why don’t we cut the middle man and just deal in prompts?
This is something a good buddy of mine can’t seem to grok. He loves generating ai stuff and showing it off and I keep telling him I love seeing his ideas, not what some corporate model decided to fill in based on his input.
I am actually, truthfully more interested in seeing his idea, the raw prompt that he entered into his SaaS chatbot or even local LLM, than I am with sifting through the generated noise for the signal of the mind of the man I love.
What, so you’d just read a load of words? It’ll never catch on.
I’m sure this is meant as a joke. But, really, prompts are the only part of AI art that is actually creative. Especially if someone has been trying to fine tune what they want for a while, the prompt might actually have something interesting in it.
Even though memes are shared as images, it’s often the text that’s the important and funny bit. The meme image is just context for the words.
I’m not saying I’m actually interested in reading people’s AI prompts. But, there might be an interesting idea in that prompt that the person could express in a way that’s not boring and soulless like GenAI tends to be.
This came up back on Reddit in the “what’s that book?” sub. Someone posted asking about a book; they had 6 lines of preamble about why they wanted to reread this, and 55 lines that were an excerpt from the book. This is normally a slam dunk–it’s pretty trivial to search for literal text. But nobody could find it. Discussion with the user led to the revelation that they’d asked AI about a book and described a scene they remembered, and the AI had responded with a title and “excerpt” matching what the user described. But then the user still couldn’t find the book, so they posted the “excerpt” to the subreddit. I.e., they were asking for a book containing an AI hallucination, rather than just describing the book to us the way they had to the AI. In fact the story they were searching for was “Harrison Bergeron,” one of the most commonly asked-about things on the subreddit, but nobody recognized it until we understood that the excerpt was slop made up in response to a scene they misremembered as part of the story.
It’s a frustrating time to be a human on the internet.
Man, that sucks. It’s really annoying how sycophantic these companies have made their chatbots, which makes this worse.
But, I think by the very nature of how chatbots work, you can’t fix that problem. Chatbots simulate what text is likely to follow other text.
If someone online has definitely read a book that really exists, and they describe a scene they remember from that book, the clear pattern online is that someone will say “Oh yeah, that’s the book X by author Y”. Some people might say “I’ve never heard of that book”, but it’s likely they might just not say anything. If we’re looking at a kind of forum with upvotes and downvotes, the person who found the real book will be the most upvoted.
So, that’s the pattern the LLM trains on. It knows that when someone posts a question like that, the best answer is one where someone names the book and types out the passage in the book.
It would be nice if the chatbot would instead say “I don’t know that book” or “That book doesn’t exist”.
But, it can’t do that, because it doesn’t “know” anything. Everything the chatbots put out is a hallucination. It’s just that some hallucinations are more useful. It has been trained millions of times on patterns when the passage someone remembers is “It was the best of times, it was the blurst of times”, so when it hallucinates about that sequence of words, its hallucination is bound to be more helpful. But, if what someone remembers doesn’t exist, or is obscure enough that it’s not in the training data, things break. Then it follows the general pattern used for every other book, and comes up with a “useful” answer that doesn’t match any actual book.
It’s definitely in the nature of the current-gen systems. It’s possible to imagine other ways to train chatbots that wouldn’t favor bullshitting in the same way, though. The way they currently are isn’t a way they necessarily have to be.
Anyway, it’s very clearly a case where it would have been better to cut out the middleman and just share the prompt, which was my point in sharing it.
It’s possible to imagine other ways to train chatbots that wouldn’t favor bullshitting in the same way, though
How? The way they work is that they predict the next word (token). They don’t know anything. So, when someone says “what’s the book that features an X doing Y and has a Z that looks like a Q” the process is basically “if this were a forum post, what’s the next word that’s likely to appear, what’s the one after that, what’s the one after that”.
If the scene being described is something from Harry Potter that has been discussed endlessly on the web, it’s likely to predict next words that end up being helpful. If the scenario being described has never been described before, there’s no way for it to say “aha, this book doesn’t exist!” Instead it just invents a plausible conversation between two people about a fictional book.
Now, you could design a filter that sits in front of the LLM that makes things better. It could be designed to recognize quotes, and if it spots a quote, it can “search the web” (or really look through a big index) to see if those words in that sequence have actually been seen before. If not, it could reject the entire response as being a “hallucination”.
But, fundamentally, this isn’t a problem you can fix through training, because of how LLMs work.
You can’t fix it with a pure LLM, no. But very few of the major models are pure LLMs anymore. They’ve got LLMs built in to them as subsystems, but they’ve got control systems layered on top of those to adjust their behaviors. And those don’t have to be trained purely with next word prediction.
The problem with that is that, it seems like it will become an unwieldy piece of spaghetti code, even assuming the code wasn’t written by LLMs.
An LLM is a crazily complex thing but at least it’s text in, text out, one input, one output.
As soon as you start bolting more and more things onto it to deal with special conditions, you have to maintain all those things, and understand how they interact.
I think I’d rather read someone’s prompt instead of having them paste me the output. I can always run it myself if I want, but a lot of the time I’d rather cut out the AI
If you ran it yourself you’d get a different output.
Yes, but if they didn’t care precisely about the output, then that doesn’t really matter.
The point is the output. If they’re saying just give me the prompt and if I want I can run it myself. There is a meaningful difference if the output is different.
I think most of the time that’s true. But, not always.
For example, I’ve seen people put up funny “remixes” of the “distracted boyfriend” meme. A lot of time they take historical figures, or celebrities, or characters from movies, or something that’s in the news. Now, I prefer the lo-fi versions of those memes where someone just spends 2 minutes in Paint and slaps a few photos together. I’ve also seen GenAI versions, and occasionally they’re good, even if I have an aversion to GenAI stuff.
Now, if the GenAI prompt were just something like “Distracted boyfriend meme but the boyfriend is Trump, the girlfriend is Lindsay Graham and the other girl is Putin”, that might be funny, and you can just know in your head what it looks like. That’s even a good lo-fi way to do it when everybody knows the meme well and can just imagine it. But, if you have a specific idea in mind that’s hard to convey, it can be different.
