I already have a lot of ideas. So, I wanted to ask if you have any tips for new writers? I've been really inspired by your works and I wanted to start posting as well. <33333 (I'm new to asks so I have no idea if I did this right 😭 but tysm <3)

Me?!?! Little old me for writing tips?!?! I'm actually so honored, thank you so much 🥹 I actually put a lot of thought into this and all the different things I've learned along the way, so I hope this helps! And if you have any more questions, feel free to ask!!

Read. Okay, so I know that this sounds so cliche, but reading is literally a lifesaver for writing. I read a lot of fanfictions that pertain to the fic I'm writing to see how others did it and help guide the way I do it. Does that make sense?? I hope so 😭 Obviously don't copy anything, but like for my newest fic (that's coming out hopefully soon if I can beat this writer's block!!) rewrite the ending in every lifetime, I felt really inspired by @.forlix's setup for her fics and @astraystayyh's when the snow falls, we fall apart for act 3 of it. It also really helps broaden your vocabulary and allows you to get a feel for certain ways to describe body movements, tones, and facial expressions (something I've always struggled with!!).

Punctuation. I've always been SO BAD at punctuation because I was never taught how to properly use it at my school. Also, when I'm writing, I never use it because my thoughts are so flighty I feel like if I stop to put a period or something, everything will leave me. SO, I will always make sure to copy and paste my work into a free punctuation checker like Scribblr (it's pretty good, though there are a few things it will miss), but the one I like the most is ChatGPT. Like, I'm going to be so fr, it has SAVED my ass, but the only thing is you have to paste whatever you want punctuation checked, and if you don't want Chat to change the way you say something, just in parentheses say (Can you punctuate this without changing any of my words please). This will still check punctuation and spelling but will fit your personal voice.

ChatGPT. I literally don't care what anybody says; this resource is useful asf. Like, I use it for synonyms, definitions, and if I don't know the word I'm trying to think of, I'll input something along the lines of "Chan (word) his arm around her waist, leaning down to press a kiss to her head." (What would a good word be for the (word) above?) It's SUPER helpful for getting stuck on words like that!!

Ask for help. Like, I'm so serious, I will ask many of my very amazing mutuals (quite often, I might add) about the way I worded something, if my poetic quips make any sense, which way I wrote something sounds better, bounce ideas off each other; honestly, just about anything I can think of that has to do with writing, I have asked the amazing @yongbun, @jeonginsleftcheek, and @luvtak. So if YOU need anything, always feel free to ask me or anybody else that you feel comfortable with :))

Feel it. Okay, so this is helpful for getting stuck on really deep emotional scenes. I would get so into trying to make it look pretty that I didn't get down into the actual emotions. It's really helpful to just set your computer down, shut your eyes, and imagine the scene. Feel everything they feel—the anguish ripping inside their very soul, the way it feels like an earthquake has just ruptured in the base of their spine, the universe seems to be tearing apart with every breath they take. I don't know, just me personally, I love anything that pertains to the soul and super dramatic emotions. It makes me feel a lot less alone about my deep emotions :)

Writer's jealousy. I had to write something about this because being an artist, you are always going to see somebody better than you, and that can cause some really nasty conflict inside of you. Sadly, there is nothing I can do to stop that, but I can say really try not to let it get to you and instead view it as a learning experience from those people and how they write. Also, never compare yourself to others because all writing, whether good or not, is your most vulnerable parts because it is unique to you. And also, try not to please everybody—I promise you won't be able to. The same way people pay millions for a piece of artwork, another person wouldn't pay a penny for; it's just simply the human mind, and that's okay. But this also ties into Write for yourself. Nobody else. Just yourself. When you're writing, try not to even think that somebody else is going to view it. It will allow you to write what you really want but also get you to be a lot more vulnerable.

The first draft is shit. I'm so serious, every single one of my first drafts has been shit, and that's okay. You just have to edit your way up to something readable. Like, if you don't like how you wrote something the first time, that's okay. The first draft really is just you telling yourself the story, while the second, third, fourth, fifth draft is you telling others the story lol.

I don't think that there is anything else, but if I think of something, I'll definitely add it!! I hope that this helped you at least a little bit!! Thank you so much for asking me of all people for writing tips; it did wonders for my confidence lol.

