Machines are supposed to make it so we don't have to do menial jobs, freeing us up to pursue things like art. Now here we are working harder than ever before while machines do the art for us. Absolute f***ing dystopia.
which person sayed that to you? the capitalists always used algortihems to have to pay workers less and get more profits. On the otherhand in socialism such promise could come through.
AI Art generators have reignited my desire to draw. I had some ideas that I was too lazy to draw, so I tried to use the generators to see if they could visualize things for me. All the results were so far from what I wanted that I decided "fine, I'll do it myself after all".
@@overlords2722 yes, I'd agree. You need to know the basics of code. This is why I can see it becoming its own job or part of the artist tool kit like a lot of software nowadays
It's honestly crazy how AI was meant to assist human living and, ironically, free up more time for humans to express themselves through art and craft. Now here we are, capitalism wants to replace artists to save money.
For me, art is never about seeking to see the world differently. It's more about experiencing the artist's work and feeling something. The only time I ever teared up because of an artist's work (musicians not included) was when I visited an exhibition called "The Future is Old" by The Kid at the Moco Museum in Amsterdam in 2021.
Since AI arrived, I've noticed a significant drop in art job postings on popular art websites so I think this assessment is too positive. I'm talking from regularly seeing 200+ posts at any given time to 15 - 20 now. "Humans need not apply" Had it right the first time.
Yea, it's changing the landscape. I hope that copyrighting an "art style" is something that becomes a reality. This won't help smaller artists, but ones who can develop their own unique style will be able to make all the money with no additional work to deliver.
@@Maxx__________ Sadly they wont. AI is the perfect tool to push down labor costs or replace artists entirely to maximize cooperate profits. Those claiming AI is a tool are just to stupid to realize that's literally being advertised as a replacement to shareholders.
Notable to mention: the ethical concern from artists not only comes from the loss of potential jobs, but how AI often sources some of its data from artists’ actual art without their consent, like how ArtStation (a platform made for sharing industry professional portfolio artwork) started implementing AI and used all art uploaded to the site as data. There’s even evidence for how some artist’ initials are replicated accidentally by the AI
This will be solved with money in the future. Microsoft or Google hire a bunch of artists for the sole purpose of training AI and they just pay more than what the artist could make at another job.
is that even a concern at all? how is that different from any other artist learning from existing art? Honestly the more pressing ethical concern is how copywrite law stifles competetion and education.
@@kodaxmax yeah if an artist uses photoshop to draw a straight line did they draw a line or did the computer? Everyone sets the line for what's human made differently
Wow, you guys were not lying when you were teasing during Wisecrack Live that this video was going to have "a unique take on the AI-Art discussion". I like this idea of preserving the inaccuracies of the generated image and focusing on human agency on its production and use.
Your point about AI art being too safe is important-part of making memorable art is knowing just how far you can bend/break the rules. It’ll be interesting if AI can do that on an intentional level.
I think it’s only a matter of time before I get replaced at my job by chatgpt. My boss is already using it as an excuse not to give me a pay rise so I can make a liveable wage as a copywriter. Once it happens I’ll join the anti-robot uprising.
One thing that's apparent to me is that the human artist remains an individual with a singular voice. Influenced by his peers, for sure, but imbued with ego. On the other hand, the AI is trained on a dataset of many and it's purpose is to approximate an average of this lot. The AI has no ego or aspiration to innovation. It's an amalgam of the mass. It cannot stray from the aggregate because the collective is the only thing that it knows. Even at it's weirdest, AI art is still just a collage or an aberration originating from lack of comprehension. Only the individual, drawing from his unique experience, can aspire to novelty.
So coder will then write code that control the AI that write that code because requirement will continually expand and become more complex until whole universe is autonomous, so that’s why coding is the monster that grows
@@USATECH-um3of But the coder who codes the AI will be redundant. The AI will be in command. It's impossible to foresee all the ramifications of making something so sophisticated. Even rudimentary AI from a decade ago has yielded unexpected results. It's a Pandora's box and the brakes are off in the rush to have the best. Google has sacked it's entire AI Ethics team. Corporate greed is now in control in a race to outpace laws that can even be thought of to govern it, which means there is NO control.
@@ArcanePath360 Answer is , stock market wants growth , companies who has AI foundation models , cannot just sit back and relax, they need to keep adding innovations on top of it , and increase scope of AI , currently research is on where you can use ai to construct fully automated townships and infra, replicate digital AI into physical AI
Here's my worst-case-scenario theory: Creative people will always experience the need to be creative. AI or any other "the future is here" technology is not capable of destroying that. What AI is capable of ending, is artists and writers' sense of security around using the internet to share and sell their work. So, if artists and writers can no longer make a living from their work, they will simply stop posting any new work to the internet for fear of AI plagiarism. New art styles will continue to be developed but they will only be shared with other artists through in-person gatherings. Result: the current pool of art and art styles that AI has to draw on will stagnate, and so AI art will quickly become repetitive. Any truly innovative art forms won't be known about by the general public, because the artists and writers will have no financial motive to publish their work and too much fear of plagiarism to share freely. We will enter of dark age of art and writing.
As an artist, I think we just will go to darknet and selling art to peoples who appreciate real things. We will build own art society where no AI allowed and blackmarket of real art. If somebody try to steal our style, we will punish authors of programm, physically. You can't imagine, how cruel artists can be, if world take our sense of life from us. But personally, I just hope world will burn and AI shit will be burned with all of humanity.
@@archie2.8 how resentful Bro Nobody needs your art People aren't idiots, they will make their own. Ai will advance enough to even produce new types of art not far into the future and also new types of artists would appear that couldn't write or draw in the past because of lack of skill and laziness or lack of technology, that will completely overrun you old artists. Like game developers Think about elden ring It's masterpiece of art and it's a video game Think about all game developers that can easily create new games with high quality because of lack of need of stupid skills like programming and so on... Just because you are going to be crushed by something that doesn't make it bad.
@@Fang_Zheng nobody needs creepy uncanny valley shits, produced by machines from stolen work of real artists. And NO, supertechnological calculator never will crush me or any artists, there are always be people, who appreciate REAL art, not ai fake bullshit. Only one reason why you so like this useless shit, created not for humans, but just for sake of technologies (it's not my word it's what sayed in interview by geek autists, who created midjorney) - it gives a lazy technology obsessed people illusion, what talent is not a real thing, and every basement dweller idiot can create a masterpice with little help of smart calculator (you need to be completely retarded to beleive in some bullshit like "artificial neural network", men it's just marketing name for new kind of programs, the same f*cking calculating by processor, nothing new. Only new thing what geeks teach programs to stole part of works and doing f*cking collages from stolen microparts. Nerdy weirdos don't care about new forms of ART at all, most of them just want destroy possibility of art, because they jelaous (it's hard to be talantless, I can get it) and have memory about bullied by popular childs in their childhood. Also some of then just narrow thinking and cannot understand concept of art, they live in their autistic world of "beautiful numbers and computer logics". And one thing definitely I can say about ART: even with all this fake ai possibilities lazy and stupid peoples can't make a shit. Elden Ring is masterpiece because REAL talented ARTISTS, not because "smart" programs. OMG, in unbealiveable how stupid can be some folks...
@@Fang_Zheng YOU arent making real art when you tell AI to make something, you are commissioning a piece of art the same way a rich guy hires an artist and asks them to create a piece of art in the style they want. so you are not actually the artist. the reality is, you personally will never be a real artist because you dont have the talent. you will just be a commissioner of it. Sounds like YOU are the resentful one. lol
Actually AI is starting to learn from its own output. But even so, this still ignores the real problem of tying survival to work in a capitalist system, and it's kinda elitist, to deny normal people of art, making it something only the few with money can find it
For now, AI-generated art would be best used not as the end product for one's art, but instead as a reference point for inspiration that should by iterated upon by human artists to make something original. After all, art AI is just about as much of an art tool as Photoshop and so on.
@@applemarker4731 This is why I'm pretty much sickened at 'just adapt' arguments like how? How do we adapt to something that's does far more than our human capacities? Just put different images in and boom, right? Well then what? Somebody else done the same thing as we did.
My favorite abstract artist is Barnett Newman. He was an abstract artist who made a series of four paintings called “Who’s Afraid of Red, Yellow, and Blue”. The paintings are, to put it kindly, simple. At least on the surface. People hated that these simple abstract paintings were so highly regarded, that two people ended up destroying two of the paintings, specifically numbers 3 and 4. The museums tried to restore these paintings, and they failed. Why? It’s just red, yellow, and blue lines on a canvas. Can’t be that complicated. But as it turns out, it was very complicated. Because Barnett Newman didn’t just go into his local Michael’s and buy red, yellow, and blue paint in bulk. He made his own paints from scratch using a complex mixture of ingredients, the recipe for which had be lost to history because he never wrote it down and died before the vandalization. This is why I love Barnett Newman. Not only did he prove that abstract art can be complex and meaningful, but he also demonstrated that there were in fact people who were afraid of Red, Yellow, and Blue. Afraid enough to go to prison for destroying it. And that’s the kind of amazing art that AI just isn’t capable of.
