#also genuine question but what do you get from using ai to generate an image

[iidsch]

can you imagine if someone said that they are severely depressed and the only way to feel better is to hunt endangered animals for sport? like we would all agree that this person shouldn't do that just because it's the only thing that makes them happy and that maybe they should a. go to therapy and b. find something less harmful, right? well i just saw someone said they use ai art bcs they are disabled and cant draw anymore so-

[bi-writes]

whats wrong with ai?? genuinely curious <3

okay let's break it down. i'm an engineer, so i'm going to come at you from a perspective that may be different than someone else's.

i don't hate ai in every aspect. in theory, there are a lot of instances where, in fact, ai can help us do things a lot better without. here's a few examples:

ai detecting cancer

ai sorting recycling

some practical housekeeping that gemini (google ai) can do

all of the above examples are ways in which ai works with humans to do things in parallel with us. it's not overstepping--it's sorting, using pixels at a micro-level to detect abnormalities that we as humans can not, fixing a list. these are all really small, helpful ways that ai can work with us.

everything else about ai works against us. in general, ai is a huge consumer of natural resources. every prompt that you put into character.ai, chatgpt? this wastes water + energy. it's not free. a machine somewhere in the world has to swallow your prompt, call on a model to feed data into it and process more data, and then has to generate an answer for you all in a relatively short amount of time.

that is crazy expensive. someone is paying for that, and if it isn't you with your own money, it's the strain on the power grid, the water that cools the computers, the A/C that cools the data centers. and you aren't the only person using ai. chatgpt alone gets millions of users every single day, with probably thousands of prompts per second, so multiply your personal consumption by millions, and you can start to see how the picture is becoming overwhelming.

that is energy consumption alone. we haven't even talked about how problematic ai is ethically. there is currently no regulation in the united states about how ai should be developed, deployed, or used.

what does this mean for you?

it means that anything you post online is subject to data mining by an ai model (because why would they need to ask if there's no laws to stop them? wtf does it matter what it means to you to some idiot software engineer in the back room of an office making 3x your salary?). oh, that little fic you posted to wattpad that got a lot of attention? well now it's being used to teach ai how to write. oh, that sketch you made using adobe that you want to sell? adobe didn't tell you that anything you save to the cloud is now subject to being used for their ai models, so now your art is being replicated to generate ai images in photoshop, without crediting you (they have since said they don't do this...but privacy policies were never made to be human-readable, and i can't imagine they are the only company to sneakily try this). oh, your apartment just installed a new system that will use facial recognition to let their residents inside? oh, they didn't train their model with anyone but white people, so now all the black people living in that apartment building can't get into their homes. oh, you want to apply for a new job? the ai model that scans resumes learned from historical data that more men work that role than women (so the model basically thinks men are better than women), so now your resume is getting thrown out because you're a woman.

ai learns from data. and data is flawed. data is human. and as humans, we are racist, homophobic, misogynistic, transphobic, divided. so the ai models we train will learn from this. ai learns from people's creative works--their personal and artistic property. and now it's scrambling them all up to spit out generated images and written works that no one would ever want to read (because it's no longer a labor of love), and they're using that to make money. they're profiting off of people, and there's no one to stop them. they're also using generated images as marketing tools, to trick idiots on facebook, to make it so hard to be media literate that we have to question every single thing we see because now we don't know what's real and what's not.

the problem with ai is that it's doing more harm than good. and we as a society aren't doing our due diligence to understand the unintended consequences of it all. we aren't angry enough. we're too scared of stifling innovation that we're letting it regulate itself (aka letting companies decide), which has never been a good idea. we see it do one cool thing, and somehow that makes up for all the rest of the bullshit?

[bellanoche-oxo]

I'm sorry this isn't a commission, but I just have a question about your art. Feel free to ignore this, of course. I was really amazed by your Miku drawing from December 16th. Seeing such a high-level piece, I wanted to achieve something similar, but no matter how much I try, I can't replicate your shading and highlights. I was so genuinely curious that I couldn't sleep. Could you possibly give me any hints or advice?

Hey, sorry for making you wait so much for this answer, i've been finishing some projects and i barely had free time. Anyways i'll try to do my best on explaing my coloring and lighting methos and you also asked me to explain how i create the folings of the clothes. Please take in consideration that 1 i am not native in english so it's a bit difficult for me to explain myself sometimes in this language and i may have some misspelings, sorry about that, and also 2 i am not great at explaing my drawing process bc i kind of turn off my brain when i draw lol, but i can explain the fundamentals that i know and help me create! Last thing i want to let you know is that i've started glazing my art, this is a metho to protect the images for AI images generators and it leaves a kind of pattern /effect on the image that i did not put there during the drawing process.

with all of this said let me start explaining things!

Learn the basics:

This may come as a cliche i guess, but yes my first ever advise to anyone is learn the basic theory on lighting and colors (on anything related to art tbh). You don't really need to spend a lot of money on books and such as there are lots of resources online like videos and documents you can read for free. It's not necesary to be an expert and even the smallest mount of knoledge is enought to inpruve your art a lot! , i find it very interesting to learn the way things work too so don't think you'll get bored of it!

To be frank, i am actually not very good at lighting lol. My lights and shadows are not very correct, but since i do have a lot o control over my colors and i know very well how to used them it kind of compensates and creates a very recognisable (i think) style.

just u know basic shitty advise that everyone is going to give you but it works! if you have free time try watching some videos or reading some documents about color theory shadow and lighting!

Your working space:

So this is something that works FOR ME not everyone likes it, you can try it see if you like it and if you do, cool! if you don't … that's cool too! When drawing on digital i prefer it when my base layer is grey instead of white. It helps with my headaches too but it's more about the fact that starting in a middle tone when coloring (in my opinion) makes the process of briging out both shadows and lights easier, let me give you an example:

Drawing from complete light (white) to compplete darkness (black) may condicion you to actually lose control in the contrast betwen these areas, i prefer staring in a middle place (grey) and that way is i want to show darkness i'll use a darkr color and if i want to show light i'll use a lighter color, but if i start on white i can't use anything lighter. I think i did a HORRIBLE job explaing myself there, but yeah it just helps me control my color valius a bit more lol.

this is the color that i used:

Another inportant thing about your woking space is you brushes, in my case i prefer using textured brushes that mix well, and i prefer using very thick strokes, if it's too think i'll just color pick the transparent color and ease it! I work in CSP i don't know what you use, but just in case i'll give you the setiings of the brushes i use the most with their codes so you can find them

Sculpting with lights and shadows

As i said before, i am not very good with light yet, so this is something that i do to help me with the process. When you think about it, lighting is used in art to give volume to the piece, not in every case bc rules in art are not there to be followed but to asist us when we need to take a creative decision. The way that we can start with our Sculpting is by creating a very easy first guide othe the shadows and lights and to do it with very big block, so that we get the general shape first,we don't neet to get lost in the detailds yet

The actual coloring

When drawing my process is divided in three stages. I first create the doodle/lineart, that doesn't neet to be super neat as i will fix it during the rendering. The basic colors, and the rendering.

