Sophie Labelle | Instagram | Bluesky
Transcript
four frame comic.
first frame shows a typical AI poster that says : All chatgpt flyers look the same, especially with this font! the text above it reads : I won’t go to your event if the poster was made by AI.
second frame shows a still frame from a Hank Green youtube video. text above reads : I won’t read your post or email and I won’t watch your video if they were made with AI.
third frame shows an AI restaurant poster of burgers and chicken in a window. text reads : I won’t patronize your shop or restaurant if it is advertised with AI.
last frame shows an MS paint window with badly written : Your bad art is better than AI slop. text reads : if you can’t take the time or make the effort to interact with me yourself, then why should I do it?
I find it really hard to even read AI text. It’s exhausting, like being on Zoom for too long.
It’s something like decision fatigue. You’re constantly having to criticize every word because you know it will make stuff up out of whole cloth. Reading something you’ve verified to have come from a trusted expert or even just a human being is far less stressful.
Yup. AI writing also tends to place emphasis in really weird places and use it waaaay too much. Every single point is the Biggest and Most Profound Thing EVER!!
It also tends to use rhetorical structures that have a flat distribution of importance, surprise, etc. across all sentences. Natural language flows, it knows how to lead you to its conclusion because it has a clear idea of what it wants to say. AI writing tends to be more like an easter egg hunt or a mr beast video in that each sentence changes the direction of the point that’s being made. It kind of ends up jerking you around. I suspect this is because that type of writing has a very small chance of generating into a corner, because it can always pivot. Each individual point stays clear and coherent, but the overall structure ends up disjointed.
Never been more self-conscious of my language as I am right now lol, but I’m not going back to edit anything. It is how it is
Gen AI “art” is certainly bad but it’s not as bad as the babyfur diaper fetish pedo slop produced by Sophie Labelle
I get it, and I do somewhat agree. But at the same time, AI helps me to execute creative tasks I couldn’t normally do and some things I can, but AI saves time.
But I don’t just accept slop. I take the time to polish to get the results I want. It’s still my idea and my vision, I just need a bit of help. This allows me to get results fast and fail faster. I can spend weeks in blender in my spare time building what I want, and I still do for some things.
This post may just inspire me to finally open Gimp and make some truly horrendous comics.
You should also check out Krita. David Revoy, artist of Pepper and Carrot, has a ton of great tutorials on using it.
What’s the main difference between the two? Also I will likely be literally drawing stick figures, so this might just be academic at this point.
I actually used to have a lot of fun making art in Photoshop back in HS, but it’s been a long time.
If you’re used to Photoshop, Krita is way easier to use, the UI is almost exactly the same. Unlike Gimp, it doesn’t try to re-invent the doorknob.
I think if you use a pen the dynamic controls work better, and brushes are less of a pain to set up. I prefer to draw in krita and then fuck around in gimp.
Krita is a powerful, free and open source, digital animation tool where Gimp is a mainly focused on image manipulation.
Check it out, it’s pretty cool and it’s got a huge community behind it. It’s also free.
I think krita classifies itself as a digital painting software. It does have animation tools in it.
Follow up question: I made something but when I try to post it here it fails with an error message of “too many pixels.” I assume the file size is too big? I didn’t really know what to set the dimensions to when I made the file. I tried making it a lower quality jpeg but it doesn’t seem to be helping.
Any idea what’s wrong or how to fix this on either Lemmy’s end or in Krita?
Not sure. I’ll check and report back.
So, I looked into it and .world uses the defaults for photo-rs. Looking at the default config file for photo-rs these appear to be the image hosting limits.
[media.image] max_width = 10000 max_height = 10000 max_area = 40000000 max_file_size = 40
Do it. We want to see them.
I love horrendous comics, if the premise is entertaining and you keep working on it then you’ll get better at it too.
But then it wouldn’t be bad art anymore.
Truly the worst outcome
I can only upvote you once. But imagine I held down on the screen for a reaaally long time when I did it.
Also web articles. The moment I see the obvious AI tells I leave and block the website from search results.
Love how derpy they made Hank Green look
Seems like a bit of an overreaction to lump him in here tbh. Didn’t he only use AI to help do research for a single video? And decided to not bother again because it sucked?
I think he was downplaying his use. He said he was getting addicted to the “dopamine” hit it gave him, which sounds like he was using it more than he claims. I don’t believe his scripts weren’t at least partly written by AI.
In my industry everyone is using AI for their logos and their advertising and they all look like absolute crap that is indistinguishable from the competition.
A company logo should look good in black and white at 1 square inch on the back of a golf card. These designs are cluttered hot messes.
That’s probably the way things are going. All of the decoration on everything you get in a store will be AI generated. Personally, I would like it if the packaging looked little more than like what I imagine 1940 military food rations did: plain tin cans and brown paper packaging with nothing but a description of the contents in plain black lettering. You could have a photo of the contents but that’s it. That might sound depressing, but is it more depressing than decoration meant to get you to buy, buy, buy? Alright, I guess the decoration is also for branding, but you can do that with black lettering too.