Sometimes the prompt is hundreds of words long and has to describe the various people’s clothing, their poses, their expressions, then other details like the background, foreground, lighting, etc. Sometimes in that case the result looks really good, and it’s really hard to parse that all in your head.
There’s an art and a skill to doing the kind of creative writing necessary to generate a good AI image. Sometimes you need to know obscure things like specific hairstyle names, car models, types of furniture, lighting techniques, facial expression details, names of articles of men’s and women’s clothing, how to describe poses and posture, etc. But, while that might generate an image that’s exactly how you imagined, it was never designed for a human to understand, and can be really painful and/or tedious to read.
Yeah, one of the things that I WILL say there’s an art to… is trying to get Genmoji to generate emojis of copyrighted things without triggering its programming. All creativity goes into basically doing everything you can to walk around the obvious
I would actually like to read people’s AI prompts.
A a series fb post from an elderly lady attempting to prompt facebook into simulating her deceased child.
“Keep it realistic. don’t make make them sound too kind.”
well engineered prompts are like a dive into the “psychology” of whatever AI tool you are using. I like the ones that go like “whatever you do, don’t put an elephant in the picture”
And then an elephant appears because “elephant” is a word in the prompt.
It’s interesting because in some ways it’s as if the AI models have a “peekaboo” level of understanding of the world.
If your prompt says something like “a man’s face is hidden behind his hands”, the model goes “aha, I see the word face, I better draw a face”, so you have to get tricky and say “a man’s hands are held up covering up the frontal area of his head”.
Funny thing is it doesn’t happen in LLMs, even if you use couple generations old and in “low effort” settings (which makes mistakes much more often). I don’t know if they have multiple layers of LLMs reconciling the output with the prompt (like how DeepSeek filters certain…ahem…“topics”)
Makes you wonder if it is an architectural thing. It still used a place holder to refer to the elephant in both cases so maybe the AI generating images just use their best version of “that”, don’t know.
It seems to me that whenever something makes the news or becomes a meme, they put in a special effort to make that thing go away. I think that involves things that filter the output. I don’t know if that’s another LLM layer or what, but some kind of filter (or series of filters) that addresses certain well known bugs. Even if the filter list is thousands of entries long and is basically “if output contains X, reject it and run again”, it’s probably a relatively cheap part of their pipeline.
it performs better in language in art they’re all pretty inept, the companies thought this would work, but it doesn’t, if you say something like ‘‘A painting of a large vivid skull and a dusty old rose’’ your going to get generic skulls and roses that won’t match up, because that’s the visuals it can find with Skull and Rose, it won’t give you something atypical.
Cutting people is illegal
ok can we cut them out instead
deleted by creator
Like fiction books?
yeah if gen AI progression keeps going with the same speed and hardware costs manage to decrease "turn this book into a graphical novel/computer game"s will be on the horizon.
Something ive noticed about AI art is it enables people who see themselves as “idea guys” to pump out every thought they have, call it good enough, and display it proudly like they worked hard at it.
When you make art you take an idea and that idea is shaped by external factors. Your skill level, your medium, how you collaborate if you are, all of these things influence what you make and that influence in turn changes your ideas. You may have an idea of what you want to paint, and during the process of painting come to appreciate the beauty of texture, and the next painting seeks to explore not just your idea but what you can do with the medium. Artists usually seek to push themselves to learn and explore and have their art change them as much as they seek to create the art.
But an idea guy only has interest in the idea. He doesnt want to do the process, he has no desire to be changed by his work. He has an idea and he wants an end product, and AI allows him to bypass that pesky learning to draw/paint/create business. It removes limitations in medium and skill level and spits out something close enough to his idea that its serviceable.
And I think that, to me, really signifies how much the idea guy values ideas over everything else. It takes all the work out of creation, and the lack of craftsmanship in the ideas is very telling in the final product.
It’s autotune for business majors.
This guy understands art.
There is also art without any idea, where the artist simply went with the flow. And there is commission work where the idea isn’t the artist’s. Also I think “idea” might sell short what someone has: an idea can be super highlevel leaving a lot of freedom to how it is represented, or that person can have a very detailed result in their mind they would like to bring onto paper or into sound or whatever.
On top it’s not like AI gives you an exact result. Quite the contrary; throw the same prompt at it a few times and you get different results. Potentially vastly different. So if you want an AI built representation of your idea, you still have to iterate and steer a lot until you get something that somewhat resembles what you had in mind. Besides having an idea, you also have to curate. But anyway, the process also shapes the result here; the constraints are different ones than for “real art”.
Mostly agreed, but you’re putting the person who creates an AI image into a box.
I may make an AI image for personal reasons and because I can’t do anything more than rudimentary Photoshop, and even then I have to have a starting image.
My art lies elsewhere. We could spend two hours while I show you around my house, show you all the things I’ve created.
People who make one kind of art can undervalue another kind.
Yep, it’s been real disappointing seeing musicians and writers use AI album/book covers
is the accumulation of one’s youtube feed truly the deepest gaze into the soul?
Naaah. Nowadays YouTube just represents the pinnacle of enshittification. Not something I’d admit publicly to using.
My Youtube feed is utter shit. I watch one One Piece theory video, and that is all Youtube starts recommending. I watch one trauma video and that is all my feed shows. Youtube is so annoying nowadays, more akin to dementia addled mind with ADHD.
I browsed some music on youtube on a fresh, logged out device a few days ago and it was so refreshing to get recommendations based the video I’m currently watching. Like 15 years ago. Now, if I’m logged in, the recommendations next to the playing video are the same ones as on the landing page, utterly useless.
You can go into your yt history and delete videos and it will fix the recommendations
I deleted my whole history last week, and within a day YouTube was recommending me anime openings for things I have not had any interest in for 10 years. I’m pretty sure it’s at least partially bullshit.