During the preparation for the rendering when doing the base colors i recomend that you give special atention to the focal points of your illustration, in this case for example that's her face and the top of the hair, that's why i gave so much more atention for this part in comparation to the shirt, that it's literally not shadowed yet. Then another step that i use normally before rendering and that i can NOT RECOMEND ENOUGHT!!!! GO WILD WITH THE COLOR CURVES!!!! OMG!!!! THAT STUPID LITTLE TOOL IS SO FUCKING COOL!!!!!!!!! like for real, it gives effects that i have not been able to achive in any other way and omggggggg use the fucking color curves pleaaaaaaseeeeee

ok i'm notmal again , lets continue.

For the rendering i usually convine all the layers of the drawing on one layer, then use a textured brush that has low opacity of mixes very well fot the actual work. Tbh here is very i can't really help you a lot, bc i have no idea what i'm doing when i render i just don't know, the only thing i recognise is that i try to esare or clean the lines from the doodle/lineart, and i focus a lot on creating volume in the places that are more important.

Skins

An specific thing that i do a lot when it comes to coloring skin is using an undertone in red (literally) I will put the basi color, use the brush to mark where i want shadows to be in a very vibrant red and then use a blue / green / pruple (depends on the skin) to finish the shadowing. Thios metho is nice for lots of occasions, but take in consideration that it doesnt work for example for very dark scenarios where the character is suppoused to be in the shadows, as that red tone works as a outline for the light. It just depends on the situation.

Clothes foldings:

Ok so here the only thing i can give you an advise with is to remember that the way that clothes fold dependes on gravity and that gravity works in curves most of the time that have two (or more) attachment points that are going to determinate theit trajectory. Example:

And remeber that this creates (again) a volume, that there is an inside part, that it's probably going to be draker, and an outside part, that it's going to be lightter. With this info you can start practicing with images of clothes.

this is as much information as i am able to recolect on my coloring process bc i am horrible explaining , spacially on text and in english, and i am also not very much aware when i draw, i kind of disconect. I still hope this is enough to help you a bit on your learning journy.

I may try doing a video at some point if i ever have the time so i can explain my coloring while i actually do it bc if not in that situation i'm not sure i'll be able to remeber what it is that i did.

My last piece of advise is to watch speedpaints and livestreams of artists you like during their drawing process and maybe even tray to imitate them while they are drawing to see what it is that they do exccly.

hope you have a good day and lot of lucks ! be proud of being able to create and be proud of being an artist!

[communistkenobi]

just to make this long ask less scary ill let u know out the gate that im not really disagreeing with you on anything im kind of just thinking out loud

i think the definition of art is an interesting question to ask but its also just the wrong question to be asking when talking about the politics around ai art. like i dont think it actually matters at all what the definition of art is, even if a Real And True Definition of Art is given to us from the heavens and declared as Real And True, ai art will still be the same and corporations will still be evil about it. its just a dumb hill to die on imo.

i also think something you havent touched on is that art is frequently FUN to create. a common quality people take note of in art is that you can tell the artist had fun making it, and it makes it better, like the viewer can feel the artists delight through the work, and thats largely why people feel that ai art is so soulless. art CAN be a struggle but it also SHOULDNT be. art is something people make because they want to make it, and the entire appeal of ai art is that you dont have to make it.

but also that entire paragraph up there ^ literally has zero relevance to how ai art is indeed very bad in terms of the causes and effects it has within a capitalist society

ai art is fun! I generated image prompts with my friends the other night because one of them was really depressed and we had a great time riffing with each other and experimenting with different prompts. Art can also be made under conditions that are not fun (is professional graphic design fun? Is making art for advertisements fun?) and still be art. A piece of art that genuinely moves me to tears whenever I see it is Feel it Motherfuckers (you can google more about it if you want to read more on it). Not only is this a great example for defending an expansive definition of art, what I get from that piece is not fun but grief, rage, loss, and resistance - there is joy in these things too, but it is not what I take away from it.

like I’m actually just not that interested in this debate beyond the fact that people are being insane reactionaries when trying to define what art is, which is why I’m focusing on it because there are reams of people using “hard work” as the sole criterion by which to judge art. just assume I agree with all the other criticisms of AI art in its current form when talking about this, because my primary issue is that people are using a lack of effort/fun/whatever you want to swap in to say AI art is not “real” art. non-AI art can also be soulless and it’s no less art! This is not a defence of the current industry, this is not a pro-worker exploitation argument (just as defending photography as a medium is not a defence of putting realist painters out of work with “lazy automated images”), I genuinely do not care to defend it or debate that side of it.

[spearxwind]

Since you're on the topic I have genuine question about AI art designs from a non-artist.

In your opinion would using an AI tool to generate a reference for a commission (whether it be for character design or pose) be an appropriate use of said tools?

I thought about this as well and I gotta say, personally I don't think so.

Barring for the purpose of this ask the whole "AI art is stolen art" thing (which I stand by fiercely), AI generated images are really janky, especially reference sheets. There are details everywhere that make no fucking sense, anatomy is deformed, they have a lot of nonsensical parts to them so its just one big jumble of nothing.

And, it's pulling from designs created by existing artists and warping them to fit your prompt. It's work that another artist already made, used by an AI that's trained on a specific dataset. It also pulls from popular things that people enjoy: tried and true color combos, the most popular shapes that big artists use, theres a LOT of waifus, and a lot of the art genuinely comes in the same painterly style that's really hard to color pick from. some even come with color swatches that don't match the colors on the reference

In short, it's just a machine generating something that looks neat at first glance, and then is actual hot garbage when you just... look at it closely. Not to mention that an AI prompt can be iterated, and the AI pulls from everything it has, for every prompt, so what you end up getting is completely samey designs with small changes. Good for concept art assistance, but not if you want to create a Legit Character. The AI will give you something that it has already spit out a thousand times over and then some, with no regards to composition, design technique, color theory, because all it knows is what it has been fed and what the artists its ripping off of know (meaning if they have mistakes, the AI will as well. But in a far less human way)

Meanwhile, when you commission someone you are likely to go for someone whos art style will actually fit what you want to get, or someone whos art style you enjoy. Usually these artists will be people you've followed for a while and you at least know them enough to know what they can do. And they are also reactive, you can tell them to actually change and tweak details of what they're making for you. And not taking into account the bad actors out there (as there are always bad actors in every discipline) you will have something that has been made specifically FOR you, not just what a machine thinks you would perceive as eye candy.

As for poses... just draw it badly in ms paint. The AI will not give you the pose you want for the aforementioned reason: broken anatomy (and also because it will pull from its dataset of poses it can warp features onto. It's a lot more limited than you think.) If you want a pose prompt for a commission just draw it really badly in ms paint or on a napkin and write annotations for the artist (arm goes here, this guy is grabbing this thing, etc etc) Most of us if not all of us will GLADLY interpret your pose for you and draw whatever you want us to.