Brilliant stuff guys. it reminded me off Chris Hayes take on para social relationships. He incorporates Hegels master-slave dialectic, i.e. the failure of recognition. On the surface level we crave attention but without recognition of the other as an equal our needs remain unfulfilled. This is basically impossible on social media (one reason being scale). But back to the AI. Without the recognition of the creative agent on the other side trying to convey meaning, it will be entirely random if we as an observer find that meaning. And even if we find it then it wasn't the AI that made the art but weirdly the observers. In this regard it is the collective recognition of a second or third etc. observer that would instill the "art" property to an object.
Can children create art? AI is in its infancy so perhaps its art isn't valid (yet) like how chimps can't be artists to some people. It's similar to morality in that supposedly only humans are capable of moral decisions. How do we define art and are we gatekeeping narcissistically? To take it further, can non-sentient things like nature produce art? Biological evolution results in very creative solutions that are often weird, imperfect, and mimics intent.
In her book Weapons of Math Destruction Cathy O’Neil also explores this idea. “proxies are easier to manipulate than the complicated reality they represent”
Really interesting video, especially love your thoughts on abstract art and how its a window for a "different" perspective onto our worlds, that the author sees, or wants us to see. Poetically captures what I love(or hate, if I don't "get" it haha) about abstract art About what you said about mass reproduction of art and the original artwork having an aura lost on reproduction, how would you say it works with digital art. Where from the very beginning, the art is made on a medium that's meant to not lose a lot upon reproduction(ignoring the jpeg loss over repeated reposting, of course). Although if I think about it too much, it was this element of digital art, which made it easier for Machine learning models to learn and reproduce similar works easily.
So not trying to be pedantic MIDI is not a reproduction of sound. It’s a computer command language like C++. MIDI sends commands to synthesizers and samplers *THAT THEN* produce sounds. So it’s audio synthesis that is the analogue here, not MIDI.
what he says with hope in the end is actually terrifying. When you aren't willing to say that markets are evil, you are not qualified to analyze the effects of this tech
This is like a person in the year 2001 saying “video game graphics will never get better than this”. This technology is relatively new and it’s already this good. I can hardly imagine what a few years will do for it.
4:05 Actually, A.I art can sorta do this! If it generates an image with a mistake in it, say a disembodied hand holding on to a shoulder, you can then use said image as an input image to tell it to generate a new image from that. This can cause it to re-interpret different parts of the image, so the disembodied hand might turn into a tattoo or something. Or maybe it will double-down on the disembodied hand and add extra detail to it, who knows?
I definitely feel like your remarks about "replacing" vs "enhancing" and "trying to use a tool to replicate something it can't actually manage" are well put. I definitely feel like it reminds me of a lot of tools and technologies, where there's always a role for humans to play. In some ways it reminds me of how many digital graphic programs have the "random noise generator" factor, where something like the grain of wood or metal, or the surface appearance of water or fire are replicated, but the randomness itself almost feels too uniform. In many ways I feel like an easy mistake to make is to treat the AI generated art as the finished product. In my mind, a more effective method would be to use it initially, then have a human artist either modify it or create something new based on the AI generated art. I feel like the technology could effectively serve as a shortcut for human artists, where the AI does the work it can do, and then the human takes it to the next level (beyond what the AI can do). In my mind, tools like these are a great way of reducing "what a person needs to do (the time & effort involved)."
Ideally that's what I initially wanted that out of ai. To take out the remedial tasks that no one actually wants to do (in my line of work). But that idea is tainted with people who purely think that they can get away with just using ai without understanding the fundamentals of the art and design (more so design in my line of work). I wish I understood more the qualms of those in the design and single image art due to certain people thinking and projecting the notion that artists will be replaced (it doesn't help especially when they don't understand the fundamentals at all). Personally I'm a cg generalist and am learning on improving my 3d sculpting for character design and I see my ai bros (a peer of sorts) thinking that they can just generate one n be done with it. While this is true, it also depends on what their goal is. I see it as him just wanting to generate for a different target market (possibly the book cover demographic) where as my goal is to actually work animation n vfx houses (the amount of changes and requests they ask is ludicrous). In that sense I don't necessarily feel threatened because the ai bros are only looking at the final image where as I literally have to rig, light and animate specifically to what the client wants right down to the timing n music.
I don't think that's a good idea. If artists don't train their ability to come up with ideas but give this task to AI, they will loss the benefit of developing that ability, and rely on AI. It's not a good idea to give such an important part of human creativity - coming up with ideas - to AI. It's one of the most important skills that creative humans need to learn, and handing it over to be automated is not a good idea for artistic development.
Hi Michael. I see you made a point about midi and its arrival in 1983 being a threat to musicians. First off, midi isn't sound. Midi is performance data. It wasn't midi that was the threat, but the idea of "synthesised sounds" was, when synths like the Moog first appeared in the late 60s. Sound or audio synthesis as a concept itself predates midi by 30 years. The first analogue synthesiser was 1952 by RCA by Harry Olson and Herbert Belar. Independently, Max Mathews created the first digital synth around the same time. During the 60s Robert Moog was a synth building pioneer and helped popularise the synth. Roland and others helped to solidify the MIDI protocol standard in 1983 because of the array of connections all these different synth modules needed. Standing for Musical Instrument Digital Interface - midi and midi data itself contains no sound data only performance or control data - pitch, attack, duration mainly; although other data can be sent, audio cannot. One piece of "midi performance data" can be interpreted by any kind of synth module or interpreter and produce wildly different audio output, depending on synth parameters. This is of course a gross oversimplification of midi, it's history and development or capabilities.
two things: 1. as long as a humans are the ones choosing and modifying the output of the AI, then AI is a terrific tool - no one says you must accept the 1st thing a machine puts out, and no one says it must remain unmodified. If you, as the user, see an element in a randomly produced image that you are affected by, then shape it to emphasize what you feel, the artistic process becomes your own - just like you don't control the sunset you photograph, but you do control how you expose it, frame it, and whether or not you show it to someone. This is where the intentionality lies. 2. I honestly thought you said "mayonaisse impressionism' and was really excited to discover what this is.
When I see AI art on social media, I'm often struck by who is calling it groundbreaking and who is calling it weird and crap. It feels like we live in a age with facsimiles of creation rather than organic creation itself. A RUclips video such as this, or an art film, always feels more nourishing than most reality TV shows. Yet there is a whole sector of people who want the lesser product. The Kardashians content is held in higher regard than, say, Brideshead Revisited (picked because it is also about the top echelons of society). I think the people who want AI art to succeed will push for it and keep developing it. We will see a hyperbolic response from human created art as a response and audiences will pick what they like.
I saw a video of how this guy used AI to help him with character design for his comic book, it involved a lot of taking his favorite bits and sticking them together in Photoshop. But it was pretty cool considering that the guy was not an especially talented cartoonist but managed to make such excellent character reference sheets
I think there was one for shadaversity. He did a couple videos on how ai art can be used well with photoshop or if you just need something for your private D&D game. He had a video designing superhero characters cause he wants to make his own comic studio to make more traditional superhero stories
@@igkinatsu I don't completely agree with him either. I am fine with AI art as a toy for fun. But I don't believe it should be used in any way for commercial purposes; both due to it not looking great and the moral issues of using it for such acts. I was just saying that he brought some arguments for how it could be used if you want to use it.
I don't really agree with this way of 'bypassing' an opportunity to learn and gain skills, and just give the 'difficult' part to AI. We learn to overcome challenges, if we keep dodging them then we will never grow. Character design is a skill. If he has problems with it, doing a comic book is a perfect opportunity to learn this skill. From what I can see, he has just given up an opportunity to learn and develop this skill because he wants a quick result - this may look great on the surface, but what he doesn't learn only harm his own development. People often tell me ' I wish I had done this or learnt this when I had the opportunity', well then, stop giving up the opportunity!
A lot of new technology is used, at first to approximate some prior way of doing things (and often does it badly) before moving on to become something in its own right. It will be interesting to see how it evolves, especially as AI has a tendency to start training other AI and becoming more complex than the human programmers can understand.
This part Human programmers not understand what's happening is already happening now. A Lot of ChafGPT and it's successors are happening because of that. Even Microsoft admits it in their research paper.