[bio-mechan1cal]

Genuinely losing my mind over how normal using AI (chatgpt specifically) is, but ESPECIALLY at a college level. I know way too fucking many people who use it to write papers or even simple 250 word discussion posts!!!! I just recently had to deal with finding out that a childhood friend was stealing my work, putting into chatgpt, and rewording it to pretend it was her own. This was a short answer discussion post about literally any topic we desired, no requirements other than length. Like we've just gotten to the point where we are blantantly stealing from other people and not even being ashamed of it anymore. I am so insanely concerned for the future of like everything considering the next generation of what is supposed to be professionals is a bunch of kids who don't even know how to write a paragraph on their own. I'm currently a STEM student looking into going into the research side of things for my career (which is a whole other topic considering NIH defunding and such) and the amount of students who use AI to write their research papers or GRANT PROPOSALS !??@?!?! actually disgusts me. Stop going into a career based on the need for original, diverse, creative thoughts and ideas just to make a computer do it for you (and do it worse, frankly). This does not increase the current body of knowledge at all, which erases the entire point of being a scientist. Please for the love of God bring back the desire for academics and be passionate about learning and discovery again

[rpcburnbook]

On the general note of AI, I would, and will, and do STRONGLY caution all writers to avoid using ChatGPT or other generative AI as, quite literally, they are and will be stealing everything you think is important to an author. Please, do not feed it, but more importantly, protect your own interests and your own SELF, because AI will, without a doubt, erase every single one of us as artists and we will lose every image and credit we deserve through built-in copyright laws. Please, just stop.

~

[project-do-over]

Violet_M00n is available

Violet_M00n: Heyyy sorry for the delay! Got caught up with work stuff...

Bookwyrm1982: literally no time has passed.

Violet_M00n: Right. Right. Habit, I guess.

Violet_M00n: Anyway, what else did you want to know about the future?

Bookwyrm1982: Well you said you had AI, right?

Violet_M00n: Well... Yes and no. We have programs people *call* "AI", but they're really just advanced machine learning. They can't actually think or anything, but they can put together a surprisingly human sounding sentence, and draw things that could at first be mistaken for art.

Violet_M00n: But of course it's awful. The results are full of factual errors or have way, way too many fingers, companies are trying to use it to replace creatives, and it burns a ton of energy doing essential nothing of value.

Violet_M00n: So could you go on ChatGPT and talk to a convincing facsimile of a human but underneath it's just a more advanced version of Dr. SBAITSO.

Bookwyrm1982: that's a shame. But then again at least you don't have to worry about them taking over the world, right?

Violet_M00n: Luna, at this point I'd welcome our robot overlords. Better than the fucks we have running things these days.

Bookwyrm1982: Are things that bad?

Violet_M00n: *sighs* no, I suppose not. I still have a job, a family, I can exist in public without fearing persecution, and I'm mostly free to do as I please.

Violet_M00n: But trust me when I say the people who very much want to take that away have much more power than feels comfortable.

Bookwyrm1982: that sounds scary though.

Violet_M00n: More enraging than scary, really. Just so many people who can't or don't want to see things from anyone's point of view from their own.

Violet_M00n: Well, that, and capitalism.

Bookwyrm1982: I thought capitalism was good though?

Violet_M00n: *sigh* we have so much to learn.

Violet_M00n: Honestly though, and you should be able to find these online if not in the library, but read some Marx.

Violet_M00n: It may not resonate a lot yet, but it will.

Bookwyrm1982: I always thought that Communism was a good idea in theory but it needed a global revolution to actually work.

Violet_M00n: You may be on to something there. And someday, hopefully in our lifetime, we may get there. But it's a long, long road. Especially here in America, where it's been used as a boogeyman for like 80 years now.

Violet_M00n: (55 for you)

Bookwyrm1982: Wow, that's.... I'm not sure I want to grow up now.

Violet_M00n: Well maybe your timeline will invent actual time travel and you can keep that wish. Luna knows I wish I could.

Bookwyrm1982: so

Bookwyrm1982: um

Bookwyrm1982: Can we talk about something more fun? Like, what's something good in your time?

Violet_M00n: Well Magic the Gathering is still pretty good.

Bookwyrm1982: We're still playing? I kinda lost interest and stopped following it a year or two ago.

Violet_M00n: Oh yeah, we're still playing, and the game is... Well, it's way different from your time but also at its heart the same.

Violet_M00n: Like it's still Magic but also there's D&D and cowboys and Gandalf, for some reason. It's cool but it's also kinda scary how much they're pumping out.

Bookwyrm1982: Oh that sounds cool! Is it just D&D and LOTR?

Violet_M00n: They've done a ton of crossovers, they call them "Universes Beyond". They've done, let's see...

