#AI is a good and useful tool!

[adaines-furious-feast]

Gonna write a novel where the leaders of two countries keep sending more and more heated emails to one another despite advisors telling them not to and all communications to them is really soulless and hard to make out who's side other countries are on. And then they nuke each other.

And it turns out that both leaders were trying to send peaceful messages but they relied on AI to communicate because apparently people these days can't even write an email without going to frigging grammarly to do it and this is how the robots are going to kill us all.

[bogkeep]

there's something so deeply dystopian to me how tech companies don't understand that a forced convenience is not a convenience at all. i'm sure autocorrect is helpful for many, but a function that forcibly changes my actual written words and punctuation is taking away my language. photo filters can be nice but i need to choose using them myself or else i have lost the ability to take the picture i want. i don't want a machine to draw or write for me. taking away the option for me to do things manually feels like violence!!!! all this talk of endless opportunity, why are you RESTRICTING me

[alelelesimz]

you guys do realize that AI has been a thing for a long time as an algorithm tool that is used in things other than generative AI, right? many games have been using that type of tech for a long time, but the term "artificial intelligence" just got popular recently

[carlyraejepsans]

HELPP i found out about character AIs and im losing it over calling deltarune sans "a more manipulative personality" 😭😭 like whatt we've only seen this man like 2 times

sorry im being a hater hgfjdks but you're the sans understander

lmaoo character.ai, look where the fall of ai dungeon got us, smh. thank you for the compliment though!

[sincerely-sofie]

Me: Ahhh yes. I have finally finished the first truly completed draft of Better the Wool's outline. It's got goofy scenes and gut-wrenching plotlines, tragic backstories and a central theme of power and control versus trust and uncertainty... plus shipping! This project is taking a hot minute, but thankfully I don't have anything demanding I finish prepping the fanfic ASAP :>

Meanwhile, the looming urge to participate in NaNoWriMo again in November:

[abortionado]

I really think it should be illegal for publicly available generative AI to create hyperrealistic photos/videos of children. Firstly, it's trained on images of real children, who inherently have not consented to their image being shared online, and secondly name a non-nefarious purpose that this technology would serve. How does anyone besides pedophiles benefit from the ability to generate an extremely realistic video of an infant

[lgbtlunaverse]

"AI is bad and copyright law might be able to stop it, so we should support more copyright laws–" [EXTREMELY LOUD INCORRECT BUZZER]

"Copyright is bad and AI is in violation of copyright, therefore being against AI is bad because that means you oppose common ownership of ideas–" [EVEN LOUDER INCORRECT BUZZER]

[jellypawss]

any response to butteredfrogs post?

I did use ai for that image before I knew how shtty ai was and haven't used it since and I apologize.

As for posting text written by chat gpt, I had chatgpt write an idea for a sims 4 expansion pack, similarly to how students use chatgpt for homework and general knowledge, which I don't see a problem with.

[llatimeria]

i just found out about one of the most powerful adhd tools i've ever seen online. i think i'll actually use this one in real life. holy shit

features:

Magic ToDo, which helps you break down tasks into smaller tasks

Formalizer, which helps you adjust your tone depending on your situation - it can make you sound more or less formal, or more or less decisive/clear

Judge, which assesses your tone to help you understand how you're coming across, or you could put in someone else's text to see what it thinks they meant

Estimator, which can help you guess how long a certain task is going to take you

and last but not least, the Compiler, which takes an amorphous braindump and magically rearranges it into a list of tasks, which can then be sent to the Magic ToDo list and broken down into more managable pieces!

[fixyourwritinghabits]

Let me elaborate just a tiny bit on what I mean by using chat AIs as tools because I'm already seeing some alarm in the tags. To be clear, Chat AIs in their current form... Suck, basically. They've improved a lot, but they still suck, which is why you should not use them to replace research or work. You don't know if the AI is giving you accurate information. The AI doesn't know if it's giving you accurate information, because that's not what it's designed to do.

However! It's still filled with some exciting possibilities and will greatly aid in a lot of tasks in the future. I'm not against using Chat AI to augment doing schoolwork or assisting in creative projects. They can be used to fix up grammar, better structure your paragraphs, and work out plot points. You do need to be aware that chatbots take the information you give it and keep it to use in the future, so be very careful about using personal material or identifying information. Don't feed it your whole novel or anything.

You also need to be wary in how you use it to assist with doing schoolwork. Like online proctoring programs, education chatbot detectors are designed to assume the worst when it comes to plagiarism, but you can also use them for your benefit. Some detectors, like Grammarly, offer a free tool for students to check their own work to make sure it sounds original, and you can also talk to your instructors on how they would prefer these tools to be used.

And let me stress, a chatbot is merely a tool! If you use them to write computer code, for example, you still need to check that code to be sure it is correct. If a physical therapist asks a chatbot to create a therapy plan based on certain symptoms, they will still need to verify that plan is the best path forward based on their own education and experience. A mechanic or an electrical engineer using AI as a diagnosis tool still needs to verify that diagnosis actually fits the situation. If you use a chatbot to work out a thorny writing problem, you will still need to be sure it works with your own instincts for what's good with your story.