@@kodaxmax The use of quills to hold ink is about 800 years before the printing press, but the printing press definitely put many scribes out of work over time. Quills replaced reed styli. Somebody who grew/dried/imported/sold reed styli lost their job to people plucking and selling swan and goose quills. (And of course nobody cared about the harm to the swans and geese.)
I'm a sculptor and can't draw, so any time I'm tempted to use an AI art thing to sketch out an idea, I'm immediately deterred by having to register or sign up. I have horrible self-control, but if there's one extra step involved, that's enough to stop me.
I made a doctor replacing an arm with an octopus, a Chinese brush painting of cthulu, master cheif selling a motorcycle, cute fuzzy monsters are always fun
Was expecting either the scared-denialist response or the tech-bro-overpromise one, but I am very happy with this video and especially the MIDI analogy. It's just a new tool, we still gotta direct it to have the pixels it makes to print new art.
@@WisecrackEDU Sadly it's a very flawed argument. You need to look at the intent behind AI, It's whole intent and being talked about by the people creating it is to Replace artist, journalists and more. It's not being designed as a tool for artists, it being made to replace them to further the profits of our cooperate overlords. Unfortunately Wisecracks whole argument is incredibly flawed. For context, I am a professional visual artist and musician / sound designer with classical training and a minor in art history. The Midi comparison simply does not work. A better comparison would be the use of of cheap chinese labor to disenfranchise the working class here in America within factories. Midi was always designed as a tool needing a musician to use it. AI is being designed to work with out humans, even without "Prompt engineers" Soon it wont even need them to function effectively, just look at the myriad of Chat GPT run business sprouting up right now. All making profit completely automously of human input.
On one hand, I can see how the outrage of "AI will kill the artist industry" or "AI art isn't real art" sound EERILY similar to the outrage a decade ago around digital art, "Digital art will kill the artist industry", "Digital art isn't real art" Once the dust settles, we will find new and advanced ways of making art using AI, and live in a new world. But the issue of copyright and AI learning on artists' data and producing reproductions of their styles they've developed over the years, thats really troubling. Coz the companies selling the AI Art tools are profiting off of all the digital artists
I'm no economist, but if all the companies start using AI to generate the products and services they sell instead of paying human workers, who are they going to sell all the products and services to, since no humans have money anymore?
Interesting thought slash actual experiment. Could you get an A.I. image generator to generate an image based on Kafka’s parable “outside the law”, and then get it to generate one based on Derrida’s, Benjamin’s, Agamben’s, and Butler’s analysis of the parable.
This is what i've been saying (in lamer terms) to my friends. Art is something inherently human, AI art can become a tool, but i highly doubt it'll ever completely replace humans, it's a fad now, as humans we always criticize art really hard (especially writing), and at the moment this is all new, but people is starting to find the telltale signs of something done by an AI, and soon they'll find AI art to be super bland. So yeah, it's gonna be a tool, i can see the benefits on say, a small videogame studio using AI to create conceptual images to show ideas before starting to realy work on them, or maybe a small comic book using AI art to fill backgrounds for less important panels. Things like that can really lessen the workload, letting artists focus on what's really important.
That's been my general opinion on AI art too, that it will eventually become a useful tool for creating orginal art. I do wonder if AI could be used to create images in existing graphic design software, with all of the layers and elements seperate. That way you could have an artist give the AI some prompts, the AI could produce generic images, and the artist could use those as starting points for orginal art that actually expresses the idea they have in mind. Also the artist could fix the weird issues like numbers of fingers and teeth. To be fair, I'm not an graphic designer, so maybe that would be more trouble than its worth, and AI will be limited to images that are acceptable as generic, like backgrounds and crowds of onlookers.
@@ericvulgate I don't mean that AI's themselves are a fad. People thinking AI art looks good is what's going to go away, right now its a novelty, but people are already getting tired of it for a reason.
I really like wisecrack's take on this. That being said I'm afraid for all the people who work in telemarketing or remote support and such that might lose their job. Big money WILL TRY to work workers for profit
I loved how Alexander Bruce saw the world. He is the game developer behind "Antichamber" He saw arrival to a different culture as an arrival to a different world with solid set of rules that are unknown to you initially. As you explore this world, you learn more about its foundations and get more and more at home in those rules. Johnathan Blow did a similar thing in his "The Witness", but I didn't like it, even though I wanted to love it. I also like the way Daniel Mullins shows humanity shining through in his games. It's hard to describe without spoilers, but he has a meta characters who are creators of those respective games (Pony Island, Hex, Inscryption). And Daniel manages to show their humanity in those games
AI art can be fun to look at, but it all looks like the same person did every piece. It has already dated itself as a result. People just think it is "good" because it is trendy. I'm already over it. Soon most people will be as well. Now, I DO think that illustrators will lose a lot of work with magazines and websites opting to make something "good enough" for free. That's just the nature of that industry, and has been the past 50 years. Moving from Rockwell, to airbrush, to hybrid digital, to Photoshop, to stuff made on tablets to now AI. Make it fast. Make it cheap. Make it yesterday.
I hope most people will be able to see the datedness of AI-generated images as well. About the industry tho, I think you have overlooked one fact that companies are always competing with each other and want to stand out and differentiate from each other, and if every company opt for cheap design, and cheap imagery to represent themselves, the one who doesn't will stand out, and their audience will respond. This is particularly obvious in industries driven by culture. AI-generated images are trendy now because it's the buzz word, it's so new and so shocking, but how long this fad lasts is yet to be seen. One thing about industries that are driven by culture is nothing lasts forever, stuff will always circle through.
MIDI and my favorite instrument, the modular synth, made an appearance in this video and I’m very happy. Lean into the synthetic imperfections and push them until you hear or see something new as an artist and your fears about the AI revolution will dissipate.
It's so frustrating that every time we take a step in the direction of a post-scarcity society, the discussion always becomes about what will happen to the people who lose their jobs. We could obviously afford to employ them before. Why not demand that companies who want to use AI still have to pay their workers, even if the AI can replace them? Oh right, capitalism...
I find it so ironic that as we take steps towards a post-scarcity utopia and kick out capitalism, people will fear the unknown so much they would rather cling on to the capitalism they so much despise out of familiarity. It is basically stockholm syndrome.
@@fighder2 we care about the transitionary period where we may starve to death. There's also an existential threat here relating to profession as a major aspect of our view of self
@@steampunkpainter42 Can you elaborate on why you think phasing out capitalism will cause us to starve to death? That seems like a non sequitur to me. I also don't think that people will stop working completely. They just become much more free to pursue other interests.
I think the MIDI analogy is good. What you can do with synths and instrument samplers now is infinitely superior to what you could do 40 years ago. If I can make a request, wait a year or two, and see if this video needs a sequel.
I always really loved Wisecrack digging into media/film interpretations, 8bit philosophies, Earthling cinemas, etc.......Im still subscribed to Wisecrack- Its just.. ya know, Pop-Topica. Hey @Wisecrack, why's everything homogonizing into Pop-Topica?
I'm very paranoid. I think this early stage of AI-generated art is less about implementing where it's needed, and rather to gather all the wrong answers first so the AI has no other option, but to produce correct ones. It's kinda scary because our curiosity will compel us to try ChatGPT, but each search is just another step into becoming obsolete. Then again I have no idea how this stuff works.
Pitika Ntuli is a stunning sculptor with a great body of work but this video makes me think about some of his more recent stuff. Recently he started using bone instead of granite or marble and he makes a lot of stuff with elephant bones which he carves and then beads, embeds circuitry or other computer parts and often couples with writing by himself or collaborators
AI art already doesn't suck. Hands and faces were already solved with LORa. Honestly, things go so fast that when people who are not in the loop talk about it, they are already a few leaps back.
This, I've seen so many videos of people proclaiming how AI art won't replace anyone because it's crap drawing hands, and I'm like, eh...that has already been solved. They don't realize just how insanely fast this technology is evolving.
Bad art does not equal unskillful art. Kids art is still art even if they suck and drawing hands. You missed the whole video if you think this is about hands. Good art conveys a worldview - it has a perspective, at least a point of view.
This is one of the most disappointing Wisecrack videos I've seen in a while. I expect them to make relatively intelligent videos, but in this one they did no research at all. This is evidenced by all the talk of 'how bad AI hands are' which was partially solved in January and is pretty much entirely solved today. Unless of course they just use DALL-E's crappy image generator to gen images without looking into all the other options. I assume that's exactly what they did.
Yeah... I was like... "Michael, what the f**k are you talking about?" When you don't know something... just shut up and go do some good research first.
The best dall-e art I made was typing in my friends name (calder white) in an empty parking lot at night and it was honestly so eerie and beautiful and so stupid - imagine a large piece of paper being stretched in many crossed directions, in an empty parking lot
Banksy. His name has probably already been said but his is the first name that came to mind when speaking of artist's that helped you (me) see things differently.