Violet_M00n: Dr. Who, Warhammer, Assassin's Creed, Final Fantasy, they're doing Marvel soon, Transformers (those are Hasbro though so they were among the first), The Walking Dead, Fortnite, Stranger Things (you... Don't know about those yet, don't worry), um, lots more stuff too that I'm forgetting, but those are mostly in like five or ten card bundles.

Violet_M00n: Unlike LotR which was a full set, with boosters and everything. And the best selling set of Magic in all time, unless Bloomburrow has passed that already.

Bookwyrm1982: Really cool! You'll have to send me some pictures sometime!

Violet_M00n: I'll be sure to downscale them appropriately this time!

Bookwyrm1982: What else do we like? Is Star Trek still running?

Violet_M00n: It had a long break there where it seemed we weren't going to get any more Star Trek.

Violet_M00n: But then JJ Abrams (a director/producer of some renown) made a Star Trek movie that was meh, but good enough to get people interest in the franchise again.

Violet_M00n: Soon after that Paramount spun up Star Trek Discovery, which had a rocky start but Grew The Beard soon enough for them to greenlight Star Trek Picard. Then Lower Decks, Strange New Worlds, Academy, and probably one or two others I'm forgetting (not to forget Short Treks and Very Short Treks).

Violet_M00n: Prodigy! I forgot Prodigy!

Bookwyrm1982: The online service?

Violet_M00n: No, Star Trek Prodigy. It's a CG animated series for kids made by Nickelodeon.

Bookwyrm1982: You're making that up.

Violet_M00n: I swear, it's true. Lower Decks is animated too, but 2D, and it's for adults and probably the best thing Star Trek has ever created. It's hilarious!

Violet_M00n: SNW follows Captain Pike on the 1701

Bookwyrm1982: And Discovery?

Violet_M00n: Complicated! It starts out pre-TOS but... Spoilers! And Picard is... Also here!

Bookwyrm1982: Is that about young Picard or something?

Violet_M00n: Old Picard, but close.

Bookwyrm1982: Hey my mom... our mom... just told me to get off the computer so

Bookwyrm1982: ttys!

Violet_M00n: See you in literally no time at all!

Bookwyrm1982 is away

[romsabombs]

its that time of the month again!!!

malevolent part 50 the company notes

"and i see it" yeah YOU see it🙄

the woods are lovely dark and deep

so did the witch heal them or

the blackstone is our macguffin

take it?

companions??

one bite at a time :(

"i dont know what lay in store for us when that moment comes" im sure harlan does

"we continue on, like we always have"☹️

his shit eating giggle bruh

REALL only religious when your lifes in danger. didnt arthur say that at some point

YORICK :((( I MISSED U MY AUTISTIC SON☹️☹️☹️☹️☹️☹️

the others??

ITS BEEN A WEEK???😀

miles☝️

NOOO i liked the jaw clacking :(

"i wish i could see everything youve described" ohh honey

hes so hopecore pilled rn . something terrible is going to happen

oscars letter 😞💔

excitement😟

they yassified him

chekhovs sword....

do u think theyve explored eachothers bodies

"your presence will be missed" oh definitely

we're selling our daughter to onedirection💔

we dont need money we just need plot armor

I MISSED HIS VOICE TOO :( :(

the lighter mentioned

HAHA i missed him sm

"we'll just keep it in view" poor choice of words

these two men so mysterious its gonna be someone we know

"oldest" friends . in what way

OHHHHH OHH OHHGGOHHHGG

CRAZYY THIS IS CRAZY

NOOO DONT TELL HIM 😀

I CANT BELIEVEEEE HES BACK MY FAVEEEE LOSERBOY COWARD

how old is this guy bro

bro this is literally insane rattling the bars of my enclosure

LESTERRRRR 💪💪 ARTHURR

canon gayass!!!🏳️‍🌈 love wins

hes gonna die brutally i fear #buryyourgays

GAYASSES hashtag pride #boykissers

horse jumpscare

WHAAT autism . saw a movie four times be fr . also talking to john about seeing a movie is just so cruel

ok pause i just googled . ride em cowboy is a 1942 movie . but malevolent takes place in 1934ish . well not rn coz we're in medieval england but u get it. hmmmm . this is like that waylay song that wasnt released yet when the show takes place

ok back to the story . we're so friendship is magic pilled rn

"all of what remains good-" "myself included" mmmokay?🤨

this weakass twink🧍‍♀️ be fr youve done crazier things come on

"you left the prison pits" THATS WHAT I SAIDD

john wayne mentioned ...strange

not canon . john would think arthurs talking to him and say "who's wayne" trust🙏

"so long, evrard" ok either boykissers or hes gonna die

BOYKISSERS ok sorry

when is noel coming back is he here :( pretty please

technically 🤓 we're with five ppl rn

day of wrath mentioned

WHAT motherrr mother is mothering

why r we ...😟 shes gonna kill us

bro dont get off ur high horse u wont be able to get on again

she wins the idgaf war

OKAYYY WHAT diva

radio static.......