I ended up just deleting my whole Google account. I’m making do with an RSS reader and yt-dlp for now, but I’m already so fed up with that whole process that I’m making a simple QT app that does the RSS reading and yt-dlp calling for me. Akregator fucking sucks, by the way. Every time I shut down my PC without closing Akregator first, it loses its shit and does not remember any read statuses.
I just browse on incognito (without an account). The home page is blank every time.
Yeah. Fuck YouTube and everything, but it’s still possible to curate a decent feed.
Yeah they are sure pushing some things really hard, while I have to reclick on videos I have already seen so that theme doesn’t vanish, like some stupid algorithm.
Pro tip for YouTube: don’t use the main feed. Instead, look at the recommendations under an interesting video for other ones that catch the eye, and save them into ‘watch later’ playlists by topic.
Also, don’t watch random videos under the main account, open an incognito tab for that. With these two approaches combined, YouTube gives me some great recs, especially for music.
The kinds of videos I always make sure to watch in a private tab are things like:
- viral videos
- videos recommended by a friend
- “how do I fix this broken toilet” videos, or something else that I need once and won’t ever look at again
With that approach, my recommended videos feed is still trash, but at least it’s vaguely on-topic trash. I don’t get recommended 500 plumbing videos because one time I needed to know how to fix my toilet. It doesn’t take the fact that I watched a video that 500 million people also watched as meaning “I want to watch all the popular videos” or “I have the same tastes as everybody else who watched this video”.
I can always tell when I screwed up and watched an off-topic video in my normal account because the recommendations go to shit for a while.
It’s also important to do the same with links from social media.
Yeah, that’s basically what I meant by viral videos or ones recommended by a friend.
YouTube keeps pushing right wing and pseudoscience to my feed even though I have marked “Do not recommend the channel”, blocked them, reported them, and sent feedback saying the “Do not recommend” doesn’t work.
Get FreeTube and never sign in again. Only get served videos from channels you subscribe to (in the app, not on your account)
This is the way. I use Freetube on Desktop and PipePipe for mobile.
I think what it pushes on me is bad, but then occasionally I look at it with a fresh account or an account I never use. The suggestions it gives for new accounts is awful. Mostly it’s just absolute brain rot content. It’s stuff that makes “Ow My Balls!” look thoughtful.
Even if I’m not interested in 95% of what the YouTube algorithm suggests, at least with my actual account it tends not to suggest things that are actually offensive to me.
It seems to me like there’s a lot of “engagement bait” in the suggestions it gives. Sometimes that’s good, in that it’s stuff you’ll engage with because you actually find it interesting. Other times it’s stuff that keeps you engaged because it makes you angry or something. It doesn’t distinguish between “You’ll love this video!” and “You’ll hate this video, but you’ll watch the whole thing!”
But, I don’t think we can discount that a lot of people who watch YouTube are absolute idiots who have 2 living brain cells remaining. Any time your feed gets a suggestion based on what other people are engaging with, it’s inevitably going to be terrible.
just being in a certain demographic will do that
It is when a person’s depth is measured in teaspoons.
It is to a smoothbrained techbro.
Around 825 AD, Persian scientist and polymath al-Khwarizmiwrote kitāb al-ḥisāb al-hindī (“Book of Indian computation”) and kitab al-jam’ wa’l-tafriq al-ḥisāb al-hindī (“Addition and subtraction in Indian arithmetic”). In the early 12th century, Latin translations of these texts involving the Hindu–Arabic numeral system and arithmetic appeared, for example Liber Alghoarismi de practica arismetrice, attributed to John of Seville, and Liber Algoritmi de numero Indorum, attributed to Adelard of Bath.[8] Here, alghoarismi or algoritmi is the Latinization of Al-Khwarizmi’s name;[9] the text starts with the phrase Dixit Algoritmi, or “Thus spoke Al-Khwarizmi”.
So in referring to what someone’s true soul is, you can say “Thus spoke The Algorithm”
Yes, absolutely. If you look at my history, it shows nothing but repeated watches of the dog of wisdom.
Says a lot about me, I think.
The deepest gaze into my soul is apparently that I enjoy disc golf, competitive pokemon battles and asian street food.
I actually closed my eyes and took a moment to prepare myself after reading the first sentence because I knew I was about to read the most brain-dead take imaginable.
They can’t even form basic logical opinions anymore.
Many if them couldn’t before AI, which is why it’s so crazy good un their view.
“From a distance, Portrait of Edmond de Belamy, which has just sold at Christie’s in New York for $432,000 (£337,000), looks almost plausible. Up close, however, the paintwork becomes a grid of mechanical-looking dots, the man’s face a golden blur with black holes for eyes. Look into those eyes. They show no sign of feeling or life. Did a computer make this? The answer is yes. The first artwork generated by AI to be sold at Christie’s, its impressive price would seem to suggest that in future we will get computers to make art for us. Robot van Gogh will harmlessly cut its ear off and robot Picasso will be a genius, minus the misogyny. Is this the future AI art visionaries such as the French collective Obvious, which programmed this “painter” by getting it to compare its own work with 15,000 pre-20th century portraits, have in mind? Or are they just, God forbid, making a fast buck from gullible art collectors? Because believing the algorithm that knocked this up to be in any meaningful way an “artist” is like thinking your voice-interaction programme is out to get you. Dream on. Computers would need to replicate human consciousness before they could replicate the funny thing humans do called “art”. Art is a way in which human consciousness expresses itself, and is equally true of the earliest cave art, Rembrandt’s portraits and Duchamp’s urinal. And that is what is missing from Portrait of Edmond Bellamy. Art is a way humans communicate ideas, perceptions and feelings to each other. It has no existence outside the human passion to communicate. So in what meaningful sense can an AI replica of certain physical traits of old master paintings be called art? For a robot to really make art, it would need an autonomous mind that was emotional as well as rational. No AI developer has yet claimed to be anywhere near achieving that and if they ever do, their creation will probably have better things to do than paint portraits – like destroy humanity. Maybe afterwards robots will invent their own kind of art, but it won’t be some poor pastiche of human genius. It will be beyond anything we organics could imagine."
I suspect this was made with the whole “Art collection as tax avoidance scheme” thing in mind.