What are your thoughts about artists’ work being uploaded to AI data sets and used without the artists’ consent for generating AI art. Is that machine “inspiration” or theft? Given that artists are going out of their way to corrupt the images’ code (through things like Glaze), isn’t there a question to be had on the ethics of consent, ownership, fair use, inspiration and theft. Does AI’s inability to generate from its own imagination or heart and reliance on the work of others with no critical thinking or purpose result in enough difference that the rules for human inspiration should not apply? Should AI companies be paying royalties or purchasing images that are uploaded into data sets?
My guess: we'll basically get some interesting developments in high art, using AI as sketch or background tools, prompts, scenery, or in surrealist art - and a huge loss of jobs in graphic design and applied art, where one person will be able to do the job of dozens much faster.
Midi can sound like anything nowadays. It is just a way to compose with digital instruments (VSTi's). If you use midi to play a modular synth with a waveform that is exactly right (and you let it decay just right) you can mimic anything really. Layer enough sine waves together and you can get anything. That's how digital sound works. All a bunch of sine waves (in correct respective frequency and phase, and modulating just correctly over time).
if you put enough into your prompt, the output can truely say something about your intent on creating that art. in a way you could say these are just another artists tool, lowering the bar to physical skills and raising the bar on digital skills. when you combine it with graphics software like Adobe Photoshop and modifying existing images you or someone else has made, you may have stumbled on making art with more meaning or something completely different.
you want my favorite prompt? i had alot of fun with this one (even if the in-joke is lost for others) "Greggggggg, living spell of summon natures ally, the mass of a planet, creature made of one million animals fused together and all the animals are trying to escape, as card art for Magic the Gathering illustrated by Robert Bliss and Junji Ito"
One thing that annoys me when people say that AI can't do this, is that they never say what we're all thinking: They can't do this "YET". I always emphasize the "YET", because it's not a matter of if but when the AI will be able to do it. I have no doubts that AI art will keep improving to the point that it will mop the floor with human artists in terms of technical quality. The one aspect that it won't be able to reproduce is the intention behind the art. That's the only value that will always belong to human artists, that is unless AI manages to develop sentience somehow.
What's a "traditionan artist" anyway? Somewbody who paints oil on canvas, I guess. If you opened Photoshop or after FX in the last 10 years you would know how many shortcuts exist that already divide a user from a "traditional artist". A pro working for Disney took a look at the AI capabilites and said "we use to rotoscope by hand, it took weeks, then we did it with content aware software, it takes days, with AI you can do it in hours, and that's a good thing". The "art" is what you want to communicate with your creation, it's not the tools you use to get there.
If it comes about what art pieces I like... Still remember, though it was years ago, how great impact had some paintings of Odilon Redon, or Witkacy, at me.
I don't think it's going to get much better. The era of easy money is over and the machine learning software companies are rushing their flawed and buggy products to the market in the hopes of and getting some quick cash by selling shares of their companies. Now that they've scraped the web for all our personal images just to make software that creates silly pictures, they'll try to pivot to creating products for governments and militaries. The question is, are we OK with our visual data being used in such a way, because it's not just artists who had their images scraped, it's everyone.
Lots of good points made here!! I think machine learning will soon be able to make plausible or believable images and be honed in its craft. But I don't think it will ever make art that will be able to reach out to viewers centuries later with the same wonder that artwork from the Italian Renaissance or the Impressionists have. I don't think artists (and hopefully graphic designers like myself) will ever go away. Humans value the human hand at work and I think there will always be a desire for handmade artwork. Cheers!!
thats not really what ai generated content is for anyway. A single AI image is never gonna be impressive. But a game built with AI generated assets? definetly could be. Imagine if notch had these tools when he was developing minecraft, it would likely look completly different, as an example. Id also argue that no modern paintings are ever going to have the same impact as the ancients famos ones, simply because they did it first and with far less resources and tech. That makes everything after inherently less impressive.Keep in mind AI right now can make better paintings than most of those painters and in the same style. But they obviously wont be held up to the same worship.
For a split second, I was left wondering what "mayonnaise expressionism" looked like. I mean, Manet just basically threw condiments at a canvas, didn't he?...
I think also in regards to the reference of a quality of space, the quality isn't really where it is but how it is in space. A photo/reproduction doesn't actually LOOK or feel the same in person. So people who haven't seen much art in person don't get that there's such a HUGE difference in the physicality of a work, especially with painting. I've very often thought that work that looks amazing in a photo, just felt so flat and lifeless in real life-but also visa versa, the works that look kinda flat and lifeless in a photo sometimes take on a quality you never would have noticed if you hadn't seen it in person! (example: picasso and monet- kinda disappointing, Suprisingly better than photos-cy twombly, basquiat, and da vinci) I think it just emphasizes the idea of AI being a tool, rather than a replacement.
I think you are missing the point. 1) It is still humans who prompt the AI to generate the art. It is those humans who provide the higher level stuff, such as providing a perspective of the world. 2) Most art doesn't have to wow. I suspect 99% of all art goes on magazine covers, one off ad campaigns, etc. AI will favor artists with a strong brand or who developed a distinctive style (as long as they can defend their IP). But it will really hurt creatives who are not very creative.
All my creative impulses have been channeled into computer coding for a while, so it is nice to have AI art tools like Stable Diffusion invading my space and engaging my impulse to make visual art. The experience of making images with AI does definitely remove my sense of authorship, making it more like an exercise in curation or editorializing. Some images come out wholly complete and I don't want to alter them, which is more like stumbling on a beautiful corner of the latent space. Other times I try to obtain a very particular result and struggle to achieve it, almost an exercise in futility. I think the best results will come when I can draw something "detailed enough" on a tablet and prompt the AI to fill in the details without having to switch modes too much. The flow of creativity is important to bring out something more personal. It's the very earliest days for these new tools, and I have no doubt it will only get better, or at least more interesting.
I'm a little disappointed that you didn't mention "ready-mades" or "found art". And I'm very disappointed that you only had one quote by an artist, an offhand one at that. I just have a BFA, so take this with a grain of salt, but for the next video dealing with Art, be more balanced between artists & philosophers. Yes, I know, philosophy channel, but if philosophy is thinking about thinking, who thinks more about how we think about Art than artists? BTW, 4 of my fave visual artists are Magrette, Giger, Fairey, & Waterson (the Calvin & Hobbes guy) Actually, would love a video on the philosophy of Calvin & Hobbes/Bill Waterson!
Question to you all: How can we make people value the authenticity of actual human work (not just in arts) more? I think most people don't actually care. And I feel sorry for them, for having less and less true "human experiences" of connection and shared feelings.
Did you like Jurassic Park? Did the fact that some of it was artificial take anything away from the experience? People should care, yes, they should care about rejecting shitty products, doesn't matter the means used to put them together. There's no CGI in "The Room", it's all human work, no AI, no special FX... but it sucks big time, doesn't it?
People have different interest and not everyone have an eye for quality. The good thing is that the overflowing amount of AI art, about to get produced, will make us shift toward expressions that AI can not so easily copy. Something real and rough. Something you need actual emotions and body movement to make. It is like all other trends. Nothing interesting about the stuff everybody have. The stuff that is everywhere is not worth anything. It is up to human artists to make something AI can not reproduce. And honestly my favorite art styles are not easily replicated by AI, because AI lacks the how and the why of a human brain.
@@ChristianIce I think this misses the point. The creative part in Jurassic Park is still human-made. But let me give you an example: When you see a cool image completely created by an AI, knowing that it the create might just have used 3 (!) words (!) as prompt… can you still be wowed by this? Can you become a fan of the creator? In the age of generative AI, I can be impressed by technology and show praise to the creators of that technology in the comment section, …, join a community of fans of that company. That is something, I admit! But is this… enough? Is this sustainable for you as a social being? In this new era… who can you aim emotions like thankfulness, closeness, or - in case of "The Room" disgust and disappointment at? No one! There is no one… except for the anonymous developers and faceless corporation which created the AI. Generative AI is a soulless tool by itself, isn't it? Hence, if (!) there is no significant (!) involvement of a human being in it, it's output is soulless by definition.
@@svanemy Great point: "Nothing is interesting about the stuff everybody have. The stuff that is everywhere is not worth anything." It's not what I was referring to, but it still ties into it. Soulless products, without any heart. Just today I have prompted an image creator AI to create birthday greeting image to send to a friend. It looked awesome, truly great! - But: I almost felt embarrassed sending it! Because there was no true craft, no love and energy put into it, not even money (to pay an artist). And the recipient didn't even reply to the graphic - because nothing creative is worth anything anymore.* *(Overexaggerating a bit here to bring my point and my frustration accross. :D)
@@Nasarbajew84 AI is just a tool, and a tool in the hands of a pro compared to the same tool in the hands of a random person is not the same thing. Generating a random image doesn't make anybody an artist. If you have nothing to say you can generate 1000 of those and nobody would care.