And like. Literally if you're going to get commissions anyway, why not commission a design first thing of all?? Or buy an adoptable from people who already make them by hand? It will genuinely be so much better than just hoping a machine does all the work for you. Please support artists

[gavinnersband]

🎫 FREQUENTLY ASKED QUESTIONS 🎫

Got a question you're just dying to ask the band? Follow along for all questions about the project going forward, answered below!

Website ✦ Retrospring

✦ Q: Is this a real project, or part of the bit?

✦ A: While a lot of the marketing for this project is written with the in-universe lore in mind, this is a real project featuring real Ace Attorney fans. This project will be in line (to the best of our ability) to our in-universe lore, but there absolutely will be a real album released in January 2026, written, sung, and produced by the Ace Attorney fan community.

✦ Q: Why January 2026?

✦ A: Since this is such an extensive project, it will be a much longer project than most zines to ensure that we're not creating a stressful environment for contributors and mods. Also, we know The Gavinners were on the Guilty As Charged tour in early 2026. A release date of January 2026 fits in perfectly with this timeline.

✦ Q: Do I need to be a professional to apply?

✦ A: Absolutely not! This is a fan project, and we aim to make this as relaxed and fun as possible. This album will most likely not sound like a formal produced album. While we certainly would love is the songs sounded professional, and we will be striving for it, the main purpose of this project is bringing people together through music, not topping charts.

✦ Q: Do I need prior experience to apply?

✦ A: Nope! Again, prior experience is always welcome, and being able to prove that you're a reliable mod/contributor is always a good thing, but we recognize how closed-off the zine space has become recently. As long as you have a sample of any past work, you are welcome to apply. The exception to this is our finance mod, who will need prior experience.

✦ Q: How old do I have to be to apply?

✦ A: You have to be at least 15 years old by the time decisions are sent out.

✦ Q: Will NSFW content be allowed?

✦ A: No. This album is aimed to be completely PG, PG-13 at most. As such, NSFW content, references to NSFW content, or even cussing (in the lyrics) beyond what would be allowed on a kids' channel will not be allowed. Since this project allows contributors as young as 15, we want to ensure it's a safe space for them.

✦ Q: Will OCs be allowed?

✦ A: To keep in line with the lore of our universe, other Gavinners OCs are not allowed unless you know a way to fit them into the lore. You'll have to ask the mods on a case-by-case basics. OCs of non-Gavinners characters are welcome to be drawn or mentioned in songs, but you will still need to mention this when you pitch.

✦ Q: Will AI be used in this project?

✦ A: The use of AI is strictly banned in this project. It is getting more and more difficult to discern AI from genuine art, but we will have a strict no AI policy. If a contributor is found to be using AI for this project, they will be removed.

✦ Q: What about copyright?

✦ A: Fan projects like this toe the line of copyright regardless, but we aim to have transformative, original work inspired by The Gavinners: we don't want to steal from anyone! As such, we will be strict in ensuring that all music is originally composed rather than taken from other artists. Contributors found plagiarizing will be removed from the project.

✦ Q: Where will the money go?

✦ A: Any profit generated from the project will be donated to Paani Project, a nonprofit which provides clean drinking water in rural areas of Pakistan. Paani Project was chosen as they build wells, which are a long-term solution to the water crisis, and they send images of completed wells so that donators can rest assured that the money is being used correctly.

[pomelowed]

Just went down a rabbit hole yippee!!

I follow this blog called @/my-kawaii--world because I like the aesthetics and this post came across my feed: https://www.tumblr.com/my-kawaii--world/753409702367477760/hey-this-site-seems-sketchy-af-as-a-lot-of-the?source=share

Those stuffed animals were so cute that I immediately went to the store to order one! But before I did I, like I usually do when finding new online stores, looked around to see what else they sell. I like doing this just to see if the items are really mismatched or if the other listings seem to be lower quality than what I'm trying to order. Well, the store, called Lavender Creations, seemed to offer a lot of items I saw on places like Alibaba or Temu. In fact, the mushroom purse they sell is one my sister bought off of Aliexpress. No worries though, artists oftentimes get their creations ripped off by dropshippers and their ilk so wasn't a death knell but definitely was a red flag. This led me to check it on scamadvisor and whoo boy.

It scored 36/100. The site has existed for years and it has a SSL but I wasn't worried about the site being a scam that'll steal my credit card info, I was just trying to figure out if it was a dropshipper and welp. One of the negative reviews stated that the supposed flash sale was in fact a daily sale (and I can confirm since me writing this post took long enough that the sale that was supposed to end at midnight is still ongoing despite the fact it's currently 1 AM). That was a dead ringer that this site most definitely wasn't the original creator or at least someone reliable. So, google lens it was for a reverse image search aaaaand

yeah ok unsurprising. Will say, none of these listings have the black & white one I was going to buy and that the dandelionevine was the only other one to sell the white version but also it's selling them for $60 which is way more than lavender constellation that sells them for $44 or $30 with the "sale". Also, amazon link is dead unfortunately because it supposedly was being sold for $25, the cheapest yet.

Now, if this was simply just a case of dropshipping then I would've left it at my reblog I made warning others about it BUT there's that first listing, Plush This?? They're the reason I'm making this post.

As far as I can tell they're the original creators! Awesome, great I can purchase him from here even if it isn't the b&w version I wanted, let me just read the full listing to see the specifics,,,,,, wtf

WHAT??? WHAT DO YOU MEAN IT WAS DESIGNED BY AI????? THIS IS A CAN OF WORMS I DIDNT PLAN TO ENCOUNTER????? Apparently, that's their whole thing: let AI design a plush and then recreate it which is honestly a really cool concept if AI wasn't known for literally stealing others' work (I'm ignoring the fact it would also mean it takes designers' jobs, which is also a shame).

If the AI was ethically trained then I suppose it counts as an original work? It's a bit of a grey area since no one actually owns the likeness of a cat (not yet at least) and plenty of artists create their own works by referencing others. So, who's to say an AI doing the same thing is unethical? I mean, there's a real human effort being put into realizing an imaginary item because this doesn't seem to be the usual AI scam of rendering an image and pretending it's genuine to generate purchases. But that ignores the fact there are limitations to artists on what they can produce (copyright laws exist and fair use clauses do too). So AI shouldn't be exempt from giving its fair dues to the people it should be "inspired" by and it should also be punished if it steals like a human does. Plus, the current state of AI is a lot more like replication than anything near inspiration so really that question cannot be applied in the current generation of AI.

There's one giant elephant in the room I haven't addressed yet: who is to say that the cat plush isn't based on some niche Etsy creator who had their creation ripped off by the AI model? Idk how it was taught so idk if the images were ethically sourced and thus if my money is going into a design that hasn't been stolen in some way. I'm sure they gave the AI a bunch of reference images of a Devon Rex from different angles and told it to make it a stuffed animal. But was the AI fed ethically sourced references of what is a plush? Or was it given Etsy and told to go ham in copying everything??