he just cant stop killing daughters

OHHH CUNTTT💅 motherrrr

😟mmmm

is it Lillith

OMGGGG LILLTHHGHH I WAS RIGHT

LILLITH I CWNT BELIEVE IT THIS DIVA

MOTHERRR MOTHER LILLITH

HAHA casually curses you

"are u helping us" dont be stupid😐

I KNEWW ITTTT I KNEW IT I KNEW IT

LILLITH KAYNE DAUGHTER CANON

guys im on top of the world rn

he called his own daughter a bitch in part 20

omg princess luna.. nightmare moon

omg shes a not!them

I DID NOT ALLOW ANYTHING☝️👺

MALEVOLENT MENTIONED

its the black stone

his dumbass

KILL HIMMMM DO IT‼️

YESSS FUCK HIM UPPPP QUEEN

harddd

"i look forward to watching you die" OKAYYYYYYYYY whos this diva

i was right about everything about lillith

antoine my husband who returned from the war . i trust him with his life

is there a mrs nyalathotep?? 👀

day of wrath weirdmageddon core

yorick u are like chatgpt unable to generate 9/11 images

bro no WOMEN are EVIL☝️ and theyre ABSOLUTELY RIGHT

YAAA HER FAVORITEEEE💥 SLAY

everyone hearing arthur mutter stuff under his breath- BRING BACK FAROE???? EXCUSE⁉️

ya no he loves lying asf

we have no choice :/ fucked up

"i dont think we should hope for that" okay😞

no one important 😟

ohhh???! 👀

A SERIF?? like a FONT?

"she can see- well not see" this happened to my good friend arthur lester

wait ur telling me yorick can just speak freely now🧍‍♀️ did i miss that

antoine casually dropping insane amounts of lore on us

john being actually useful for once

KELLINNNN☹️ MY FAVE

arthur being casually racist to yorick

😦 wuh oh . this is like being in trouble at your minimum wage job

we're just gonna tell this guy everything huh

"i trust you" rookie mistake

ummmm.😀

guys i dont think hes fully cured

"i can almost feel her gaze" HARDDD

"john wayne!🤓" "WHY👺"

was??? blud wants to be human too

hey girlll :3 whats uppp🎀

"you carry two souls within you" okayy

they both have plot armor then☝️

his heart is good :(☹️

oh hey finally some real hope

WHAT☹️ nooo 💔 diva dont die🕊️

shes spoiling the story bro

:(( oh :( ☹️ :( he wont give up

this is definitely about john

AW :( "but what do i know"

prison pits mention (again)

two broken men trying to hold hope

TONIGHT? ohh..😟

"come now, dont stop yapping on my account😁" ok bro :|

oh no he doesn't know anything hes just a freak now😦

yorick starting his human arc

His Own Thing

wait yeah no coz everyone knows about yorick hes freely yapping why were we shocked alya knew about him

[caffeinated-composure]

GenAI prompt rant

Okay so today I came across something that, to me, perfectly encapsulates one of the biggest problems with genAI in general and the use of LLMs in particular.

Someone was joking about how they were testing their LLM and asked it to produce a random number. That question returned 42 like half the time. They thought it was funny.

Someone else responded by suggesting that this could be "fixed" by adjusting the temperature settings or inserting additional stipulations into the prompt settings, such as "remember the list of numbers already provided, and provide a number not on that list."

That won't fix shit.

GenAIs are basically just giant statistical machines. They read everything on the internet and spit out the most statistically common sequences of words. They don't understand the content of the responses they provide - they literally just repeat what they've read elsewhere.

This means, among other things, that they will never actually produce a random number in response to the prompt "Give me a random number." What they will do instead is provide an answer based on the ways other people on the internet have answered that question. Unsurprisingly, people on the internet respond with "42" far more often than other numbers, so the LLM will also do that.

Adjusting the temperature will, at best, ensure that 42 is appears only somewhere in the first 5 responses, rather than being the first response the vast majority of the time. But it will always appear somewhere near the top of a list of "random" numbers.