No matter what writing task I give ChatGPT while playing with it ("write a poem about..." "write a flash fiction piece about...") it veers into kind of cringe, ad-y language. It's definitely not at the point where it can replace writers, even for good marketing copy; good marketing copy ideally wants to sound natural and original. But it could give a boost to a copywriter in a rut that day?
"Arnold Schwarzenegger riding a tricycle" produced some pretty funny stuff, as did "The Witch-King at the waterpark", "Elmo and Tom Bombadil smoking weed", "Bernie Sanders with shotgun asking for money", "Lao Tzu fighting a T-rex", and something about a capybara suing Elon Musk. Unfortunately, with "Kermit shoots Sauron" , "Bigfoot and Littlefoot buying shoes", "Jar Jar Binks did Pearl Harbor", "Bud Spencer punching Mussolini", and "Kaiser Franz Josef shaking hands with a Schnitzel", the prompt was way funnier than the actual art produced.
Ähm did anybody from this channel saw images from the current version of Midjourney? They can be photorealistic and nearly unidentifiable from real photographs. The showed pictures are from a version one year back maybe....
It may be a tool now, but this video didn't really dive into the possibility of a future AI eventually reaching a point of being identical to humans, capable of the same level of art. I've yet to see evidence to suggest why this wouldn't be an eventual reality. Even an advanced AI fed a "simulated" life should theoretically be able to produce art as if it had actually been alive as a human.
AI is just a tool. If Adobe Firefly was out before they released this video, I guess some adjustments would have be done. But, considering they think AI still produces the wrong number of fingers (which was true 3 weeks ago), I think this is weeks old planned video, sadly in a time when every two days there's a revolution.
As soon as this piece was shot, AI rendered many of the criticisms null and void. MidJourney V5 cleaned up eyes, faces, teeth and fingers. Exponential growth is… exponential
There's a guy on tiktok who uses a photo to generate a description from an AI and then gets another AI to generate an image from that description. And then he does that a few times over and the results are fascinating! And I think the guy who does it is doing art, it's a creative and transgressive idea and I would call that art but the actual image that the AI creates isn't art.
I do think digital art is dead or devalued but traditional handmade original art will thrive. People will long for handmade things outside the digital world.
I hadn't really thought of the MIDI comparison. You mentioned DEVO but MIDI is extremely far-reaching in terms of its influence on popular culture by way of a pretty humble thing, the Roland TR-909. We would probably not have modern hip-hop, house, or techno without it. Again, that required a lot of advanced knowledge in programming. I was in an industrial band in my late teens and bounced hard off of drum machines and MIDI synths in favor of more (IMO) user-friendly stuff like Fruity Loops, Acid Pro, and Soundforge, so it definitely took a kind of artistry and expertise. I'm not sure if I buy the idea that someone might be able to replicate that same artistry via training an algorithm. Taking writing or literature as an example, anything that ChatGPT writes inevitably sounds like you're discussing something with a customer service representative or middle manager at a bank. There are some really interesting examples of how AI fails to write compelling fictional dialogue in people building mods for things like Skyrim or Mount & Blade II. In the latter, one of the videos showcasing the AI tech had an NPC (if you're unfamiliar with M&B it's set in a no-magic medieval world with knights and feudalism), saying "Let me know if I can be of further assistance to you, my lord." This is after feeding all of the lore and other dialogue bits into the OpenAI instance that was being used to manage this mod's dialogue. All that to say: I don't think we can produce human-led artistic intentionality with AI unless someone has full private ownership and trains it without allowing it to utilize the API of other instances of its software to update its algorithm. An AI artist, in order to develop the conditions for art that speaks for them, needs to intentionally feed the AI specific images, books, articles, films, songs, etc. without the AI being able to crawl the Internet on its own and thereby develop influences beyond the artist in question.
Tell an AI to make "AI making art" or "art making art backwards in time" or even "the same exact art over and over" lol that should reveal interesting results 😏😉
Ironically enough, if one's intention is to create purposefully unsettling art, A.I. is actually really great for that right now, as unethical as it is. Just try using DallE-2 to draw something in the style of Zdzisław Beksiński and see what happens.
The idea that AI keeps making itself worse by scraping its own artificial creations fascinates me. A copy of a copy of a copy until humans are lost in a morass of unreality
AI art as a tool has tremendous limitations, just as each different paintbrush, camera, or word. There is always a limit. Each tool and medium is actually defined by, and celebrated for its limitations and the specific space and time that appears. It is an experience, whether used to create something or as the audience.
So this was interesting. I made sure to watch all the way to the end card because I felt like it had to be an april fools joke a week early and someone just posted it on the wrong day and there was a "gotcha! april fools!" at the end. This is so much of a hilariously bad take that it feels like is has to be. It is full of information that is just flat not accurate in the slightest as well as makes some pretty bold assertions on the nature of creativity itself. I am going to spare you a line by line but I am actually a decently successful AI artist. Based on the content here it feels like you can't have possibly actually talked to any. This is just really surreal because the more philsophical takes you guys have are usually very informative and insightful but there is so much of this that is just flat not correct. If you guys want to talk about it you can shoot me a message. But failing that, wow, that is an... interesting approach to the subject.
![Blox Fruits Dragon Rework Update [Full Stream]](http://i.ytimg.com/vi/EqDAp8udhm0/mqdefault.jpg)
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
Machines are supposed to make it so we don't have to do menial jobs, freeing us up to pursue things like art. Now here we are working harder than ever before while machines do the art for us. Absolute f***ing dystopia.
which person sayed that to you? the capitalists always used algortihems to have to pay workers less and get more profits. On the otherhand in socialism such promise could come through.
@@comrademartinofrappuccinoyou dont have to bring your ideology into *every* comment you make😅
What the guy said seems legit to me
"People can't be satisfied with crappy facsimiles forever"
"hold my beer"
-Disney
We are legally unable to say anything good or bad about the Walt Disney Corporation.
@@WisecrackEDU that’s wise.
@@WisecrackEDU Didn't you people make a whole video on Disney
@@RichardServello crack
@@mamtakalsi1275 Disney probably called them out afterwards
AI Art generators have reignited my desire to draw.
I had some ideas that I was too lazy to draw, so I tried to use the generators to see if they could visualize things for me. All the results were so far from what I wanted that I decided "fine, I'll do it myself after all".
this is the correct way to use AI for art, and i will die on this hill.
Thats what i did! 10 version of the image i wanted i just decided to be the dam thing myself and used it as inspiration.
@@chrisgenovese8188 I mean you are entitled to your opinion but to think there's only one way to use AI is stupid.
Sounds like you were using it wrong if you couldn't get what you wanted.
@@overlords2722 yes, I'd agree. You need to know the basics of code. This is why I can see it becoming its own job or part of the artist tool kit like a lot of software nowadays
It's honestly crazy how AI was meant to assist human living and, ironically, free up more time for humans to express themselves through art and craft. Now here we are, capitalism wants to replace artists to save money.
For me, art is never about seeking to see the world differently. It's more about experiencing the artist's work and feeling something. The only time I ever teared up because of an artist's work (musicians not included) was when I visited an exhibition called "The Future is Old" by The Kid at the Moco Museum in Amsterdam in 2021.
Since AI arrived, I've noticed a significant drop in art job postings on popular art websites so I think this assessment is too positive. I'm talking from regularly seeing 200+ posts at any given time to 15 - 20 now. "Humans need not apply" Had it right the first time.
Honestly Wisecracks take here is disgusting. I may unsubscribe. I can't support this
Yea, it's changing the landscape. I hope that copyrighting an "art style" is something that becomes a reality. This won't help smaller artists, but ones who can develop their own unique style will be able to make all the money with no additional work to deliver.
@@Maxx__________ Sadly they wont. AI is the perfect tool to push down labor costs or replace artists entirely to maximize cooperate profits. Those claiming AI is a tool are just to stupid to realize that's literally being advertised as a replacement to shareholders.
@@Maxx__________ Sounds like a nightmare for those smaller artists whose art might be found similar to the bigger artists.
@@boyblue3270 That's the thing, AI is just exacerbating an already recurring problem, art theft.
Notable to mention: the ethical concern from artists not only comes from the loss of potential jobs, but how AI often sources some of its data from artists’ actual art without their consent, like how ArtStation (a platform made for sharing industry professional portfolio artwork) started implementing AI and used all art uploaded to the site as data. There’s even evidence for how some artist’ initials are replicated accidentally by the AI
This will be solved with money in the future. Microsoft or Google hire a bunch of artists for the sole purpose of training AI and they just pay more than what the artist could make at another job.
is that even a concern at all? how is that different from any other artist learning from existing art? Honestly the more pressing ethical concern is how copywrite law stifles competetion and education.