AI is good for prototyping stuff or creating dirty one-shot application for simple tasks. In this case, you can try to get inspiration on how the logo should generally look like and then you go to a professional with the concepts you like and work something out that is actually showable. People are using this stuff wrong - but i am not surprised anymore after reading the article where they middle/upper management of a business was burning stupid amounts of tokens to convert pdfs to slideshows instead of using a specialized tool for that.
If you tell workers they have to use AI they will use it as much as possible on things that it is not well suited to because there are dedicated tools that will do the same thing. Malicious compliance.
The market is saying something different. To the outside, you look like an old man screaming at the kids on your lawn.
From my “outside”, you look like a tech bro pushing the new NFT’s.
Note:
frames one and three should not exist anyway. They are advertising. It was already malign before generative AI.
And Hank Green, despite the morons trying to sound clever by saying he did, did not use an LLM to write the script for the insert segment in that interview. He used a phrase, which was already a common diplomatic phrase, in a completely normal way. The phrase has made its way into ShartGPT outputs because it was already common, and people desperate to seem like they aren’t fooled by LLM output are grasping at straws to find things to point at as ‘tells’ and turning into anti-tech-priests, consulting auguries to check for demons in the words of strangers.
There are no nice, neat tells. There are dozens of models, each with dozens of versions, each with some things that MIGHT serve as a tell, but you can’t keep track of them, and the right prompt will disrupt them enough to ruin the tell, so you can give up. You can’t reliably tell. That’s the whole point. Enjoy living in the derealized, boiling world our annointed leaders have made. Hug your loved ones.frames one and three should not exist anyway. They are advertising. It was already malign before generative AI.
Advertising as a concept is needed, though. Telling someone to come to your house on Friday is technically advertising.
We certainly need to regulate the shit out of mass advertising, but I always found the small flyer on a landline pole very homely.
Inviting someone over is not advertising.
describe or draw attention to (a product, service, or event) in a public medium in order to promote sales or attendance.
Meets the definition. I just want to make sure we make distinctions between general ideas and why we actually hate the concept in the modern day.
Telling someone to come to your house
public medium
An ad is an ad. It being a bad idea doesn’t change that fact 🤷
You don’t see the contradiction in those two concepts? I was hoping you’d take a look and figure it out but let me explain, and hope you get it.
‘Telling someone to come to your house’ specifies someONE to whom you are directing the invitation. Public communication is broadcast communication, not in the sense of radio waves but in the sense of ‘casting’ (throwing out) the media to a ‘broad’ (non-specific) audience. You are trying to combine two contradictory levels of specificity.
You don’t see the contradiction in those two concepts?
No, not really.
I was hoping you’d take a look and figure it out but let me explain, and hope you get it.
Why would I when you didn’t give me the same grace? from what I said upstream
I just want to make sure we make distinctions between general ideas and why we actually hate the concept in the modern day.
Well here we are arguing definitions and ideas and not the reasons within these modern ads are oppressive. Not talking about the algorithmic nature of mass surveillance and privacy breaches needed to power modern ads, but “no this ad is different” surface level thinking. But if you want to talk about it, so be it.
‘Telling someone to come to your house’ specifies someONE to whom you are directing the invitation
Okay. make it a group you’re at. Throw a flyer on a telephone pole. Same deal. We had ads before the interet.
You are trying to combine two contradictory levels of specificity.
I don’t think the levels change the intent and definition of an ad. You want to tell people about a thing, you use something to tell them thing.
What if I use AI to generate images that would be as bad as if I did it by hand?
Read the comic again, because the answer is clear.
It’s never been more acceptable to post Microsoft Paint art in legitimate business spaces than it is now. Well, except for briefly in the late 1980s.
Except of course if your paint art is to abstract or surrealist,
then we must miscredit it and grill someone else for it being an out of context social commentary that holds the wrong opinion.
Such is the way.
If you’re talking about those MAGA douchebags then yeah, fuck those guys.
Speaking as a hard left, anti-fascist, anti-imperialist anarchist, i cant believe we are speaking about the same people.
And yet we are.
Anyway sorry to bother you, i just wanted to check out the source of the current festivities in person and thought that comment was too ironic not to comment on.
The restaurant AI slop is the worst… even a shitty cell phone picture of your food is better using AI on your menu… i truly dont understand
I would take Word (or LibreOffice Writer) bullet point list with prices and ingredients list over “pretty” AI menu at this point
I love the high effort approach I saw in Japan, where they make little models of the food and put them in display cases (of course not all restaurants but some)
Online clothes shopping has been rendered completely unviable now.
On amazon, sometimes I want to see the four different colours something is available in, but it’s the exact same image colourised four different ways, so you all you know is that it doesn’t look quite like the picture.
I love this story.
“Yum, slop” is the title of a Reddit post from June 30 that has more than 670 upvotes and features a close-up picture of the Lozanos’ menu display. “I reckon utilizing AI imagery for the storefront menu board doesn’t correlate with a long tenured small business in this neighborhood,” the poster commented.