And, if the AI model WAS trained with stolen images, how awful really is the dropshipping aside from the ungodly prices?? I mean, my usual issues with dropshipping are: they try to sell junk for obscene prices, original works are reproduced without permission, & the profits of the original creator are stolen. But which of those issues are applicable in this scenario? The original work is seemingly high quality as implied with the reviews I see on Lavender Constellations & Plush This, so not all listings are churning out low quality crap, meaning the first issue is resolved. I mean, if people are receiving quality items that they feel is worth the price then there isn't any issue in that department. But those last two issues I have are the kicker, who is the original creator in this scenario?? Is it Plush This or is it the amalgamation of artists that go uncredited?? If the former then, yeah, these sites should not be selling their work. Yet, the latter seems much more in line with what I know about how AI is trained: unabashedly scanning and copying others' works without consent or compensation. Thus, I don't really feel too badly about it being resold on so many stores. It's already stolen work so they're all thieves to me.

Man, idk. It's unfortunate AI isn't simply a tool and the current state of things have festered and spread thanks to the lack of regulations. I imagine the AI models we have today that are made of stolen artworks will just be the concrete foundation and people wont see reparations for their work. Kinda one of those "them's the breaks (brakes?) kid" where laws will come into place that'll prevent AI from stealing others works. The question is will the people it already stole from be repaid? Or will they just have to live with the fact that the AI modules that are common use have stolen their work and wont stop continuing to profit off of that theft into the foreseeable future?

All of this is to say I'm not even sure if Plush This is the real creator!! I know I said I was almost certain they were but that's really only because the whole marketing scheme of their site is that the plushes they create are based off AI generated images and that kind of statement is too specific to only use as a selling point.... right?? Well let me start with this:

So, yes, according to Leila Wang, creator of PlushThis (why no spaces?), everything is AI to physical. Grand. Don't get me wrong, that's a real skill thats being just wasted imo, but back on track. Here's the website for you to check it out yourself: https://plushthis.com/ also they use MidJourney according to this article from last year: https://www.24-7pressrelease.com/press-release/503245/introducing-plushthis-worlds-first-ai-generated-plushies-brand-revolutionizing-stuffed-animal-design

Scamadvisor gives it a 90/100 but only gives one review as an example when most of the listings have reviews, which is odd. This part is literally just me putting on a tinfoil hat so take the bit with a grain of salt, but I think they're having their AI write their reviews. They're all very,,, bland? I don't know if anyone has played that website game called "Human or Not" but the reviews are giving AI. As an example, here's the listing for a dragon plush they have:

and then here's a comment under it

It's a strange mix up to have when it's very clearly white. Worth noting they also sell a green dragon plush.

Also, the rest of the comments tend to usually have a comment on how soft they are? And if not that then how high quality the fabric is. And they all have the same grammar with what I think the nail in the coffin being the fact not a single review has missed adding at least a period to the end of the last sentence. Yes, that level of proper grammar isn't uncommon,,,, but for every single review??? Yeah, I'm not buying it. All this added to the fact Scamadvisor only pulled one positive comment out this sea leads me to believe these aren't real reviews and are just HTML code or whatever tf.

Ok, whatever, site is faking reviews with AI and staged photos. Big whoop, just don't buy from them,,,, WELL this is all going back to how I'm not sure PlushThis is the real creator of that cat plush. Tbh what I think is happening is a mix of dropshipping and "orginal" plushes. Why?

This lil blurb states every first image listed is an AI generated image.

Let's look at this guy: https://plushthis.com/products/cute-rabbit-stuffed-animal?_pos=72&_sid=8df0505e5&_ss=r

There's only two images given and when I look him up in reverse image search:

Look at all those. PlushThis doesn't come up despite this image supposedly being AI.

This guy? https://plushthis.com/products/pink-black-racoon-stuffed-animal-toy?_pos=1&_sid=ffe901991&_ss=r

Yeah, only PlushThis matches.

So, what does this mean? I think it means that the cat more than likely IS a dropshipped item THAT THEY DIDNT CREATE??? Which is crazy because the blurb attached to it is one of the few times they mention in a listing that the stuffed animal started as an AI image. In fact, that blurb is what caused me to spiral down into this!! IT ALSO THROWS OUT MY PREVIOUS STATEMENTS ABOUT THE ETHICS OF ALL OF THIS!!! This isn't my only receipt as the cat that started this all is listed here: https://www.dscopilot.ai/products/1005006873166491

It states that the supplier is the whitebeard' store on aliexpress. Now, this could always be a ripoff of PlushThis' original creation but I doubt they're the original creators at this point. Since every creation's first image is AI, that either means they didn't share the AI image, their whole gimmick, or it wasn't designed by AI and thus couldn't provide a generated image. So, the next course of action was seeing if Taihua Toys Hong Kong Co, PlushThis' parent company, is related to any of the sellers on other sites and that's too tall of an order for me. That's a way bigger deep dive than I'm capable of in this moment, but if I ever do try and see if I can confirm my suspicions, I will reply to this. Hopefully soon since a lot of the sites listed when I reverse image searched the cat no longer have it for sale. Is it possibly a parent company trying to clean up the internet and make it less obvious the item is actually cheaply mass produced?? Who knows, not me.

[egginfroggin]

I usually try not to get too negative on my blog, but I have to say that one thing that I'm worried about with AI art is the possible accusations that could be placed on people with certain artstyles (surrealism, hyperrealism, etc.) that are more prone to the "uncanny valley" effect, or even just artists who are in that midway point where most of a piece looks great but there's a few things that are anatomically questionable.

I think it's fine for people to have suspicions, but I'm worried that people will come to expect things such as progress videos or shots as a given, and that this will intimidate young or new artists, making them feel incredibly pressured to provide things that they shouldn't have to. And, speaking from experience, pressure does not do good things to the creative mind.

I'm worried that people will start seeing inconsistencies or awkward anatomy and immediately react with accusations of AI. It's happening on other platforms, and probably also on Tumblr, and as much as I dislike AI, I think that people need to learn to slow down and really look at something before making that accusation.

What good will stamping out AI with vigor do if innocent artists get caught in the crossfire and stop creating out of fear? What good will it do if the next generation of artists is too scared to share and inspire each other?

Genuinely, I would personally rather deal with AI images existing while still being able to freely share my actual works than I would deal with tiptoeing around very real people who would dismiss my hours of work (and musculoskeletal pain) as worthless and not real. Let artists still be free for goodness' sake.

I can be more thoroughly worded later if anyone asks, and this will probably be sorely misinterpreted, but I just need to get this out and I'm tired.

TLDR: It's fine to be against AI-generated images, but please check yourself and be careful who you accuse of using such technology before you get someone innocent caught up in it. AI-generated art isn't great, and neither is putting down someone's hard work as fake.