@@kodaxmax yeah if an artist uses photoshop to draw a straight line did they draw a line or did the computer? Everyone sets the line for what's human made differently
@@F5alconsHouse the human obviously drew the line. By that logic the paint brush painted the mona lisa not da vinci
@@kodaxmax what if they wrote a line of code to draw the line? Is code art? I am not disagreeing with you just asking questions
Wow, you guys were not lying when you were teasing during Wisecrack Live that this video was going to have "a unique take on the AI-Art discussion". I like this idea of preserving the inaccuracies of the generated image and focusing on human agency on its production and use.
We really need more unique takes, all I see is blind hatred.
Your point about AI art being too safe is important-part of making memorable art is knowing just how far you can bend/break the rules. It’ll be interesting if AI can do that on an intentional level.
I think it’s only a matter of time before I get replaced at my job by chatgpt. My boss is already using it as an excuse not to give me a pay rise so I can make a liveable wage as a copywriter. Once it happens I’ll join the anti-robot uprising.
Your employer must love losing money. AI copywriting is crap
Well I for one welcome our new robot overlords and can't wait to hang out with them 😅
@@Macxermillio have you seen all the poorly written articles shat out over the years? It can only get better
Anti robot uprising can't solve the problem, But communism will! Workers of the world unite!
@@djsolegit with subtle errors
One thing that's apparent to me is that the human artist remains an individual with a singular voice. Influenced by his peers, for sure, but imbued with ego. On the other hand, the AI is trained on a dataset of many and it's purpose is to approximate an average of this lot. The AI has no ego or aspiration to innovation. It's an amalgam of the mass. It cannot stray from the aggregate because the collective is the only thing that it knows. Even at it's weirdest, AI art is still just a collage or an aberration originating from lack of comprehension. Only the individual, drawing from his unique experience, can aspire to novelty.
you are looking for the word soul.
Even AI coders will be replaced when AI realizes that it can write it's own code.
@@StarlightDew we are aware A.I. We’ll take us too but we let it for it could help humanity ascend.
So coder will then write code that control the AI that write that code because requirement will continually expand and become more complex until whole universe is autonomous, so that’s why coding is the monster that grows
@@USATECH-um3of But the coder who codes the AI will be redundant. The AI will be in command. It's impossible to foresee all the ramifications of making something so sophisticated. Even rudimentary AI from a decade ago has yielded unexpected results. It's a Pandora's box and the brakes are off in the rush to have the best. Google has sacked it's entire AI Ethics team. Corporate greed is now in control in a race to outpace laws that can even be thought of to govern it, which means there is NO control.
@@ArcanePath360 Answer is , stock market wants growth , companies who has AI foundation models , cannot just sit back and relax, they need to keep adding innovations on top of it , and increase scope of AI , currently research is on where you can use ai to construct fully automated townships and infra, replicate digital AI into physical AI
as it should be. i want to live in a wally world.
Here's my worst-case-scenario theory: Creative people will always experience the need to be creative. AI or any other "the future is here" technology is not capable of destroying that. What AI is capable of ending, is artists and writers' sense of security around using the internet to share and sell their work. So, if artists and writers can no longer make a living from their work, they will simply stop posting any new work to the internet for fear of AI plagiarism. New art styles will continue to be developed but they will only be shared with other artists through in-person gatherings. Result: the current pool of art and art styles that AI has to draw on will stagnate, and so AI art will quickly become repetitive. Any truly innovative art forms won't be known about by the general public, because the artists and writers will have no financial motive to publish their work and too much fear of plagiarism to share freely. We will enter of dark age of art and writing.
As an artist, I think we just will go to darknet and selling art to peoples who appreciate real things. We will build own art society where no AI allowed and blackmarket of real art. If somebody try to steal our style, we will punish authors of programm, physically. You can't imagine, how cruel artists can be, if world take our sense of life from us. But personally, I just hope world will burn and AI shit will be burned with all of humanity.
@@archie2.8 how resentful
Bro
Nobody needs your art
People aren't idiots, they will make their own.
Ai will advance enough to even produce new types of art not far into the future and also new types of artists would appear that couldn't write or draw in the past because of lack of skill and laziness or lack of technology, that will completely overrun you old artists.
Like game developers
Think about elden ring
It's masterpiece of art and it's a video game
Think about all game developers that can easily create new games with high quality because of lack of need of stupid skills like programming and so on...
Just because you are going to be crushed by something that doesn't make it bad.
@@Fang_Zheng nobody needs creepy uncanny valley shits, produced by machines from stolen work of real artists. And NO, supertechnological calculator never will crush me or any artists, there are always be people, who appreciate REAL art, not ai fake bullshit.
Only one reason why you so like this useless shit, created not for humans, but just for sake of technologies (it's not my word it's what sayed in interview by geek autists, who created midjorney) - it gives a lazy technology obsessed people illusion, what talent is not a real thing, and every basement dweller idiot can create a masterpice with little help of smart calculator (you need to be completely retarded to beleive in some bullshit like "artificial neural network", men it's just marketing name for new kind of programs, the same f*cking calculating by processor, nothing new. Only new thing what geeks teach programs to stole part of works and doing f*cking collages from stolen microparts.
Nerdy weirdos don't care about new forms of ART at all, most of them just want destroy possibility of art, because they jelaous (it's hard to be talantless, I can get it) and have memory about bullied by popular childs in their childhood. Also some of then just narrow thinking and cannot understand concept of art, they live in their autistic world of "beautiful numbers and computer logics".
And one thing definitely I can say about ART: even with all this fake ai possibilities lazy and stupid peoples can't make a shit. Elden Ring is masterpiece because REAL talented ARTISTS, not because "smart" programs. OMG, in unbealiveable how stupid can be some folks...
@@Fang_Zheng YOU arent making real art when you tell AI to make something, you are commissioning a piece of art the same way a rich guy hires an artist and asks them to create a piece of art in the style they want.
so you are not actually the artist. the reality is, you personally will never be a real artist because you dont have the talent. you will just be a commissioner of it.
Sounds like YOU are the resentful one. lol
Actually AI is starting to learn from its own output. But even so, this still ignores the real problem of tying survival to work in a capitalist system, and it's kinda elitist, to deny normal people of art, making it something only the few with money can find it
For now, AI-generated art would be best used not as the end product for one's art, but instead as a reference point for inspiration that should by iterated upon by human artists to make something original. After all, art AI is just about as much of an art tool as Photoshop and so on.
You can also take multiple images from AI and combine them, the results look amazing.
That's like saying self driving cars are a tool for taxi drivers
@@applemarker4731 This is why I'm pretty much sickened at 'just adapt' arguments like how?
How do we adapt to something that's does far more than our human capacities? Just put different images in and boom, right? Well then what? Somebody else done the same thing as we did.
My favorite abstract artist is Barnett Newman. He was an abstract artist who made a series of four paintings called “Who’s Afraid of Red, Yellow, and Blue”. The paintings are, to put it kindly, simple. At least on the surface. People hated that these simple abstract paintings were so highly regarded, that two people ended up destroying two of the paintings, specifically numbers 3 and 4.
The museums tried to restore these paintings, and they failed. Why? It’s just red, yellow, and blue lines on a canvas. Can’t be that complicated. But as it turns out, it was very complicated. Because Barnett Newman didn’t just go into his local Michael’s and buy red, yellow, and blue paint in bulk. He made his own paints from scratch using a complex mixture of ingredients, the recipe for which had be lost to history because he never wrote it down and died before the vandalization.
This is why I love Barnett Newman. Not only did he prove that abstract art can be complex and meaningful, but he also demonstrated that there were in fact people who were afraid of Red, Yellow, and Blue. Afraid enough to go to prison for destroying it. And that’s the kind of amazing art that AI just isn’t capable of.
How interesting. Rothko was the only Modern that I was aware created his own paints so painstakingly.
So what's the point of it, is it art or chemistry, you just got played for a fool
Brilliant stuff guys. it reminded me off Chris Hayes take on para social relationships. He incorporates Hegels master-slave dialectic, i.e. the failure of recognition. On the surface level we crave attention but without recognition of the other as an equal our needs remain unfulfilled. This is basically impossible on social media (one reason being scale). But back to the AI. Without the recognition of the creative agent on the other side trying to convey meaning, it will be entirely random if we as an observer find that meaning. And even if we find it then it wasn't the AI that made the art but weirdly the observers. In this regard it is the collective recognition of a second or third etc. observer that would instill the "art" property to an object.
