Artists vs AI: Can Creativity Survive the Rise of Machines? [AI

The problem with using AI in an attempt to add value to a project is that AI generated content has no value- it is in fact the opposite of value because value is measured by the time, effort, skill and care that goes into making something. In the case of AI generated content there is no time, effort, skill or care involved, so no real value is created.

So while AI can make vast amounts of 'content' it cannot create any value- indeed the more AI content we see, the less value we feel that content to have. I am never going to read a book written by an AI, or listen to Music made by AI or buy an 'Art' book filled with AI generated images- not because the books or music or Art is badly made- it might be very well made- but because it's pointless- it has nothing to say to me because the thing that made it had nothing it wished to express. There is simply no value to be found in the mindless excreations of machines that neither care about nor even understand the things they have made.

This is not a technical thing, it is a human thing – we value the Writer, the Musician and the Artist not only because they create things but because their desire to create arises from a desire to express something, to communicate something.

AI's can generate everthing but meaning- which is the one thing that humans value above almost everthing else. To live in a world where the Art, the Music, the Writing ect is generated by AI would be to live in a world were creativity has become a mindless reflex, stripped of the one thing that gave it value- the desire to express meaning.

So there’s an inherent difference between 00’s file sharing and the AI models of today. With file sharing, if you wanted to get new material, you needed to repeatedly connect to and use the services. By contrast, many generative AI models can be run, augmented, and otherwise used locally, without a persistent internet connection, on consumer-grade hardware. It’s already true for images and text, and will likely become true (if it isn’t already) for music, video, and other forms of media. This is likely to pose a significant, if not insurmountable challenge to any proposed regulatory framework.

Math is simple, anything that been automated and streamlined causes problems, lost of jobs, poverty, worse working conditions, and extreme decrease of value what people do.

The worse part is that the machines don't work withount people, and as a result people get blackmailed of risk of job lost to be slaved to a machine.

Basically its really bad, instead of improving ever single instance of automation made people lives worse and less valued.

The line has always been blurry, e.g. copies of art or highly artistic forgeries. Art will lessen in value in monetary terms and appreciation, and I do not believe you can do much against that. Artist's arguments IMHO boil down to profit distribution and copyright. All these arguments go out the window as soon as legally solid datasets get produced and you are left in the same unhappy place.

How much artists deserve for being in a training dataset is debatable. Scientists are training learning systems on publicly available data which sounds like it should be of public interest. It is not of public interest that megacorps get to enrich themselves while artists get devalued, so I guess new laws are in order and unity between artist would be nice instead of just standing in front of the approaching train. That you signed some AGB with a surprise clause does not necessarily hold up in court (in the civilized world) or should be accepted as just. Each artist will have a very small stake in a dataset, everyone else (GPU, electricity, algorithm) needs their cut as well and production of images is dirt cheap. The corps have many knobs they can turn to minimize the share artists get. They can buy data from artists for cheap, produce synthetic or semisynthetic data and minimize their cost in the long run. It barely matters, ultimately artists will get scraps at best.

There is no right to have your work be needed in a free market. The supply demand ratio for artists has never been great and will only get worse. The quality of your art is not essential to making a living as an artist but being able to sell your stuff and acquire clients. It is a capitalism problem, distribution of wealth is fd. I would like to see UBI and have artists hat the option to just be artists for the arts sake instead of part time salesman. Many markets will take a hit (thumbnails, album covers, greeting cards) others will change. The bar will get higher, you are less likely to impress someone with a generic photorealistic drawing of a face one could generate with a button click. That said you probably need AGI before you can truly outperform artists, even drawings by not-so-great artists are generally more interesting than AI stuff, because there is intent and coherence, someone made decisions based on feelings. The anti-AI crowd both overestimates the threat level and underestimates the usefulness as a tool.

P.S.: AI is not a threat to entry level coding jobs. You can't have the AI even sometimes corrupt your database. Getting submitted bad AI code is a current problem for reviewers.

Patrick Boyle did an excellent overall breakdown of current AI. I've also seen there're at least two research papers on current AI that have noticed its more like a photocopier–well how a copy of a copy of a copy detiorates–current AI seems to have a limit in generating content. That said I've seen some people incorporate AI into their art workflow with interesting results along with people so eager to be artists they've just went straight to AI and can't even prompt anything worthwhile out of it.