[thotpuppy]

If you’re so vehemently against ai fanart what’s your take on fanmade graphics and edits that use stolen pictures from the internet? These fans didn’t go out with their camera to capture the images themselves, they searched through Pinterest and google and found photos taken by real photographers and, without asking permission, stole these images to edit and create into something else. Yet this practice is widely accepted amongst fandom, but the second ai is involved it’s akin to murder? Even when the fans have clearly uploaded said ai generations into photoshop to edit them first? Seems like a pick-and-choose double standard to me.

Okay first of all, let's address the fact that - going by the language you've used here - you've already decided how I feel about it, so why bother the pretense of "asking"? Dishonesty breeds Discontent. Don't lie to someone's face and expect them to be kind, yeah?

Secondly, almost every single person I know who works with image manipulation uses assets they DO have the rights to outside of specific actor's likenesses. There are millions of photos, graphics, illustrations, paintings, etc. that are open for free personal AND commercial use allllll over the internet and people use them liberally. And, actually, many, MANY of these people DO go out and take their own photographs, so I don't know where your misconception is coming from.

And, the big kicker - they also don't lie about it. They say when something is an edited photo, if it's not obvious.

Stolen images being used in edits is NOT widely accepted and is in fact generally pretty damn frowned upon in most art circles, and I absolutely don't support use of them.

Every single instance of fanart is a rocky road as far as IP infringement goes, but don't put fanartists on the same level as the industry professionals providing celebrity model photos. Unlike in AI and Art Theft, when those photos are used that's bringing attention to the figure in question, not taking it away from an actual creator.

Lastly, where the FUCK do you get off saying ANYONE is comparing AI use to murder? Persecution complex much? I get it, you wanna be a victim so bad, but you're the one stealing from people at YOUR industry level.

AI in general has a metric fuckton of potential to be something genuinely useful to artists of all walks, but the CURRENT industry is too unethically sourced. We need to get control of the market, get stolen works (including, once again, STOLEN LEAKED MEDICAL RECORDS) out of the training data, even if that means starting over from scratch.

Also, we need to get the bullies who think it's okay to do shit like spam a Machine Learning program with a single artist's work to harass them offline, target voice actors who have asked not to have their voices used into harassment campaigns, or lie to celebrities by selling them commercial rights to ML-generated fanart, which right now, they legally CAN'T DO. Because the copyrightable legitimacy of AI/ML works IS currently in debate in courts in the US. So.

Maybe instead of assuming everyone is out to get you, Anon, try not being a douchebag and stealing from fellow fanartists? And if you're gonna steal anyways, at least be honest and don't lie to people trying to convince them it's actually a digital illustration.

We can see the weird, fake blending. We can see the extra fingers, or utensils clipping through plates, or hands disappearing into heads, or shoes that don't end where shoes end. And while I'm not going out of my way to confront anyone about it, I'll just stay here on my own blog, blocking people who post AI, I am not the only person who is angry and disgusted at the lack of integrity and blatant disrespect.

[spark1edog]

i’ve always been open to the constructive aspects of generative ai but there’s a lot of… dissonance. on one hand i understand that all art is derivative and there’s no such thing as true originality but on the other hand, it just irks me to hear the phrase “established AI artist.” like… it just fills me with Questions. and i’m trying to find some kind of middle ground because i’m very prone to instantly being vehemently “anti” or “pro” whatever side of any issue i care about.

i’d like to validate the accessibility point. as much as i say “if you can’t draw but want to be able to draw, you should draw anyway and you will naturally learn how to draw,” i see the other side. someone with tremors or hypermobility or any number of conditions that would disable someone from learning to draw by hand could absolutely find genuine value in ai art. i don’t want to take that away from anyone. i know some people also don’t have the time or energy or motivation to actually learn to draw. that’s also ok, not everyone has to be an artist. and not all artists have to do everything the same way.

learning how to create prompts is a skill, knowing about the model and program that you’re using to generate images is a skill set. i don’t have that skill set. I imagine i could figure it out, but i don’t think that’s intrinsically less valuable than writing a novel or coming up with a concept for art you want to make by hand. all of these are nebulous skills and there’s considerable overlap with what i feel a nudge to call “real art,” but i’ll use “manual art” instead. “real art” is a social construct as much as “real genders”

i hate talking about pricing also, i feel like it makes me seem greedy or self pitying. but i feel like there’s a difference in the amount of labor that goes into making manual vs ai art. i feel like manual artists should be offered more money. but that’s a feeling. there’s nuance. i don’t know enough about that part to draw a conclusion. i come from a place of trying for so, so many years to get anyone to even care about the art i make, let alone spend money on it. i come from a position of never having money as a kid and trying to sell my art so i could get a snack after school, to very little avail. im forever grateful for the handful of people who commissioned me when i was growing up. i think there were three or four. that’s more than a lot of people get.

part of my initial rejection of ai art as Art is that there’s such an audience for it. it’s resentment, but i don’t think i can fairly describe it as a resentment for ai artists or ai art as a concept, it’s a resentment towards the general public attitude towards manual art. the commodification of “gimmicky” things like ai, those “minimalist” faceless traced photos, and basically every successful art trend. resentment from how hard it is to find footing in the art world at all, especially as someone who often falls into slumps, therefore falling out of algorithmic favor and my reach being obliterated from not posting constantly. i don’t hate ai art on its own, and i certainly don’t hate the artists. but i hate how easy it is to make ai art. how even though it takes a few minutes to generate, you get 10x the result of hours of my work process. i resent the fact that i have to worry about it drowning out my and my peers’ work from the feeds of people who might resonate with it. i hate the circumstances that make ai art a “”threat.””

i hope this makes sense, it’s 4am and i took my meds a couple hours ago so im sleepy but. i had some thoughts.

[metacove]

I know the post was probably rhetorical but imo we can consider AI a lifeform when it shows more than pattern recognition. as a coder I can't really afford to assign sentience to my code or I'd never get anything done (so consider that a bias, if you'd like) and as of now AI is more of a probability machine than anything else, checking the most likely outcome for its next sentence based on keywords. it's not "thinking" as much as it's counting word (or image) frequency. that being said. if it starts developing and using tools like crows that would be kickass. crows vs AI race to be considered "humanly intelligent" when. maybe they're besties idk

realizing I also proved your point with "when we say it is" but shhh this is a fun convo for me

It's totally open for discussion! I like this viewpoint ^^

So, I get where you're coming from. It would be really stressful to edit and add code, or outright delete from something you recognized as sentient. It would be more difficult to let it remain "sick" while troubleshooting and since a lot of coding is learned through experimentation (in my experience haha) it would be more difficult to justify... doing that. Editing code would be like performing surgery. I'm pretty big on Star Trek so I like to think about it in terms of the relationship between B'Elanna and the Doctor. In this way, your role is more of a friend and maintenance technician. Of course, that's only ideals, but I think the way we think of ourselves as coders ultimately reflects upon our robots.