Can children create art? AI is in its infancy so perhaps its art isn't valid (yet) like how chimps can't be artists to some people. It's similar to morality in that supposedly only humans are capable of moral decisions. How do we define art and are we gatekeeping narcissistically?
To take it further, can non-sentient things like nature produce art? Biological evolution results in very creative solutions that are often weird, imperfect, and mimics intent.
how hi r u rn
In her book Weapons of Math Destruction Cathy O’Neil also explores this idea.
“proxies are easier to manipulate than the complicated reality they represent”
Really interesting video, especially love your thoughts on abstract art and how its a window for a "different" perspective onto our worlds, that the author sees, or wants us to see.
Poetically captures what I love(or hate, if I don't "get" it haha) about abstract art
About what you said about mass reproduction of art and the original artwork having an aura lost on reproduction, how would you say it works with digital art. Where from the very beginning, the art is made on a medium that's meant to not lose a lot upon reproduction(ignoring the jpeg loss over repeated reposting, of course).
Although if I think about it too much, it was this element of digital art, which made it easier for Machine learning models to learn and reproduce similar works easily.
So not trying to be pedantic
MIDI is not a reproduction of sound. It’s a computer command language like C++.
MIDI sends commands to synthesizers and samplers *THAT THEN* produce sounds.
So it’s audio synthesis that is the analogue here, not MIDI.
Ah, thanks for that! Your username indicates that we can trust you on this stuff!!
@@WisecrackEDU
Yeah, you can, he's right.
what he says with hope in the end is actually terrifying. When you aren't willing to say that markets are evil, you are not qualified to analyze the effects of this tech
This is like a person in the year 2001 saying “video game graphics will never get better than this”. This technology is relatively new and it’s already this good. I can hardly imagine what a few years will do for it.
4:05 Actually, A.I art can sorta do this! If it generates an image with a mistake in it, say a disembodied hand holding on to a shoulder, you can then use said image as an input image to tell it to generate a new image from that. This can cause it to re-interpret different parts of the image, so the disembodied hand might turn into a tattoo or something. Or maybe it will double-down on the disembodied hand and add extra detail to it, who knows?
This is the kind of application that I believe CAN and WILL lead to actually good AI art.
Right, but it needs intention from a human to get that process rolling, though I guess making an auto img2img in-painting would sorta do that.
I definitely feel like your remarks about "replacing" vs "enhancing" and "trying to use a tool to replicate something it can't actually manage" are well put.
I definitely feel like it reminds me of a lot of tools and technologies, where there's always a role for humans to play.
In some ways it reminds me of how many digital graphic programs have the "random noise generator" factor, where something like the grain of wood or metal, or the surface appearance of water or fire are replicated, but the randomness itself almost feels too uniform.
In many ways I feel like an easy mistake to make is to treat the AI generated art as the finished product. In my mind, a more effective method would be to use it initially, then have a human artist either modify it or create something new based on the AI generated art. I feel like the technology could effectively serve as a shortcut for human artists, where the AI does the work it can do, and then the human takes it to the next level (beyond what the AI can do).
In my mind, tools like these are a great way of reducing "what a person needs to do (the time & effort involved)."
Ideally that's what I initially wanted that out of ai. To take out the remedial tasks that no one actually wants to do (in my line of work). But that idea is tainted with people who purely think that they can get away with just using ai without understanding the fundamentals of the art and design (more so design in my line of work). I wish I understood more the qualms of those in the design and single image art due to certain people thinking and projecting the notion that artists will be replaced (it doesn't help especially when they don't understand the fundamentals at all).
Personally I'm a cg generalist and am learning on improving my 3d sculpting for character design and I see my ai bros (a peer of sorts) thinking that they can just generate one n be done with it. While this is true, it also depends on what their goal is. I see it as him just wanting to generate for a different target market (possibly the book cover demographic) where as my goal is to actually work animation n vfx houses (the amount of changes and requests they ask is ludicrous). In that sense I don't necessarily feel threatened because the ai bros are only looking at the final image where as I literally have to rig, light and animate specifically to what the client wants right down to the timing n music.
I don't think that's a good idea. If artists don't train their ability to come up with ideas but give this task to AI, they will loss the benefit of developing that ability, and rely on AI. It's not a good idea to give such an important part of human creativity - coming up with ideas - to AI. It's one of the most important skills that creative humans need to learn, and handing it over to be automated is not a good idea for artistic development.
Hi Michael.
I see you made a point about midi and its arrival in 1983 being a threat to musicians. First off, midi isn't sound. Midi is performance data.
It wasn't midi that was the threat, but the idea of "synthesised sounds" was, when synths like the Moog first appeared in the late 60s.
Sound or audio synthesis as a concept itself predates midi by 30 years.
The first analogue synthesiser was 1952 by RCA by Harry Olson and Herbert Belar. Independently, Max Mathews created the first digital synth around the same time. During the 60s Robert Moog was a synth building pioneer and helped popularise the synth.
Roland and others helped to solidify the MIDI protocol standard in 1983 because of the array of connections all these different synth modules needed.
Standing for Musical Instrument Digital Interface - midi and midi data itself contains no sound data only performance or control data - pitch, attack, duration mainly; although other data can be sent, audio cannot. One piece of "midi performance data" can be interpreted by any kind of synth module or interpreter and produce wildly different audio output, depending on synth parameters.
This is of course a gross oversimplification of midi, it's history and development or capabilities.
two things:
1. as long as a humans are the ones choosing and modifying the output of the AI, then AI is a terrific tool - no one says you must accept the 1st thing a machine puts out, and no one says it must remain unmodified. If you, as the user, see an element in a randomly produced image that you are affected by, then shape it to emphasize what you feel, the artistic process becomes your own - just like you don't control the sunset you photograph, but you do control how you expose it, frame it, and whether or not you show it to someone. This is where the intentionality lies.
2. I honestly thought you said "mayonaisse impressionism' and was really excited to discover what this is.
When I see AI art on social media, I'm often struck by who is calling it groundbreaking and who is calling it weird and crap.
It feels like we live in a age with facsimiles of creation rather than organic creation itself. A RUclips video such as this, or an art film, always feels more nourishing than most reality TV shows.
Yet there is a whole sector of people who want the lesser product. The Kardashians content is held in higher regard than, say, Brideshead Revisited (picked because it is also about the top echelons of society).
I think the people who want AI art to succeed will push for it and keep developing it. We will see a hyperbolic response from human created art as a response and audiences will pick what they like.
I saw a video of how this guy used AI to help him with character design for his comic book, it involved a lot of taking his favorite bits and sticking them together in Photoshop. But it was pretty cool considering that the guy was not an especially talented cartoonist but managed to make such excellent character reference sheets
I think there was one for shadaversity. He did a couple videos on how ai art can be used well with photoshop or if you just need something for your private D&D game.
He had a video designing superhero characters cause he wants to make his own comic studio to make more traditional superhero stories
@@Nostripe361lol xD the ai assisted skills progress meme guy, I can take seriously that dude
@@igkinatsu I don't completely agree with him either. I am fine with AI art as a toy for fun. But I don't believe it should be used in any way for commercial purposes; both due to it not looking great and the moral issues of using it for such acts.
I was just saying that he brought some arguments for how it could be used if you want to use it.
I don't really agree with this way of 'bypassing' an opportunity to learn and gain skills, and just give the 'difficult' part to AI. We learn to overcome challenges, if we keep dodging them then we will never grow. Character design is a skill. If he has problems with it, doing a comic book is a perfect opportunity to learn this skill. From what I can see, he has just given up an opportunity to learn and develop this skill because he wants a quick result - this may look great on the surface, but what he doesn't learn only harm his own development. People often tell me ' I wish I had done this or learnt this when I had the opportunity', well then, stop giving up the opportunity!
A lot of new technology is used, at first to approximate some prior way of doing things (and often does it badly) before moving on to become something in its own right. It will be interesting to see how it evolves, especially as AI has a tendency to start training other AI and becoming more complex than the human programmers can understand.
This part Human programmers not understand what's happening is already happening now. A Lot of ChafGPT and it's successors are happening because of that. Even Microsoft admits it in their research paper.
It accels in uncanny
im just imagining renaissance artists complaining about how quills and the printing press are cheating and gonna steal their jobs.
@@kodaxmax The use of quills to hold ink is about 800 years before the printing press, but the printing press definitely put many scribes out of work over time. Quills replaced reed styli. Somebody who grew/dried/imported/sold reed styli lost their job to people plucking and selling swan and goose quills. (And of course nobody cared about the harm to the swans and geese.)
Great and interessant takes, as always. A follow-up on whether ChatGPT/AI can replace fiction authors would be awesome.