Another wrote “I’d rather see s—tty photos taken in a light box with a 1 megapixel camera than whatever this AI s—t is supposed to be.” Others called the focaccia in the image akin to a “loofah” or “lung,” said the “Frappuccino” drink looked like it was wearing “fishnet stockings” and wrote that the slaw topping a pulled-pork sandwich reminded them of “the mutated cordyceps from the Last of Us.’’
I spent a little bit but not too much time on this, I hope you enjoy it.
Every generation that decides a new image machine is rotten thinks it noticed first. It didn’t, and there are receipts.
Grant Taylor’s essay “The Soulless Usurper” walks through how the art world greeted early computer art, and the complaints are the same ones being made about generative images right now: mechanically sterile, nobody actually decided anything, a machine getting slipped into a chair reserved for the artist. This you?
Taylor documents sabotaged machines and physical attacks on artists in the late '60s and early '70s, plus a long tail of milder stuff — eye-rolling, flat refusal to call it art at all — running from the medium’s start in the 1960s to the early '90s. His verdict: computer art was “possibly the most maligned art form of the twentieth century.”
Two dates worth holding together. That period is half a century behind us, and the takes aged badly — the plotter drawings critics wouldn’t call art hang in museums now. And the essay came out in 2012, a decade before Midjourney and Stable Diffusion. That’s what makes it useful: Taylor isn’t arguing about AI. He’s describing the previous fight, written before anyone had a stake in how the comparison would land.
Source: Grant Taylor, “The Soulless Usurper: Reception and Criticism of Early Computer Art,” in Mainframe Experimentalism, ed. Higgins & Kahn (UC Press, 2012), p. 18.
You’re conflating ideation with creation.
Walter Benjamin, nearly a century ago now, established a fundamental argument that most cultural critics of today have forgotten or simply gloss over: fascism twists art into a simple aesthetic rather than understanding art as a means to express change in the material conditions of the culture and its people. By relegating more and more to the machine, the machine gives us a false sense of expression without ever actually affecting our own material conditions. We are in the era of mass creation of the simulacra that Jean Baudrillard warned about.
Critics disliked Van gogh in his time too. And he used all the traditional kit.
Art had a sort of schism once photography took off. Pro-ai art seems to get stuck on one side of it. Traditional art had a long, long run of ‘good art’ being well composed and skillful reproductions of reality: people, places, events—with a few visual metaphors for spice.
But then came photography. And all those people using art to look back in time had options, cheaper and accessible options. The demand for the skill to make precise replicas nearly collapsed.
What has sustained was not the technical method of making art, but the expressiveness of the artist. The spice becomes the meal. Things become more referential, abstract. It envokes states of mind often contextual to the times and circumstances the art was produced, and drawing on concepts outside of the image itself. It’s not just about making pretty pictures from paint, charcoal, or photoshop.
Once you have the dadaists, Duchamps and Pollocks running around, arguments about ‘correct’ techniques to make art are always going to be shortlived. It’s not the advancement of technology that people question. Longish ago most of us collectively decided method didn’t really matter, intention did.
If you scrolled through your camera roll would you consider it art? Can you use ai to quickly copy a technique? Yes. Can it be applied critically to success? Maybe. But if you outsource the practical skill and contextual knowledge that makes good, deliberate art, what are the odds that happens?
Thanks for taking the time to put together this viewpoint and sourcing your argument.
I disagree with the comparison to digital art. Computer generated art takes skill, effort, and an artistic eye to succeed. Not just anyone can pick up a stylus or mouse and make good art, though it is a bit more accessible than some traditional mediums.
AI on the other hand requires no effort or skill on behalf of the creator. Furthermore, any artistic style is just mimicry. AI models have scraped real artists work without permission and is being used to slap together plagiaristic content, no accreditation of any kind to the source.
It took less time than you think for exactly the reason you might guess. I guess I just see a different ‘brush’.
I’m glad you disagree. I’m fortunately old enough to remember being told that digital art wasn’t real art, and do be alive today where nobody cares, oil painting is cool, brushing up your oil painting in GIMP is also cool, using GIMP alone, still cool.
Just take your same argument and fast forward 30 years and you’ll get ahead.
There are undiscovered art forms and new/reset expectations just around the corner. The first real artist (with real artistic genius) to adopt the tools is going to be in every MOMA around the world.
Your mistake is in positing that the MOMA is in any way relevant at this juncture.
uh, i’m not an AI fan but I couldn’t care less if mom and pop diners or volunteer ran orgs are using ai to generate posts for their causes. getting huffy over it is dumb af. If I see a multimillion dollar corp using it then yeah, get wrecked. They have the resources and the means to pay an actual artist.
![Your bad art is better than AI slop [Sophie Labelle]](https://lemmy.world/pictrs/image/c2feca8b-da2b-4e53-9ec3-04b436e86a0d.webp){kind=link}