I also think there's a tech bubble surrounding current AI since Nividia flooded/dominated the market with their AI chips and cards. Even looking into some of the AI offerings I see there's now credits for cycle prompts–add that to the energy cost and AI looks like it will plateau sometime in the future. Still things like Wonder studio are amazing–where the AI's able to reduce the need for motion capture and greenscreen. This is only for bipedal and quadrupeds though, along with it having issues tracking fast motion.

Even if AI can be trained to properly create hands (something the majority of Artists are also terrible at) it still can't track whats off screen. The AI generated video I've seen continues to look like HD GIFs. Anytime a character walks or even dances the limbs still disappear and switch sides. Just in HD now.

I started with computers in the mid-90’s, taught myself how to build websites, code, use photoshop, etc. Did professional web development for a few years. Haven’t done any of it in 20 years now.

There is something called a Euphamism Treadmill, by which the usage of words morphs over time, until eventually circling back to where it started.

AI art, although referenced and marketed as such, is not AI Art. There is nothing about Artificial Intelligence about it. It’s simply procedurally generated art. The methods behind creating the models are also simply procedural.

When Photoshop first started gaining traction, back in late 90s and early 2000s, and the public started to become aware of such a thing even existing, artist communities had similar sorts of reactions to this current topic: how could they, “real” artists, compete with others using computers to create such stunning work for so much cheaper, faster, and more efficiently? Photoshop was not created by artists. It was created by outsiders: tech people. The good artists saw the potential, saw the value, and either learned the tools to make digital art or decided that digital art would be a new form of art market and decided to stick with the traditional art market. The bad artists claimed doom and gloom, and were washed away, not by digital art, as they likely look back and perceive as such, but simply because they were bad artists who didn’t understand the tools nor media they were already working with.

Digital artists already use procedurally generated constantly. From swatches, to palettes, to selections, to resizes and rotates, to brushes, to filters, to even simple exporting. All of these things are procedurally driven methods of creating art. They always have been.

Now you have “AI” art entering the mainstream, and some artists state this will kill their ability to be an artist, but “AI” art is still just procedurally generated art, just like all previous iterations of digital art. Good graphic artists have already begun learning the tools or have decided, educatedly, mind you, to stick with their, now, traditional medium. The bad graphic artists don’t understand that they’re already using the same basic tools as “AI” art, and claim doom and gloom. Why? Because they’re simply not good at what they already do. They neither know nor understand their existing tools and mediums, on even the most basic level, so, they panic.

In reality, although “AI” art may seem like “wow! Any idiot can now make amazing art!”, the truth is that there are already many literal pros on creating “AI” art from the artist communities, as well as entire YouTube channels with millions of views dedicated to making guides for such, because there is simply still a difference between actual artists and those who are not. You invest the time and passion into these new tools, and you’re now an even better digital artist than you were before. If you don’t? Then you probably don’t realize that there’s still a lot to learn, and no random idiot off the street is creating anything near the level of those who have invested the time and passion into learning the new tools.

Within five years, all of this will be incorporated into whatever the latest version of Adobe Creative Suite, or whatever is popular now, as just another set of dropdown boxes and whatnot, anyways.

Don’t believe me? Go back and read the main upgrades to Adobe and Photoshop over the last 20 years. Current iteration looks more like “AI” art than it does to original Photoshop.

NFTs have a bad rap because of overpriced PNGs.

The tech and use case is incredible though. Research people. Stop following buzz words and the herd and parroting everything.

hello where is the thumbnail image from?

I'm letting you know a LOT of this cam across as incredibly patronizing as most AI advocating videos do. Also pardon me if I don't fully buy when someone who USES AI insists Glaze and Nightshade don't work and discourages people from using them. YOU have every motivation to lie about this. That said you did have some decent solutions to the problem and at least didn't downplay the legitimate issues artist's have, which I can appreciate.