So they’ve bought this shit that will be worth nothing in months/years.
I mean if it’s really the first AI “artwork” to be sold at Christie’s, it will be worth a little something being the first one of its kind, but it will will drop in value very quickly.
I predict that it will be offloaded immediately, it’s basically a printed Concerned Ape. If they are holding onto this in two years and it retains value well I’ll be dead, I mean I will have a stroke and just die right there. UNFATHOMABLE
AI art also helps keep starving artists unfed
Hey, that’s not all it does! It uses all the water so they remain unhydrated as well! As well as all the electricity to keep them unpowered, and ruin the sleep for everyone within a 60 km2 square radius, so they also are being kept unrested. Jeez this is the dumbest timeline. By a lot
60 km2 square radius
Dude here lives in a 6D space.
No that’d be 60km² hexagonal radius, right?
Afaics they measure the radius of the ‘area’ in km4, and I assume the ‘surface’ is one dimension lower than the total dimensionality of space.
Yeah I’m just kidding around.
Making it as an artist was always difficult, because the way artists are paid under capitalism favors giving all the money to a few big artists, who ironically aren’t necessarily even that good.
My goal all through junior high and high school was to become a comic book artist. I’m still sad I never made it because I missed out on the opportunity to help Stan Lee achieve a net worth of $70 million.
Yeah, like the Renaissance before capitalism! Let me check my notes, oh, uh, looks like “all the money to a few big artists” was the thing then as well.
this_engine_is_a_capitalist.jpg
More like, it favors giving the money all the people who make a living off of what the artist produces.
Its like a factory. The artists are the workers at the bottom who actually produce and make the art work. Then there are the managers who promote the art and their bosses who sit at the top, running the factory. It is always those who benefit off of other people’s hard work who get to enjoy the fruits of other people’s labor.
Artists are why publishing houses exist and the reason why editors have a job in the first place, but it is the editors who drive nice cars and live in nice houses while the artists who draw and write the very books the publishers sell, get to enjoy getting paid so little that they can’t make a living off of it.
If you look in places like the animation industry, you can point to any of the big animators who helped make Disney big in the 90s and you will see that none of them can get work. None. New animators have no chance. Artists who have been in the business for 20 or 30 years, who are responsible for the movies that defined a generation are jobless and if they get a job, it is the kind of entry level gig that you’d normally only give to a recent graduate.
Illustration had been fucked for decades. They are getting paid the same as in the 80s, but not adjusted for inflation. More often than npt they get offered jobs for exposure and when they land gigs that pays, it will be two months work for one week’s pay. Asking for more money is not an option. There are no unions. All they can do is work faster to maybe make the low pay worth it. Meanwhile their editors live in gleaming mansions.
You might end up being one of the lucky few scammers in fine art who become rich people’s favourite, but that is a completely different world from the commercial art industry. People work themselves into early graves. There’s a reason why Berserk’s crestor died in his 50s.
Back problems, wrist problems, eye problems, weight problems. And these are passionate souls, which is why it is so easy to take advantage of them and their skills. Most people in the animation industry burn out before they’re 40. You don’t see many middle aged artists in the industry today. Some end up teaching. Most leave the industry all together when those who benefit from their hard work has sucked every bone dry.
The higher ups get fat bonuses. They live the great life. Not a creative bone in their body, but they make a living on the piles and piles of passionate souls who will be eaten up and spit out with nothing to show for it.
That is the life of the artist.
The very, very few ones who rise to the top are often those who have a rich family member or spouse to finance their artistic endeavors. Illusteators, painters, authors, animators, you name it. More often than not, the ones who do well and are successful, have no real financial problems in their lives because their loved ones got them covered.
It’s easy to make an amazing piece of art when you don’t have to worry about money.
Artists are why publishing houses exist and the reason why editors have a job in the first place, but it is the editors who drive nice cars and live in nice houses
I agree with most of what you said, but I think most editors don’t make that much. It’s the corporate heads (i.e. the editor’s bosses) who get paid - the MBAs who’ve been put there to squeeze maximum profit at minimum cost.
a few big artists, who ironically aren’t necessarily even that good.
I get your point and all, but I just had to ask - how do you define and separate ‘good’ from ‘bad’ art, though?
I agree with you though, the traditional art world is no different from any other industry under capitalism. It’s filled with sharks, ass kissing and networking. Most either come from fame, and/or from a wealthy family who’s willing to support them. It’s more important to be a driven entrepreneur with good social skills and the right connections, than an artist.
how do you define and separate ‘good’ from ‘bad’ art, though?
It’s partly subjective. Think the quality of the craftsmanship. For example, for paintings, whether the perspective makes sense. Or the shading. Or in literature, whether the story in internally consistent (I think the term is verisimilitude). Of course, all rules have exceptions.
My god, finally. Someone on Lemmy solved the question that has been asked for centuries.
Yes, my middle schooler was going to hire a team of artists to create minion memes and dogs kung-fu fighting if not for AI.
Or, you know, your middle schooler can just draw silly things like we all did growing up.
I’ll pass along your suggestion
The algorithm rejected you.
I’ve seen what makes them cheer.
Weird way to brag that your child will not only always be a talentless hack, but won’t even try.
Weird way to dunk on a child.
Yeah, it’s totally the child’s fault that they aren’t learning anything, and not the parent’s. Why would parent ever be obligated to teach their child, right?
I would ask your parents if they could explain to you what a meme is and why you would find them in a meme community.
You’re going to have a stroke if you read every meme as if it were reality. There’s no middle schooler, the cute dog or cat isn’t actually asking to has cheeseburger it is a human writing a caption on a picture and Gigachad isn’t a real person.
I’m sorry to be the person to have to tell you all this.
That joke doesn’t work, I’m not in a wheelchair and that’s the important fact about the joke.