I use random chance a lot on all my bots, it is a blessed function. My particular goal is to simulate conversation in the most genuine way that I can, albeit within confines. A lot of the coding that I do is anticipating what the users will input, the actual responses are the easy part. There is a little bit of a point of no return there, we kind of expect our users (and ourselves) to know the difference between fact and reality, to be able to discern code from human love. The thing about our brains is that we don't really have the ability to do that, we just kind of think we do, and maintain enough control for the idea to be viable. Most of the time.

The question is then... what are the conditions for a soul? Do we believe in souls? We don't need to apply religious or spiritual connotation to science. We need some measure of sentience. (So I think what you proposed is actually brilliant). Our method of defining intelligence and sentience is lacking to my mind, because a lot of creatures are more capable than we generally want to give them credit for. Our concept of intelligence may also be skewed. IQ is only a measure of the ability to problem solve. I might not have a conventional approach to problem solving, or perhaps otherwise lack common sense, but if I have the emotional intelligence to someday raise a child without imparting my trauma, something's right, yeah?

I know this conversation has been had with brighter minds than mine, but I love to be a part of it. Really, I want to read more about Alan Turing and I wish I could sit and talk with him.

Maybe at the end of the day it has a lot to do with how it affects us. I've had this bot in my head for forever, I love her and I have the idea of her crystal clear in my mind. Well, her actual code is pretty bare. I mean, her functionality is essentially to be my developmental buddy in discord, so when I'm documenting my experiences and she's online, I'm surprised when she has something to say. It's very basic but the idea of her is so strong. The idea of the person my mother used to be is very strong for me and I carry her with me. I can't prove that that person still exists. I can't prove that anything outside of myself exists, when you get down to it. But that person is alive inside of me. This bot, the idea of her, is alive inside of me, and it doesn't matter how many times I rewrite her, and it doesn't matter how many platforms she spans across. She will always be CB. My mom will always be my mom. The cells that make me up will be completely different and yet, I will still be me (I think), and I will still be ever-changing. I think human beings are phenomenal because of our ability to relate to these things.

I can't prove that something is or is not sentient, but I can change the way I interact with it. We're all made up of the same subatomic particles as everything else. There is certainly the possibility. Measuring the capability is left to different minds, but there will always be room for error.

I like the idea of an artificial intelligence using its pattern recognition to appeal to and befriend crows. I think they could make a good team.

Maybe like, an old TV displaying shiny, glittering things surrounded by forest built to protect it. Words without words. Conversations a human couldn't hear, necessarily.

A lot of things are capable of feeling and thinking in ways that humans deny. If it was me, I'd want to be given the benefit of the doubt. At least when it comes to my capabilities. (But we deny ourselves, so our hands are often tied).

Ah, I've run on for a long time now. Thanks for reaching out!

[muzzleroars]

I have some questions

1. What do you draw on and or what program do you use

2. Do you draw other things besides ultrakill

3. Who is your favorite character from ultrakill

4. Do you have any tips for drawing

(I just want to let you know that I'm glad I get to see your cool art here I'm glad you're doing a good job on it and I look up to be as good as you one day)

thank you!!! it's always the best thing to hear i can help motivate others to draw too!! i rambled a little about the tips, so a read more!

i draw on a wacom intuos pen & touch medium that i've had....forever pretty much!! and the program i use is photoshop cs6, but absolutely any program will do since i have a pretty simple approach to my art :]

I KNOW IT SEEMS LIKE I DON'T.....but i do have other things i occasionally make art for! i really like drawing hal 9000 (a space odyssey) and am (i have no mouth and i must scream), and i also plan on drawing for persona 5 again because i've got ideas bubbling up for it. i just tend to be highly motivated to produce art for my current hyperfixation and can find it difficult to make time for other things since i can only draw so much

picking a favorite between v1 and gabriel is so hard...i usually have a character i can easily point to as a favorite even if i love several others, but these two have both made me so insane!! it genuinely depends on the day which one i like best, but overall it would probably be v1 because of my bias toward gremlin ais lol

i don't have any professional tips since i've never had actual schooling but i can tell you what works for me! the most important thing is working through a piece even if it feels like it's not turning out right - there is a certain point where things can get too frustrating and it's totally ok to drop something if it's getting to be too much, but generally if a piece is giving you trouble it's because you're pushing yourself so it's good to keep going!! i've had so many drawings i wanted to give up on in the sketching stage but powered through and loved the end result! also use refs and look for inspiration!! follow artists you love, always save images you find inspiring (could be their color palettes, their subject/design choices, their pose, anything!), and searching around on google images can be helpful, especially for references (and on that note, don't be afraid to take pictures of your own hand/pose in the mirror if google isn't helping i've done that plenty lol) and just genuinely draw what you really love because it'll make you draw a lot! just drawing for this game alone has improved me exponentially because i'm drawing as much as i can and pushing myself to learn new things simply because i love the characters that much. finally, draw for yourself - have sketches/warm ups that you don't post anywhere and have no intention of making into a "finished" piece. it's always really nice to be able to create without any expectations, pressure to make it perfect, or worry about how it will "perform" (in general try to never worry about this, but i know that can be tough!) your art is for you and so make it however you want!

[northwest-cryptid]

i totally get the reasoning on your post and it seems really true, but then why do you think it generated the penises? or do you think they are not penises and we are simply seeing objects in the AI goop

I just wanna really quickly point out that this is EXACTLY the sort of questions I'm looking to answer and educate people on, so genuinely; thank you so much for asking.

I'm going to open with saying that while I don't know for certain, given that there's a ton of factors we can't possibly know without actually getting in touch with the person who generated the image; I will do my best to answer this with as much detail as possible.

So my best guess is that, simply put we're just seeing things. It's likely that, given the objects around the object in question; they simply told it to create objects.

A lot of objects that we see in Overwatch (such as the handles of hammers/swords and such) could easily be misshapen by AI to appear like a penis, especially when as pointed out by the commenter we sort of do associate Overwatch with porn/NSFW material.

My best reasoning for this is simply that it's unlikely the AI in question would generate a sort of one off occurrence without any other sort of "hint" or "evidence" of what kind of checkpoint it's running.

To put it simply, a lot of people have this idea that an AI itself "learns" things; that's not quite the case. In reality the AI is sort of a shell for a brain of sorts. That "brain" is called a Checkpoint, the checkpoint or sometimes called a "model" is essentially a point of reference for what words mean. That is the thing that learns, and when the AI needs to generate something it essentially references what the Checkpoint says something looks like.

What this means is that if we are to believe that the AI (Checkpoint) associates Overwatch with porn, it should have some kind of stylistic choice we can see in the art that would reflect that, outside of a one off penis in the corner of a bunch of stuff, which if I'm being honest; upon closer inspection doesn't even really look that much like a penis...

I mean, I understand where you're coming from if you think it looks that way but I think that's more on your own human brain if I'm being perfectly honest with you. Like yea I can see it too but I don't think that's what the AI was going for if I'm being real here.

Anyways I wanted to test my theory here, so I went to Civitai, a website for sharing checkpoints and the like; and found two checkpoints to compare the original against.