Telling it not to repeat numbers will... well, first, it just means these numbers wouldn't be random anymore, even if they would have been otherwise (which they weren't). But it only succeeds in masking the problem.

Because here's the thing - before adding those stipulations, someone messing around with the LLM would have noticed that it kept returning 42, and would be able to tell pretty easily that these numbers probably aren't actually random. They'd go "Oh, hey, this is a way that ChatGPT fails, I guess ChatGPT isn't good at this sort of thing." And they'd, you know, learn something about what ChatGPT is or isn't good at doing.

But instead, now, this hidden prompt manipulation means that the user is unlikely to ever notice that there's a problem. Sure, 42 appeared - but it was alongside a set of other numbers that look mostly random. You'd have to come back to multiple sessions and ask the same question each time to even have a chance at noticing that these lists of numbers have a weirdly high number of matches across different sessions. Most people aren't going to notice anything. It makes it look, to your average user, like ChatGPT is capable of providing random numbers.

And there's a whole fucking industry of people out there whose entire job is learning how to prompt ChatGPT in a way that makes the answers look good. Sometimes, this actually means the answers are better, but often, it's shit like this. And you don't even know because this prompt manipulation is happening behind the scenes - when you talk to the IT helpdesk chatbot at work, it might be doing this kind of prompt manipulation behind the scenes, while you think you're asking it directly about random numbers.

It's wildly misleading, and it creates a false impression that LLMs can do a lot of things they most assuredly can't. It's no fucking wonder everyone thinks AI is going to be able to replace every single job in existence in the next few years.

Please, for the love of god, stop making prompts like this. Stop misleading people by making your LLM look like it's giving valid answers when it's not. I can't stress enough how deeply frustrating it is to constantly have to explain that no, ChatGPT is not a fucking doctor, or a lawyer, or anything else that requires actual understanding. ChatGPT is an actor that has played a doctor or lawyer on TV and is very very good at using all the words that sound right to a layperson, but if you hire an actor to represent you in court because they sounded good, you're an idiot. But as much as I would like to blame the idiot who hired the actor as a lawyer, I'm also gonna have to blame the person who enabled all of this, the one who provided the actor with all the scripts required for them to fake it convincingly, and who then ran around and told people that this actor actually is a lawyer.

Just... fucking stop lying to people about what LLMs can actually do. Please.

[wizard-appreciator]

Hi! Actually everything that person just said is true! AI IS extremely unethical and is actively destroying our planet! Here are a couple of articles that talk about how much fucking energy AI consumes! Also you could try doing your own research sometimes you dipshit

https://www.theguardian.com/commentisfree/2023/dec/23/ai-chat-gpt-environmental-impact-energy-carbon-intensive-technology

https://www.technologyreview.com/2024/05/23/1092777/ai-is-an-energy-hog-this-is-what-it-means-for-climate-change/

https://www.businessinsider.com/chatgpt-generative-ai-environment-climate-change-global-emissions-2023-9

And here's even a couple of YouTube videos if you don't like reading! But i would strongly advise you to do your research yourself instead of saying stupid bullshit online!:

https://youtu.be/RT8-GlA_rkM?si=j_h0xPr0PI6uAsSg

https://youtu.be/SkWzwoDIGmE?si=k_lOQgRWmf-Q9oyk

And here are even more articles! And just so you know, i will not be doing this again! Next time you do your homework yourself, buddy!

https://www.techradar.com/pro/generative-ais-energy-demands-are-accelerating-the-climate-crisis-top-researcher-warns-of-environmental-impact-of-googles-new-search-feature

https://www.cryptopolitan.com/ai-impact-on-climate-change/

Also, yes! By literal definition Hatsune Miku is an artificial intelligence (AI) in canon. She might be a vocaloid in our world, but in the canon of the vocaloid world she is an artificial intelligence that is a singer and a performer! And Miku supports human and non-human artists, but no one can call the AI we have in our current state "an artist", because for now it is merely replicating the art that has already been made, and it doesn't even have consent from the original artists, and also Miku is the owner of the world (click to open link), so i don't think she would be happy if she saw that her fellow AI's are destroying it. So yeah! Go eat shit ig idk. And id also like to see your sources, so Go diva! And by "Go" i mean Go kill yourself!

And yeah, i don't really care if people use ai for fun or personal projects that no one will see or really care about, that don't consume a lot of resources, but i DO care about people spreading misinformation! I wouldn't have even looked at this post twice if you haven't said anything about AI not doing any harm or Miku being cool with AI. Please do research yourself. And stop using AI.

@wuxingbianji