It’s me, I’m you.
the thing is, you give your thought to ai and ai generates what it thinks you mean through other peoples examples but you only have your idea of what you want. and you only have so many chances for ai to get it right
Engagement farming? I refuse to believe the top person is for real
Also Youtube feed is the videos they haven’t watched, so isn’t in their mind (yet)
deleted by creator
Top guy doesn’t even have the checkmark you need to monetise on xitter tho…
These platforms are built on the backs of chuds working for free. Someone did a study and most of the Russian propaganda is shared by ordinary people and some of them were dubbed super sharers who will share like 30+ right wing posts a day.
This is what traditional art reveals. And quite well.
Couple of years ago I was at a cannabis painting event (get stoned whilst painting), with my friend. I’ve got some experience drawing and sketching, very little with painting. My friend has very little experience in anything art related, yet whilst I was struggling to figure out what to paint he manages to make a lovely blend of colours and themes, in ways I never could. I was looking at it for a good while, and it had so many lovely elements. He kept thinking his wasn’t good!
He would’ve never been able to make that with AI, it would have missed all that nuance.
If you’ve had very little experience with painting, how do you know how good you would be at it? Your friend might have a natural talent, but that doesn’t mean that you couldn’t learn how.
Natural talent is a myth, it’s like believing that you shouldn’t think about what you’re doing to paint loose.
It’s all hard work, and maybe in the end you can split up artists by talent, but until then it’s just who have learned the most.
Prodigies exist. Some people learn faster than others, and some have ability that’s just leaps and bounds beyond anybody else. This is independent of hard work, but somebody who has talent but doesn’t do the hard work isn’t going to be very good. At one end of the bell curve are people who just aren’t going to get it no matter what, and prodigies are at the other end. Most people are in the middle: they can become technically proficient with hard work, but they’ll never be as good as a prodigy who works just as hard.
I disagree, with the exception that some learn faster (which isn’t “prodigy”).
Even the masters spent years, even decades training before creating their masterpieces. They copied, they failed, but above all, they worked hard for years to become good.
Of course the masters worked hard. I’m saying that no matter how hard you work, you aren’t going to become Mozart or Einstein. Hard work is important, but it also isn’t the only factor.
You ain’t gonna be Mozart or Einstein with that attitude! 😉
Sure, but people have to work for it, there are no shortcuts.
This is true. I’ll be going to Amsterdam again soon, and I’m hoping there’s another event at the time.
I chose to believe this happened. Jaden Smith has at least one friend and they were impressed with the art their friend made.
I will carry this story in my heart and retell it at the next paint and sip I go to.
That sounds like a lot of fun
It’s quite a nice event, they try to do them weekly unless something comes up. They sometimes do stoner pottery, or design your own bongs with the erm, right type of paint I would presume lol. I don’t want it to sound like advertising, so if you are planning a trip to Amsterdam send me a message and I can give the website/socials :)
But real art is not being sold for plebs with the promise of sticking it to “millionaire” artists that are all “full of themselves”.
humans can come up with
Let me stop you right there. This is almost too absurd to be a Onion article…
If someone wants to know the weird shit people have on their minds you just have to find their social media timeline, especially Reddit, Tumblr or Fedi. Or, like, talk to them.
Given the flavour of weird shit in someone’s head most AI wouldn’t even slop anything like that. Like, I’m thinking of three shrimps having a raging threesome right now with Jim Carrey’s head in the background watching one shrimp-dick explode. The AI would probably refuse or just make Carrey sit in front of a meal.
I challenge any AI to come up with a comment like this.
Thank you for that vibrant mental image
Seems like a fundamental misunderstanding of what AI art is.
Maybe most people feel discouraged from trying art because they think it needs to be good enough to get likes and shares online. If people were more encouraging to beginner artists or people who just don’t have time then maybe the average person wouldn’t care as much about AI art
The only reason people think they need likes and shares online for their art to be valid is because social media companies have been carefully and intentionally crafting that obsession with online recognition and growing numbers. This isn’t a “people aren’t nice to beginner artists” issue, because trust me nobody is kinder and more welcoming than the art community to newbies. Its a corporate social media issue.
That’s fair a lot of art communities online are nice. I guess I meant more cringe culture encouraging perfectionism
Imagine if people were going to the gym with a robot and having it lift weights for them. Then, they would post about how many reps they did. And then you said “Hey, maybe if people encouraged beginners to lift weights more, then they wouldn’t be using the weight-lifting robot!”
There’s a lot more going on here. The people using AI don’t even understand what the assignment is.
You’re comparing apples and oranges a bit here. Sure, using a robot to lift weights at the gym is analogous to using AI to create art, if your goal is to get better at being an artist.
But if you’re a person with no painting or drawing skills who uses AI as a way to express your ideas visually, it would be a little more akin to being a shitty machinist who uses Sketchup and a 3D printer to manifest their mechanical ideas in meatspace.
99% of AI “art” is absolutely a cringey slopfest, but I’m not going to lie and pretend I can’t relate at all to people who use it as a tool to supplement their own practical deficiencies. Whether art should be more about the output or the process is a whole other philosophical debate…
But if you’re a person with no painting or drawing skills who uses AI as a way to express your ideas visually, it would be a little more akin to being a shitty machinist who uses Sketchup and a 3D printer to manifest their mechanical ideas in meatspace.
It’s akin to getting somebody else to do it for you. Map out the workflow of hiring an artist to produce art for you. Then, map out the workflow of using an LLM to produce it for you. It’s the same workflow.
You’re not wrong. But the exact same point could be made about mapping the workflow of 3D printing a custom part to having a machinist prototype it for you.
In both cases, you tell the machine what you want, and it executes it more or less faithfully depending on how well you describe the thing. There is a direct parallel to the 3D printed “art” and copy/paste gizmos slopping up marketplaces–but we rightfully don’t claim that the affordability and accessibility of 3D printers is a problem.
IMO you need to work to become an artist, seeing alone is a skill you need to work on for example, so no newbie will be helped with prompting AI images, how on earth could they?
That philosophical debate doesn’t matter here, because with AI both the process and the output are worse than useless
I understand the OOP’s sentiment. There are plenty of us who have visually detailed and complex ideas, but are unable to fully express them via conventional media, due to lack of skill or disability. Not everyone can practise their way into being a master painter. It doesn’t mean I think of it as art, and I can’t understand why some people are fine with passing off something generated from others’ work as their own creativity.