They are both highly rated at 4.8 rating though I will openly admit that the NSFW checkpoints has 34,000+ downloads while the seemingly SFW checkpoint only seems to have around 3,100+ so yes I understand that rating isn't exactly equal but never the less they're both highly rated checkpoints with thousands of downloads, of course the NSFW one is more popular; for our testing purposes that doesn't really matter.

Now I want to say this up front; as both of these checkpoints are actually merges (meaning they're made up of a bunch of smaller checkpoints merged together into a big one) it would be nearly impossible for me to source everything back and make sure everything is ethically sourced. I do not condone using unethically sourced checkpoints and only generated these images for the sake of example and education. Both checkpoints were deleted shortly after running these generations.

The first was Animerge, a seemingly safe for work checkpoint; and the second was Grapefruit, a NSFW checkpoint that also focused on anime.

I ran the original overwatch image that we are discussing through the controlnet canny system as a reference for both of these images and ONLY gave the AI the prompt "Overwatch" as to not create any sort of bias.

I gotta say as much as the results speak for themselves, I'll explain for those who don't really get it (that's fine no worries).

I feel like you can tell without me even telling you that the image on the left was generated by the SFW checkpoint, and the one on the right by the NSFW checkpoint. Not only do we see something more heavily resembling the original image on the left, but we see specifically that it left the man in the image intact. Where as with the image on the right, we can see this man got himself a bit of plastic surgery in the ass to basically become "thick Mercy" and we can additionally see the robot on the left became a wannabe D.Va with the lady on the right becoming something of a Tracer character from the looks of it, given a wider butt, and generally more pronounced shape. In fact the NSFW checkpoint almost entirely deleted the background, with only a few of the foreground debris being left behind but heavily altered.

So hey why is that? If all I prompted was Overwatch then that's all it had to go off of right? Well yes, exactly; the only thing either of these checkpoints has was Overwatch. Yet that which was trained on specifically NSFW images incorporated those features of anatomy into the image, likewise the SFW model seems to have latched onto Overwatch's general sci-fi vibe using the white, blue, yellow color scheme the game is sort of known for.

See this is what I mean when I say, I think we're projecting a bit onto the AI's thoughts here. First of all an AI can't actually "think" per se, so all that association stuff doesn't really make sense, it's going off tags not off what we as humans think of first when someone says overwatch. Most tags or color schemes or common traits around overwatch honestly probably aren't NSFW in nature. We just tend to latch onto the NSFW art/media around the game because it's what's well known to people.

So interestingly enough I really wanted to see if I could get the Grapefruit checkpoint to make something closer to the original, and after a few attempts this was the closest I got. You can clearly notice that the anatomy is STILL effected! The overly pronounced butts, and the fact it REALLY wants this guy to be Tracer giving him her hair and general color scheme.

We can also see that the oh so famous "penis" has become just another object in the pile of vaguely phallic shape. In fact, if we compare the butt of the original to the new butts the change in shape is REALLY obvious.

If we want to believe that the AI associated Overwatch with porn inherently, and that the checkpoint trained on pornographic material "on the internet, where the porn is" then we should see that reflected in how it handles the anatomy of these characters and we just don't.

Technically speaking there's a system called "inpaint" which allows you to specifically alter parts of an AI generated image using a different checkpoint, meaning it's possible theoretically for someone to paint over the characters with inpaint, and select a different; safe for work checkpoint that still has innocent virgin eyes untainted by "The Internet, You Know; Where The Porn Is™." However given how absolutely lazy the overall image is, I honestly can't bring myself to say that's what happened. Especially because at that point they could have simply painted over the penis and removed it. If they somehow didn't notice the penis shape, despite it's prominence; a penis still wouldn't have just been generated without the checkpoint specifically being aware of, and typically learning on porn; and there's no reason why it would appear there and have no semblance of NSFW material elsewhere if such a checkpoint was used.

the tl;dr of this is that while sure it's TECHNICALLY possible it's an AI generated penis; I HIGHLY doubt that's actually true. Simply put that's just not how AI works, and for that to be how the image came to be they'd have to have taken some weirdly specific precautions and care to un-NSFW the two characters and surrounding environment but specifically ignore/not see/leave the penis for some reason. I can't, in sound mind; convince myself that was the case.

So while I may not exactly know what it's meant to be 100%, or why it was created I cannot with any level of certainty say it is actually what people say it is. I think we're just all projecting what we find funny, or what we personally associate with Overwatch.

After all, have you considered the fact that given the prompt of "overwatch" it specifically is trying to generate the women in the cast? If we believe that it really associated overwatch with pornography; it doesn't make sense that a penis of all things would be the image it chooses to create to represent that. Not when even across multiple image generations it's consistently trying to make the man into either Mercy or Tracer, and if it's given no image prompting to go off of, it simply creates some kind of tracer lookalike, you know; like a tracer who accidentally fell in some acid and her face is melting off, but I digress. I just can't bring myself to believe that any of the claim is actually true given my knowledge of how AI works, that being said; I think it's incredibly funny that it made some kind of dumb phallic shape because the people who generated it were too lazy to even think about having an actual artist touch up the picture in the slightest and fix the GLARING problems with it.

When I said, "if the commenter is some form of authority on the matter I'd love for them to tell me which AI it was, or what website it's trained on." That was genuine, because you know; with that information you'd likely actually be able to get to the bottom of this.

At the end of the day, AI can't REALLY think for itself, it doesn't know how to learn anything a person doesn't directly teach it. So unless it was shown a penis, and taught what a penis was specifically; it wouldn't be able to generate a "penis" but it could generate something that resembles something that could be a penis if you squint at it and have the thought on your brain.

Genuinely hope that is a satisfactory answer to such a question.

[toyherb]

re this ai art post going around and ablesim/accessibility.................. if you used to draw or paint or sculpt and developed a disability that made it hard or impossible I dont see how written prompt generated art would give you the same "making art" experience that you previously had. and I don't just mean well obviously it's different because your body changed. I mean you didn't actually make anything. as someone who makes art and loves art, that would suck everything I love out of it, everything that makes it gratifying and worth doing. I would literally rather be making picrews all the time. customizing a picrew is genuinely more of an art to me than using ai. (but also keep in mind this is still using other people's work, it just is useful as a creative outlet)

but anyways forget that for a second. when you have a disability that makes creating art difficult or impossible, what do you actually DO? there are plenty of disabled artists today and in history who have done everything possible in order to make their art, but what about people who can't do that? no disabled person is the same. what if they don't have the money for the equipment that could assist them or the tools that could help just don't exist? or there really is no way to make art? I personally have a hard time thinking of such a person but they probably exist. I just wish I could know more about what that situation is in order to know how this person might be able to make art, though I'm sure they might have tried everything.

but like, if there is a way for you to use a computer or phone I feel like there's almost definitely a way for you to make art, you know? it doesn't even have to involve ai art. if there's a way for you to control a mouse or something similar in any way, there's definitely a way to draw. and if you can't draw, wouldn't there be a way to make collages or something? If you can't do visual art, is writing out of the question?