We are absolutely going to see some shitty usage of AI and automation because, you know, capitalist hellscape and all that. For good reason people who work in the field are pushing for regulation and oversight. But a chat AI is like a calculator, not Astro Boy. Remembering it's merely a tool, not a replacement for thought and creativity, and using it as such can be for your benefit.

[oxymoronicdumbass]

if i have to sit through one more conversation in which i have to listen to someone chirp about the benefits of AI, i am going to stab someone

[sciderman]

How do you feel about the increase in really weird NSFW ads on here (advertising panels that look like sexual encounters, and AI art apps that pride themselves on porn) but will take down NSFW posts from their users, even if it isn't technically sexual.

i hate all social media and it's consistent prioritising the advertisers over the users and the internet simply was a better place before capitalism sunk its hooks into it

[torchickentacos]

I'm getting so sick and tired of AI. Someone needs to start mass-feeding it eng dub AG pokemon scripts only or something. Make it watch Spontaneous Combusken 87934573489 times. I think that would sufficiently poison its datasets if we could do it enough times.

[starflungwaddledee]

alright. look, we're going to go into this because i genuinely want to think you didn't mean harm by this, but it's unacceptable to use this sort of manipulative phrasing. especially with strangers, some of whom are kids.

this is not a call-out. i've hidden all names except my own. i just need to address this post i was tagged in, and don't want to add it onto the end of the original post. i could also have done this privately, but i want this to be here for the other folks who were @'d. i won't @ anyone else who was mentioned in this post, but many of us are mutuals, so if you see this post and you're feeling at all stressed out or bad, i recommend just clicking through because i'm going to go into this.

firstly, and i'm going to make this transparent, person who @'d me: i don't think you're being malicious or did anything purposefully bad. i don't think you intended harm or that you are "a bad person". i don't have a single negative thought about you as a person. i don't make this post to be mean. i truly think you made an earnest mistake that could easily have hurt others, and i am stepping in with the hope this can be avoided in the future! per my usual boundaries on reassurance seeking, i will not reassure about this further.

secondly, the post that you tacked this onto IS important and a helpful resource, and it is great to bring attention to it. we should be doing everything we can to not only defend against, but actively fight back against generative AI. many people cannot access the most commonly recommended tools (myself included), so a resource like this is fantastic and i'm glad to learn about it and share it! i don't speak for anyone else, but i've said before that i personally don't mind being tagged in resources that could help me or others and i'm usually happy to share them, especially if i think the latter

but, assuming that you are genuinely well meaning and don't know better, you need to know that this is not the way to go about it. i don't mean mass-tagging, which is fine in times like this imo, i mean your written add-ons that actively guilt trip every single person you tagged.

"if you weren't convinced by the idea of being a good person" and "I do hope anyone I @'d isn't a bad person" in particular.

you may not have realised, but these are profoundly manipulative and cruel things to say. regardless of how you intended them, they are inciting guilt in the reader, and especially in the people who you actively called to come and look at it. here's what it sounds like:

"hey! you! yeah you! come look at this!! come closer! now, do what i ask you to do, or you're a bad person."

there are a million and one reasons someone might not reblog something. being tired, offline, anxious, even needing to run a specifically professional blog with exclusively your art on it for your own financial survival which makes it hard to reblog important posts like this; none of those are bad.

in this case, only one thing makes them a "bad person", and it's "they're pro-generative AI and did not reblog because they want to hide this information to ensure they can continue stealing from creatives".

i'm fairly confident you don't actually think anyone you tagged here has that point of view, or that you really have any doubts about their stances on generative AI. in fact, of the folks i recognise here, they're all independent creatives, sharing artwork with fandom for free on the internet. they are the victims of generative AI, and like most of us, are facing a terrifying future and are already desperate to find a way to defend/fight back.

you do not need to use manipulative language like this to get us to care about this sort of content! this affects us all, content creators and content consumers alike!

in future if you want to direct folks to something like this, which is super helpful and it was good of you to do!, you can just @ them so they see it. you can even say something like "this is important and some reblogs would sure help to boost it!". this is still a call to action, but without the manipulative phrasing, just in case they cannot act for any reason.

in the end, guilt tripping people like this, intentional or otherwise, is dangerous.

at best it will make them feel like shit and they'll feel forced to reblog + share from you out of guilt rather than just believing in the cause. and sometimes it feels like it's most effective, especially when things are urgent; but in my opinion the risk of harm is just too high. because at worst, you could accidentally send someone into a negative thinking spiral. you can never know what people are going through offline, or outside of your spaces, and how something like this will hit them.

[another-one-bites-the-chipaguasu]

Do me a favor and block me if you have anything nice to say about the current widespread usage of AI.