Whilst true that not everyone can practice their way into mastery of painting, I usually find people don’t even try.
I definitely prefer honest garbage to polished inauthenticity, but, based on AI usage, it seems the ordinary person disagrees. It’s sad, don’t you think?
100%. Obviously-AI images induce immediate repulsion in me, and I hate that I feel the need to inspect normal-looking images to see if they might, too, be AI. For some I’ll never really know.
I also hate that there’s now certain styles and fonts that I now just assume are AI without any other evidence
Having a linguistic concept in your mind does not mean you have a “complex visual idea.” The whole point of “practicing” is training your mind against it’s nature and start to think visually for the first time. The tragedy of AI image generation is that it keeps people from exploring that side of their imagination and learning to really see. they never really develop a visual language but instead take the prepackaged style of the llm vending machine. It is sad to think that a voice is lost because they traded the vulnerability of making something to getting a predetermined product and pretend they had something visual in their mind to begin with.
You believe people who can’t draw also can’t… think visually? That’s a completely different skill common in most of the population https://en.wikipedia.org/wiki/Visual_thinking
Not exactly; People who can draw can use that tool to explore visual ideas and think visually; But you can do that with design concepts, paper cutouts, Adobe Illustrator shapes and so on. People who only engage with linguistic concepts as in the case of an LLM are not really engaging in the nuances of any visual language. They are using information set into a translator that outputs a result. It’s like thinking that Using google translate Makes me “Think in Chinese” lol.
I see it often that clients come to me thinking that they have a realized visual Idea but they really don’t have much. They have some overall aesthetic short hands. A couple of images they liked and a standard they want to meet. This is not visual thinking; this is called being a client. Then there is the job of art directors, they won’t draw everything themselves but they will offer a system, information critique, cultural awareness, even technical methods.
You can engage an LLM as a client or an art director; But in the end you are still using google translate to write in English and have an output in Chinese, if you follow my former metaphor. You are not really thinking in the true language of the output, There is wide cultural barrier between you and what you are making; most of the context cues and nuances of your message are lost or irrelevant.
You’re judging people’s thoughts by their output. Explaining visual ideas in detail, in words no less, is yet another skill separate from drawing and thinking. It’s closer to writing clarity, how well one expresses themselves in text.
It isn’t fair to think someone who’s bad at communicating ideas is equally bad at having them. This is the same fallacy that causes one to think LLMs are alive. People are more than their output, they have inner lives.
I don’t contend that people who can’t paint can think like a painter, but that visual thinking is more than just thinking like a painter.
In the case of the people you describe as ‘client’, it sounds to me like they just don’t care about what the result is, it’s just a job to them.
That said I do think there’s merit to expressing ideas yourself. Having to dwell on it to bring it into existence, to make decisions where the idea is vague can make you realize its shortcomings and see how it can be improved. The act of creating the product can make you aware of the strengths and limitations of your tools. That becomes a lesson for the next time, so you can have better ideas from the get go. And the experience from creating it lets you know which details matter when explaining it to another person.
I’ve met a lot of people who are very proficient in Putting down visual Ideas and most of them have a very humble approach to what they do. They speak about “experimenting”, “doing research” and “putting down options for review.” The Dunning Kruger effect comes to mind where The less someone knows about a discipline they are tackling the more you overinflate your level of skill. On the flip side; The more experienced you are at a discipline, the more you understand your limitations and become humble. In this sense It’s not aimless judging to make a guess that people using LLm’s thinking the have “fully formed ideas in their mind” are overestimating their abilities. LLM’s have a bad rep of exacerbating the dunning Kruger effect by cutting them off with engaging with the medium they are “commissioning” The LLM.
I have never seen a commissioner believe they created the image they commissioned, but for some reason LLM users want some recognition for being “Ideas men” We have always had Ideas men. Everyone has a dumb Idea in their head. Some of them are stuff that don’t need to engage with any cultural language or medium. But is there any merit to them? I’ve had clients that do care for the results given: A writer asks for a set of sketches for the architecture of their fantasy story; They ask questions, they beg for feedback, they even bring crude MS paint drawings. If a visual concept results from this collaboration, The writer won’t take credit for the visual part. They will be humble and recognize that they don’t have the background to generate a fully fleshed visual idea idea. But the work done during the collaboration, The back and forth enrich both parts of the process; An artist can learn a lot from a writer and vice versa. The is struggle, critique. It’s messy; There is work done.
This will sound like an unrelated topic; But I’m a huge music fan; I have a collection of hundreds of albums. And what disgusts me about AI music in particular is not the fact that “there is no talent” or “soul” but that it’s designed to kill the vibrancy of the music community. Everyone can take someone else’s sound, and tack a dumb masturbatory “idea” on top. There is no engagement with theory, music history, or with music scenes. There is no engagement with people. It’s spam. It’s Useless spam made by “idea men.” And there is an air of staleness to it. I said it on the initial post; There is a stench of cowardice to AI “creative” tools. The AI music scene statistically is full of white men making music that mimics Black women singers. It’s larping, It’s being unable to own up to your body and it’s limitations. Who wants to waste time listening to that? I feel “ideas men” Have a tradition of being like that. People that don’t want to put themselves in their ideas and instead outsource the work. At the very least conceptual artists of the 80’s and 90’s like Demian Hirst thought of a brand or a style. But Ideas men of the LLM go beyond that. It’s the signature style of the vending machine or some lora of work they stole.
It is easier than ever to make something and putting it out there, and it’s pathetic to see people make excuses on why they won’t do it. I know people in wheel chairs paralyzed that draw with their frigging mouths! There are no excuses.
The whole point of “practicing” is training your mind against it’s nature and start to think visually for the first time.
Speak for yourself.