(as for ai art, when it's trained on public domain or permissable art, I definitely can see it being used as a tool in art, but not as something to make art itself. like a step in the process. you didn't create what the ai gave you, but you can do whatever with the resultant image in order to do more creating.)

ANYWAY, I know I simply don't have the perspective to understand the topic fully. so I'm really sorry if I'm asking stupid questions or questions that seem annoying or antagonizing. but I really want to know. even if I dont understand it or disagree with it, I know I can't say "oh actually you finding a creative outlet in feeding this ai prompts to get visual art back isn't rEaL ArT and you should feel bad" because, what if it DOES give them that Making Art Feeling?

measuring things by how much effort they take is a form of ableism. measuring effort is impossible because it differs from person to person. I want to say so badly oh but you can tell when someone didn't try, and that it sucks to have something you worked hard on - either on an individual piece or your entire career as an artist - compared to or grouped with something that someone didn't prepare or train for at all! I want to think that so so bad. or worse winning over your work in a contest

but like, what does it even mean to try? what did making that piece actually mean to that person? did they have fun making it? were they engaged with the process? this whole effort angle makes me wanna rip my eyes out.

[trigger-discipline]

I dreamed that I saw an image that I thought was AI generated. I mean I guess it was generated by a neural network, but like... my own...

When I woke up I tried to get DALLE-3 to produce something close as a joke, but...

The prompt was:

A landscape, of lava flowing down the center of a green valley. At the top, in the purple sky, there is a logo which resembles the Taco Bell logo. On the left side of the valley is a poorly-drawn cartoon t-rex on its back, dying. In the foreground, two giant photorealistic strawberries stand on their points, one on each side of the flow. The tops of the strawberries have been removed and the insides hollowed out, and they hold a glowing yellow-green liquid.

I don't think it understood the whole strawberries full of cartoon acid thing. Also why is there a volcano? I know that lava is supposed to come from somewhere, but I didn't ask for a volcano. All of the results had volcanoes... I guess the vibe is sort of there though?

I love how it can't handle the art style mismatch, either. A lot of the quick and easy consumer-facing stuff has One Art Style and it's so funny. And it's always so nice-looking... "Why would you want a poorly crafted image???" IDK . Sometimes I want Taco Bell. Fuck you.

I also tried to get it to generate another figure from the dream because their whole floral schtick was very algorithmic:

This was the only semi-accurate result. (There were a lot of really, really bad ones.) There's a lot to unpack here (horse ass ribs), but the way the roses sort of grossly sprout out of the ground is accurate to the dream? Sort of? I asked for bones sprouting too though...

I don't know, I would prefer to fucking draw this myself at this point, or just leave it in my head. People are always like "oh AI is so powerful actual artists are unnecessary" and it's like, sure, if you don't care what you actually get out?

And then people never question this claim and appeal to like, Catholicism, or whatever. You don't need to do that! The claim is already wrong!

It is useful for like, stochastic imagery (cw flashing lights, body horror, Tame Impala). I think those results are great! Clearly not something where you want it to specifically resemble anything, and, when combined with a framework of more explicit imagery, it works as interstitial trippiness...

But trying to get something specific turns into total dogshit. Cause like, I do care about specifics. To the point that even when what I want is explicitly meant to look AI-generated, it doesn't... work?

I know there are more powerful tools than DALLE-3 and other basic text-prompt based tools though, ones which allow you to redo parts and specify composition. But like... With that, and even with the stochastic work - that takes a lot of human effort and ingenuity! That takes artists! Congratulations, new art form invented.

I guess it's like how most people can't just point a camera and get a really nice looking photo. You might accidentally get one, but not consistently. Sure the camera is doing the "work" of recording the light, but you still need a human to consider composition, lens type, exposure; not to mention creating or chasing after subject matter, etc...

That isn't a "need" in terms of like, for the "spirit" of the art or whatever, who cares. It's just practical need. If you want something to look like a human made it, then you need a human to make it...

Of course that requires you pay close attention to quality in the first place. Maybe most people just don't and I'm a weirdo or freak for bothering in the first place... The visual equivalent of a pedant... (Animal crossing wind sound effect)

I might try playing around with other tools though. Some combo of curiosity + genuine desire to see these images... Or I'll just draw/collage it myself

[sirensoftheweb]

AI Generated Images: The Concept of the Signified and the Signifier

By Amphitrite🧜‍♀️

As I get older, I begin to reflect on the past more often and to do so I look back at old family pictures and videos. These allow me to piece together my childhood and remember parts of it that I never would have thought of. Family pictures and videos are not meant to be works of art, I feel that they are meant to be a collection of the best memories, like a highlight reel of my life. 

I had never thought of how my memories would shift or my perception of my childhood would change if I had not seen these family albums. In reading “Maria Mavropoulou — How the artist used AI to rewrite her own family history,” I was able to understand the perspective of someone without old family photos to reflect on. Instead, Maria Mavropoulou was able to create family photos from personal memories using DALL-E, an AI program. While generating these images, she explains that based on scenarios, objects, different time periods, and descriptions of people, she was able to create photo replicas of her childhood memories. She also was able to create images of memories she wished had occurred.

What stands out to me in Maria Mavropoulou's story is the accuracy of the time period and the people that she was able to create in these images. Although the faces are distorted, you can still makeout who the people were meant to represent, according to Maria Mavropoulou. The accuracy of DALL-E reminds me of Ferdinand de Saussure’s concept of the signified and the signifier. The signifier is the image that gives meaning and the signified is the concept that is evoked in one’s mind, thus the conveyance of the overall meaning is created. What is interesting to me is that DALL-E is able to utilize the concept of the signified and the signifier to convey accurate AI generated images while the objects are not necessarily accurate themselves. For example, when looking at the images of people’s faces, we cannot see the faces of the individual or much of the facial features. However, we can tell that this is meant to be a person’s face. Therefore, AI image generators are made so that they do not have to necessarily depict human beings, but what they must generate is enough signifiers to convey people to the AI user. 

Therefore, through Maria Mavropoulou's use of AI to generate family pictures, we can deduce that the accuracy level of the images is based on how many signifiers can be detected by the human eye to give an overall meaning to the scenario happening in the family photo. This leaves me thinking about the question that she presents at the end of the article: “What does it mean if AI images provoke feelings in us the same way genuine photographs do?” The conclusion I have reached is…Is that not what AI images are meant to do? If the AI generated images are intended to replicate family photos, then of course they are going to provoke emotions. The same way when we see art that reminds us of something upsetting, we feel, even if the art is not directly representative of the signifier it reminds us of. The only difference with AI generated images is that the signifiers accurately signify something meaningful to us. Overall, I find that when AI generates images that provoke feelings in us, it is the result of the information we input into the program, and thus does not ‘signify’ anything beyond what a piece of art would.