[rollercoasterwords]

like. i still wouldn't want someone to copy and paste my fics into a large language model like chatgpt but it's not so much bc i'm worried abt my work being stolen (seems unlikely that an LLM would spit back out my exact words considering how it works, and even if it did, i doubt any individual would be able to like. publish and profit from those words, based on the nebulous status of copyright law when it comes to LLMs like chatgpt. and having my words fed into the LLM really isn't going to make much of a difference when it comes to corporations profiting off the tool in the first place; plus in instances of corporate exploitation i think there are more effective ways to organize than like...arguing for strengthened ip laws or trying to like make ip laws for fanfiction spaces). it's more because i'm wary of what that says about how a person is like...approaching my fic specifically + fanfiction more broadly. in two main ways:

1. i think it is just. basic respect to check with a writer before u take their work off ao3(or whatever fanfic-specific place it's been shared) and put it somewhere else. and like, this applies to lots of things outside chatgpt--reposting fics to other sites, posting them on goodreads/storygraph, printing + binding fics, etc. if u are treating fanfic writers as people who u are in community with, who are generously sharing a gift with u, then it seems like basic kindness to check in and see if they're alright with u taking that fic outside the space it was posted to do something else with it.

with chatgpt and similar LLMs specifically, a lot of people are wary because there's still so much unsettled in regards to how copyright laws might shake out, and most people (myself included) are unsure of how/whether our writing/data might be stored and used by these corporations that own the LLMs. i don't think ai itself is something that should be mythologized as like ontologically evil technology, but anytime a corporation is introducing us to new tech like this we need to be wary of where it's coming from and how it could be used--people have already pointed out a lot of very serious issues with the way this technology is being developed and how it could/likely will be/already is being used exploitatively--which, again, is more a matter of organizing against corporations than railing against ai tech itself, but is still a valid reason for writers (again, myself included) to be wary of having their work fed to LLMs without permission.

and like. sure, u don't have to care abt writers' feelings + boundaries and can just take their stories and do whatever u want with them. but to me that says u aren't treating fanfic as a community space, but rather a content farm in which fics are products that u are entitled to do whatever u want with. and i just think that's shitty! and if that's how ur treating fanfic then i'd rather not have u reading my fic at all

2. i honestly think it's a strange way to engage w storytelling by treating endings this way. like. story endings are usually v important + intentional, and can completely change the entire tone, themes, messages, etc of a story. i understand going to the writer and asking them abt what they had in mind for the story ending if ur looking for closure, and i understand imagining ur own story ending or even writing ur own ending to an unfinished story. what i don't understand is plugging a story into chatgpt and having it spit an ending out for u.

and like. maybe this is bc we've all been calling these LLMs ai, which evokes an impression of like. a sentient robot creating something. but that's not what these programs do! the first article i linked explains how they actually work really well, but essentially--chatgpt and similar LLMs cannot create new ideas. they can't take a story and synthesize its themes or pick apart its tone to then come up with an original idea for an ending. at the same time, they aren't just plagiarism machines that are ripping text directly from other writers and spitting it back out.

instead (to my understanding), what they're doing is compressing vast amounts of information by running statisical analyses to just save the most common trends, patterns, recurring info, etc, and then plugging that in to fill the gaps. it looks like it's writing something new, but it's essentially just paraphrasing already-existing information pulled from the internet. so i'd imagine that if u fed an ai a fic and said "write an ending," the ai would basically compare the fic to whatever similar stories it has saved and then spit out an ending that is most commonly found on the internet for that type of story. [not an expert here tho--this is just my best guess based on the bit of research i've done].

my point is--you won't be getting a new ending inspired directly by the story u put in. you'll be getting a paraphrased version of the most commonly recurring type of ending for similar stories on the web. and i just....don't see how that would be satisfying in any way. it seems, again, like a way in which someone would be approaching fic like a product, something that needs to be finished + complete bc ur entitled to it, rather than viewing fic as a piece of art with its own unique themes, message, and story that can't just be plugged into a one-size-fits-most ending generator. and like, i'm trying to avoid mysticizing writing as some sort of ethereal art form that would be blasphemously degraded by having someone plug in a shitty ending paraphrased from a conglomeration of various similar stories--i don't think someone creating a shitty ending for a story is like. a horrible evil thing. but i can understand where the satisfaction is coming from if you're writing your own shitty ending, where you get to come up with where u think the story would go + where u get to synthesize the themes u picked up on etc. but ai isn't even doing that--so again, i don't understand where the satisfaction is coming from aside from just going "well every story i read needs to be finished," which. makes me wary bc it just feels like a completely different way to approach stories and storytelling than i would hope to find in fanfic spaces, one that treats fic less as a creative place to explore and more as a transactional space where u are entitled to products.

anyway. feel like my thoughts + feelings abt ai keep changing the more i learn abt it + i'm sure they could change again, but rn my impression of this whole situation is like. i find the fact that some people are plugging fics into LLMs less concerning re: ip + ownership rights, and i don't think it's useful to exaggerate or mythologize abt what ai actually does (i think even calling it ai has kind of misled a lot of people, myself included). what concerns me more is that plugging fics into LLMs to write endings feels symptomatic of a broader culture in which people treat fanfic as an informal profit economy in which fics are product or content that a consumer-audience is entitled to, and i think that sort of approach leads to a whole plethora of other issues + makes fandom a more hostile space.