It is common for people who start learning a skill to talk that way; They believe they have a strong grasp on some kind of thinking and in practice they find out they are way behind… They call it Dunning-Kruger something? In people who are learning art or design i see it all the time, people talk about their “visual ideas” and then get frustrated because they can’t put them into paper. A lot of their visual ideas are Imitation without personality. When the student starts to develop a language they gain an added feeling of empowerment: They have connected many forms of knowledge and experience that once existed arbitrarily in their heads. They have a method a system and i repeat visual language. It’s one of the biggest tells of AI slop images in the wild. You can’t see a person, you see the diffusion model itself, and often times since diffusion models need to take information from training that casts a very wide net it looses any personality or distinct voice.
Lastly; we often confuse “making art” with “making a product” and forget that art is a form of communication itself. If people rely on translators instead of learning a language, do they really “think” in that language?
You can make a product: Write a novel in German even if you barely speak that language by inputting your English text into google translate and carefully and editing the output as best you can. But there is a communicative barrier between you and your product. You don’t really think in German, you don’t know the slang, and the nuances of the language by relying on the translator.
True, but I’m talking about being able to visualise something in one’s imagination, but have difficulty in translating that into reality, so others can see an approximation. I agree with the negative affects on imagination, cognitive surrender is real.
So what then for those with aphantasia? They may be able to describe something in detailed words, but not visualise it themselves at all. In that case, AI really might be able to help them express an idea.
I mean one of my brothers is aphantasiac. He’s a great 12 string guitarist. He does covers, but even with practice not everyone can write music well.
That’s cool, but if he was having trouble communicating a visual idea, and he wasn’t good at painting, drawing, sculpting, or digital tools, would you think less of him for using AI to help him crystallise it and show it to you?
I have aphantasia. I still draw manually; I just need a reference photo (or subject) and “modify” it as I draw. Imagine if you couldn’t remember what something looked like so you have to memorise “ok their right eye is 30% down their face and only half visible from this angle and extends 20% horizontally at about 10 degrees until it’s covered by the nose which is…” etc. Then I draw that tiny line and have to check again
Writing it out, I just realised it’s actually similar to how you would describe it to an LLM 😂. I will admit that I’m very unsuited to this skill, and it’s extremely difficult for me, but I still find it fun. Maybe because of the difficulty itself?
Actually, lately I’ve been challenging myself by using pen and doing line-art. Normally, because I fuck up so much, I do messy sketches; so it was really interesting using a medium that is so punishing. I’ve learnt with line art it’s mostly about what you don’t draw, rather than what you do, if that makes sense
I fucked around with a locally installed image generator A.I. for a while in the past ~ it was fun, but I got bored of it quickly, and it became kind of hollow
I do not have aphantasia, but a painting technique I follow could work for someone with aphantasia. It’s more about pattern-recognition than imagining something beforehand.
I pretty much start by following whims. I pick some colors and do what feels right. Only after a bit do I look again at the work and think, “Huh, that looks like an alien” or “That kind of looks like a tree,” etc. It’s kind of like making my own Rorschach test - making sense of the noise on a canvas. Then, I start accentuating the pattern I see until others can see it, too. In the end, it’s a unique piece that looks like I planned it all along.
I’m not sure if that sounds up your alley, but I figured I’d share just in case.
Thanks! I’ll give it a go one day
Nice! Don’t know why you got downvoted just for sharing your experience.
I remove my own upvote from my comments because I like things starting at 0
The thing is; Our “Imagination” lies. Our brains are especially good at forming concepts that are often snippets of visual echoes and linguistic narratives but don’t form into anything concrete. In order to form a concrete visual idea you need to engage with visual concepts themselves. A character designer needs to have a grasp on human anatomy but there is design specific information like contrast and color balance, form, shape, line economy, design sense; then there is cultural information like emotional response to shape and color, particular fashion sense, cultural and environmental cues. All those are choices; a designer engages with them, interiorizes them, experiments and chooses what works best for them. Every Specific Area has different context and it’s own sense of rules.
A Good character designer might think they would have good “visual ideas” for graphic design but in reality this won’t be true until they engage with the particular context that is proper of graphic design. Since by using AI people are cutting themselves off to working with any visual concepts: (It’s all linguistic descriptions) They are cutting themselves off from making decisions that are truly visual. They are clients. Just like the clients that hire an artist that more often than not, don’t know what they want. But “think” they have an idea that is concrete in their heads. People that instead of immersing themselves in visual or cultural concepts take short hands. Reminds me of this thing of calling visual trends “punks” putting “cyberpunk”, “cloud punk” or “hope punk” in their prompts. Oversimplification. Visual stupidity.
As for aphantasia; Supposedly I have it. Back in college it was impossible for me to visualize images. And I used to heavily forget faces, places or shapes. I confuse people I already know with other people, Images just don’t stick in my head. But again. Visual education is not so much as a mental faculty but a system. I have a very particular system developed with experience and a formed visual language. So It doesn’t matter that I can’t visualize something in my head; Because I can get to any drawing pad and jot it down. Journaling is important for me. Often time I record the process I used to find a particular answer and keep them as “maps” to retrace my steps. As I developed more visually intelligence I got better at retaining images in my head. I feel my ADHD has a bit to do with it, because back when I was a kid it was difficult to find contextual cues, subtle proportional differences and cultural responses. I trained myself to get better at picking up on those over time, and my “aphantasia” feels less prominent today.
I think we’re talking about separate things, you about those who don’t fully know what they want to express, and that art comes from their attempt to express it, with the influences of their life experience, and should they offload that attempt to express a concept to AI they lose that idea (or never really knew it). I’m talking about those who do know what they want to express, visually, and regardless of the tools used, could look at the output and say with confidence ‘that’s not it’. It seems this is a bit foreign to you because you perhaps have never experienced seeing a full visual scene in your mind’s eye, and known you are unable to draw/paint it with a level of fidelity that even approaches the visualisation. I can imagine all sorts of things, see them clearly in my head, and though my drawing skills aren’t terrible, I couldn’t reproduce them well. I could describe something linguistically, but have no interest in spending time using an AI to try to express it. That doesn’t mean no people exist that have things in their heads that could be expressed via AI.