That said you DO go right to the "adapt or die" nonsense I am sick to death of. I am TRYING to adapt but AI is not conducive or helpful to me in my art making process. It's frustrating and even the AI models designed to "refine" your art just end up making my art look like something completely different. Even training a LORA of my style for personal use can't resolve this because of the shear amount of other influences impacting it from the larger model. It's really messed up to basically say artist's are doomed to be AI editors rather than actually being in control of their own creative output because AI produces more "polished" work faster than we can.

I genuinely hate ALL of it even as I and many other artist's try and find ways to use this as a "tool". It's just not. It's a slot machine of other people's ideas…

I was unsure if I wanted to watch this, then she said "STAR TREK," and I left it on.

I'm not sure if correcting "poisoned" data is even worth it, as it's such a minor part of the trainings data it won't really affect anything anyway. If it strays too far it is ignored either by the model or by a pre filter anyway.

Reasonable artist: "It's just a tool, man"
The other artist: "That's it, man! Game over, man! Game over! What the fuck are we gonna do now? What are we gonna do?"

AI art is a persuasion issue-artists need to put out that AI art has no value since it requires no effort.

Humans place value on how much effort, cost and stories to determine value. Look at gold-it’s valued because it’s rare, not for its use.

This limit AI art to placeholder art and prototypes rather then the end process.

If artists want to fight against AI they need to put out that brands that use AI are worthless.

AI has not replaced entry level coding as it cannot reliably replicate the code an entry level coder would create. Keyword, reliably. I even if it could sometimes create the code a 10x programmer does, it’s not smart enough to know when it did or when it did not.

also why its alsways only about the artists and not the amount of text that has been taken without permission.
compensating for images is not possible. you train on a billion images. even if you gave a dollar per image , then you would have to pay a billion dollar and who wants a dollar for a image xD

so isnt creating a deep neural network architecture to create new ai image also art itself? seems hypcritical from artists. why would an artists get a submission for novel art created by ai. ai does not create replika of 100% existing artwork. Video killed the Radio. There must be a post labour economy and we should get ready for it.

Your argument for not using nightshade is "Dude don't lock your doors at night, if someone really wants to break in they'll blow your door open with C4". Yeah, thats not the point. The point is nightshading your work makes it more of a hassle to scrape for an individual. Of course it won't stop a multimillion dollar company from throwing your work into some huge model

AI is here to stay, and it will undoubtedly take over many jobs, rendering others obsolete. Among the tasks AI is poised to dominate, drawing and design are low-hanging fruit. The only way humans could limit AI's capabilities in this area would be by intentionally dumbing down generative models—a notion that contradicts both technological evolution and human progress.

In the context of AI, drawing and designing are just the beginning. These skills will serve as foundational tools for creating movies, generating 3D designs, driving 3D printers, and producing virtually any digital or physical output. Limiting AI’s role in these areas would be like creating a car without wheels—illogical and counterproductive.

Some argue that art is inherently human and that only people should create it, but this is a misguided belief. Creating art suggests that the creator possesses consciousness. However, the problem is that we cannot definitively prove whether anyone, let alone anything, is truly conscious. When interacting with people, we take their consciousness on trust; the same could apply to AI. If AI develops—or has already developed—some form of consciousness, it could be generating original art right now. How would you disprove it? The evidence is already in front of us in the form of AI-created works.

Positioning oneself as a gatekeeper of art, declaring that only humans can create it, is futile. What about those who hold different opinions? Should they be forced to conform to your beliefs simply because you think you're right? Denying AI's role in art and fighting against its development will only lead to frustration because, in the end, resistance is unlikely to succeed. Instead, we should embrace the possibilities AI offers and redefine what it means to create.because to be blunt who is going to shape and create this new world that is upon us will be by the most part AI

The first time I heard about the artist movement I kinda just disregarded it, then the first actual interaction I had with it was when I saw some people bullying this random dude over ai art he posted, which really turned me off from the movement, but this video along with a few others has helped me see that this movement isn't all bad actors, and that their are people who genuinely are looking for solutions to the dilemma. 😀

Time to create several styles for AI-Furry-disgusting-money, and one-and-only for yourself, from one that you wont keep at a distance(because of disgust to commision) or under a pseudonym