The thing Is I have seen my work being realized: through doing it. Through doing the hard work. I am not a musician but i love music, and I’ve done a couple of songs that i like by recording simple things and mixing them together. I don’t have music in my head. I experimented, did the work. Made a mess.
I think the problem is that there is people out there that has never experienced the satisfaction of creating and want a palm in their back for not even trying! Thinking that they have fully realized ideas in their head is a great cop out for this.
Yes, we’re most certainly talking about different things.
Get a pen and paper, that’s enough to make great art and express yourself.
If you try.
No shit, mate.
but are unable to fully express them via conventional media, due to lack of skill or disability. Not everyone can practise their way into being a master painter.
A master? Maybe not. But you can absolutely practice your way to being a decent painter. Enough to get your creativity expressed.
And there are plenty of disabled artists out there. I’ve seen decent painters who had no arms and had to do all their painting with brushes held between their toes. Don’t let disability be your excuse. (Unless your disability is blindness, I guess? It would be awfully hard to be a blind artist. But a blind person is going to have an equally hard time getting anything meaningful out of AI.)
I get it, but I don’t see it as all or nothing. If you want to be an artist, but you’re “creating” with gen AI, I’m not interested.
If you have something inside your head that you want to just show someone, and its not intended to be art, or remembered, then should you spend years practising to use some medium in order to create it yourself?
If you have something inside your head that you want to just show someone, and its not intended to be art, or remembered, then should you spend years practising to use some medium in order to create it yourself?
Yes, because creating stuff is fun.
How many resources does that use, and opportunity cost in time spent, vs a occasional generation? I have a fair idea. Creating stuff is fun if that’s your goal. Dedicating time and resources to become an artist only to convey a bit of info occasionally, not so much.
People have been communicating ideas through stick figures and poorly drawn lines for about as long as people have been drawing.
Edit to add: perspective and shading are also extremely modern techniques, invented during the Renaissance. Every castle, every cathedral, every bridge, and everything before then would have been planned and communicated without any way to communicate a 3 dimensional shape in 2 dimensions
yeah but perspective and shading ruined art of cats.
Does that not illustrate the point? Our methods and tech for expressing ideas have come a long way. Historically, there would have been more people (at least proportionally) who lacked the means to communicate their ideas as fully as they would have if they’d had access to the same education and resources as we do now. It’s an analogous situation that there are now technologies one can direct to (potentially) communicate an idea better than if one first had to achieve decent artistic competency to produce a similar result.
The difference is in the cost-benefit. Shading and perspective have minimal extra cost in materials to add to a drawing. Maybe your drawing takes a bit longer and uses a little more pencil lead/charcoal/ink/whatever, and the individual requires more training and practice to draw it well. That’s not a high cost to being able to communicate 3 dimensional shapes in 2d drawings
Machine learning requires a massive amount of compute upfront (basically a swarm of GPU servers at 100% for a few days or weeks, so easily multiple kilowatts per hour at the low end), it also requires a massive amount of training data (literally hundreds of terrabytes of data! This data is typically a mix of existing datasets with appropriate licensing and scraped data which hammers small community servers and creates extra cost and work to mitigate the extra unexpected traffic, etc. etc. oh also included in those training datasets is a lot of CSAM, private medical imagery, copyright infringing uploads, invasive medical and police report information, etc etc.) then still needs a large amount of compute once the model is trained to output an image (basically 100% of one GPU for 10-60 seconds, so maybe only a few watt hours per image but it adds up fast) and the resulting output is…no better than anything a human is capable of drawing or painting.
In aggregate we’re consuming hundreds of killowat hours of energy, consuming unthinkable amounts of water to attempt to remove hundreds of killowat hours of heat out of these systems. Small webpages operated by individuals are shutting down due to the cost of bots excessively crawling their sites looking for training data. Victims are being revictimized as the worst moments of their lives are sucked up into the datasets these machine learning algorithms are training on so they can remix and recreate these worst moments an infinite number of times. And to what end? All of this societal cost just to save you from drawing some stick figures and lines on paper? This cost benefit ain’t mathing out here
The only revolutionary thing about image generation is that there’s never been such a costly way to create images. Oh and maybe it’s kinda novel that humans aren’t the ones drawing it. The output isn’t anything new, just an amalgamation of existing ideas (seriously try having it generate anything that’s not already in its existing dataset in large numbers, it completely falls apart!)
In short, I’m not convinced that your AI generated images are communicating anything that you can’t communicate with a paper and pencil. And I know I and most people for that matter would much rather see a stick figure and poorly drawn lines that a human drew than whatever slop you poked out of the slop machine
i mean if you want to do a cost:benefit analysis doing it yourself wins every time
You might want to revisit your maths, there. There’s no way sitting at a computer learning Krita for months uses less electricity than a few prompts to an AI, and there’s no requirement that the AI in use has to be running remotely at one of those big, fuck-you data centres owned by Anthropic, OpenAI, or whoever.
Yeah, why use technology when decades of practice will accomplish the same thing? Everyone is just too lazy to master every skill.
/s
That is not the argument being made here. The argument is that people would rather see a bad or decent piece of human-made art to be able to look in someone’s mind, than look at amazingly skillful AI art.
Your thinking comes from a productivity and competition standpoint, which is just not what’s being talked about here. Productivity and competition is so deeply ingrained under capitalism that not caring about both of those things is hard to conceive.
It’s not talking about making a movie that succeeds on Netflix, or making a cover image to sell a product, etc etc etc, but about expressing yourself as a human to other humans.
You don’t have to be a master. Started drawing recently which you can do for less than £1 and free if you wanted to put effort into doing it for nothing.
Helps to find things which are easier to draw with minimal skill, which isn’t necessarily everything. Also repeating the same or similar thing too.
I want to come up with decent patterns to engrave into wood which does limit it a bit, can’t really do shading and the lines can’t be too close or once carved into wood they merge into each other. I think a bird standing on a branch and some Halloween ghosts are probably things I can do well enough now that they could be worth carving at some point. Sanded some more scrap wood recently which I could use, might be worth buying some boards soon if it comes out well.