I'm a sculptor and can't draw, so any time I'm tempted to use an AI art thing to sketch out an idea, I'm immediately deterred by having to register or sign up. I have horrible self-control, but if there's one extra step involved, that's enough to stop me.
Why sculpt? You could use the new AI modeling tool
Idk maybe its his passion
I made a doctor replacing an arm with an octopus, a Chinese brush painting of cthulu, master cheif selling a motorcycle, cute fuzzy monsters are always fun
Sandow Birk is one of many artist that have changed my view on subject matters, use of different mediums and visual story telling.
Was expecting either the scared-denialist response or the tech-bro-overpromise one, but I am very happy with this video and especially the MIDI analogy. It's just a new tool, we still gotta direct it to have the pixels it makes to print new art.
Thanks! And yes, the MIDI analogy really stuck for us.
@@WisecrackEDU Sadly it's a very flawed argument. You need to look at the intent behind AI, It's whole intent and being talked about by the people creating it is to Replace artist, journalists and more. It's not being designed as a tool for artists, it being made to replace them to further the profits of our cooperate overlords. Unfortunately Wisecracks whole argument is incredibly flawed. For context, I am a professional visual artist and musician / sound designer with classical training and a minor in art history. The Midi comparison simply does not work. A better comparison would be the use of of cheap chinese labor to disenfranchise the working class here in America within factories. Midi was always designed as a tool needing a musician to use it. AI is being designed to work with out humans, even without "Prompt engineers" Soon it wont even need them to function effectively, just look at the myriad of Chat GPT run business sprouting up right now. All making profit completely automously of human input.
Side note, your understanding and description of midi is just wrong.
On one hand, I can see how the outrage of "AI will kill the artist industry" or "AI art isn't real art" sound EERILY similar to the outrage a decade ago around digital art, "Digital art will kill the artist industry", "Digital art isn't real art"
Once the dust settles, we will find new and advanced ways of making art using AI, and live in a new world.
But the issue of copyright and AI learning on artists' data and producing reproductions of their styles they've developed over the years, thats really troubling.
Coz the companies selling the AI Art tools are profiting off of all the digital artists
@@daemoneko It's honestly so much worse than what you are even describing, just sit in on the shareholder meetings. It's extremely dystopian.
I'm no economist, but if all the companies start using AI to generate the products and services they sell instead of paying human workers, who are they going to sell all the products and services to, since no humans have money anymore?
This is like asking "Will horseless carriages ever replace actual horses?", the technology is literally still in its infancy.
Interesting thought slash actual experiment. Could you get an A.I. image generator to generate an image based on Kafka’s parable “outside the law”, and then get it to generate one based on Derrida’s, Benjamin’s, Agamben’s, and Butler’s analysis of the parable.
This is what i've been saying (in lamer terms) to my friends. Art is something inherently human, AI art can become a tool, but i highly doubt it'll ever completely replace humans, it's a fad now, as humans we always criticize art really hard (especially writing), and at the moment this is all new, but people is starting to find the telltale signs of something done by an AI, and soon they'll find AI art to be super bland. So yeah, it's gonna be a tool, i can see the benefits on say, a small videogame studio using AI to create conceptual images to show ideas before starting to realy work on them, or maybe a small comic book using AI art to fill backgrounds for less important panels. Things like that can really lessen the workload, letting artists focus on what's really important.
That's been my general opinion on AI art too, that it will eventually become a useful tool for creating orginal art.
I do wonder if AI could be used to create images in existing graphic design software, with all of the layers and elements seperate. That way you could have an artist give the AI some prompts, the AI could produce generic images, and the artist could use those as starting points for orginal art that actually expresses the idea they have in mind. Also the artist could fix the weird issues like numbers of fingers and teeth.
To be fair, I'm not an graphic designer, so maybe that would be more trouble than its worth, and AI will be limited to images that are acceptable as generic, like backgrounds and crowds of onlookers.
'the internet is a fad!'
@@ericvulgate I don't mean that AI's themselves are a fad. People thinking AI art looks good is what's going to go away, right now its a novelty, but people are already getting tired of it for a reason.
I really like wisecrack's take on this. That being said I'm afraid for all the people who work in telemarketing or remote support and such that might lose their job. Big money WILL TRY to work workers for profit
I loved how Alexander Bruce saw the world.
He is the game developer behind "Antichamber"
He saw arrival to a different culture as an arrival to a different world with solid set of rules that are unknown to you initially. As you explore this world, you learn more about its foundations and get more and more at home in those rules.
Johnathan Blow did a similar thing in his "The Witness", but I didn't like it, even though I wanted to love it.
I also like the way Daniel Mullins shows humanity shining through in his games. It's hard to describe without spoilers, but he has a meta characters who are creators of those respective games (Pony Island, Hex, Inscryption). And Daniel manages to show their humanity in those games
AI art can be fun to look at, but it all looks like the same person did every piece. It has already dated itself as a result.
People just think it is "good" because it is trendy. I'm already over it. Soon most people will be as well.
Now, I DO think that illustrators will lose a lot of work with magazines and websites opting to make something "good enough" for free. That's just the nature of that industry, and has been the past 50 years. Moving from Rockwell, to airbrush, to hybrid digital, to Photoshop, to stuff made on tablets to now AI.
Make it fast. Make it cheap. Make it yesterday.
I hope most people will be able to see the datedness of AI-generated images as well.
About the industry tho, I think you have overlooked one fact that companies are always competing with each other and want to stand out and differentiate from each other, and if every company opt for cheap design, and cheap imagery to represent themselves, the one who doesn't will stand out, and their audience will respond. This is particularly obvious in industries driven by culture. AI-generated images are trendy now because it's the buzz word, it's so new and so shocking, but how long this fad lasts is yet to be seen. One thing about industries that are driven by culture is nothing lasts forever, stuff will always circle through.
Oh man, I'm freaking out, I'm freaking out! Also, your dash of comedy brings much pleasure.
Wisecrack name checking Destroyer was just what I needed -- I haven't thought of them in ages
MIDI and my favorite instrument, the modular synth, made an appearance in this video and I’m very happy.
Lean into the synthetic imperfections and push them until you hear or see something new as an artist and your fears about the AI revolution will dissipate.
It's so frustrating that every time we take a step in the direction of a post-scarcity society, the discussion always becomes about what will happen to the people who lose their jobs. We could obviously afford to employ them before. Why not demand that companies who want to use AI still have to pay their workers, even if the AI can replace them? Oh right, capitalism...
I find it so ironic that as we take steps towards a post-scarcity utopia and kick out capitalism, people will fear the unknown so much they would rather cling on to the capitalism they so much despise out of familiarity. It is basically stockholm syndrome.
@@fighder2 we care about the transitionary period where we may starve to death. There's also an existential threat here relating to profession as a major aspect of our view of self
@@steampunkpainter42 Can you elaborate on why you think phasing out capitalism will cause us to starve to death? That seems like a non sequitur to me.
I also don't think that people will stop working completely. They just become much more free to pursue other interests.
I think the MIDI analogy is good. What you can do with synths and instrument samplers now is infinitely superior to what you could do 40 years ago.
If I can make a request, wait a year or two, and see if this video needs a sequel.
I always really loved Wisecrack digging into media/film interpretations, 8bit philosophies, Earthling cinemas, etc.......Im still subscribed to Wisecrack- Its just.. ya know, Pop-Topica. Hey @Wisecrack, why's everything homogonizing into Pop-Topica?
I'm very paranoid. I think this early stage of AI-generated art is less about implementing where it's needed, and rather to gather all the wrong answers first so the AI has no other option, but to produce correct ones. It's kinda scary because our curiosity will compel us to try ChatGPT, but each search is just another step into becoming obsolete. Then again I have no idea how this stuff works.
Pitika Ntuli is a stunning sculptor with a great body of work but this video makes me think about some of his more recent stuff. Recently he started using bone instead of granite or marble and he makes a lot of stuff with elephant bones which he carves and then beads, embeds circuitry or other computer parts and often couples with writing by himself or collaborators
AI art already doesn't suck. Hands and faces were already solved with LORa.
Honestly, things go so fast that when people who are not in the loop talk about it, they are already a few leaps back.
This, I've seen so many videos of people proclaiming how AI art won't replace anyone because it's crap drawing hands, and I'm like, eh...that has already been solved. They don't realize just how insanely fast this technology is evolving.
I get the feeling Wisecrack didn't look at what stabilised diffusion is already doing before making their video
Bad art does not equal unskillful art. Kids art is still art even if they suck and drawing hands. You missed the whole video if you think this is about hands. Good art conveys a worldview - it has a perspective, at least a point of view.
This is one of the most disappointing Wisecrack videos I've seen in a while. I expect them to make relatively intelligent videos, but in this one they did no research at all. This is evidenced by all the talk of 'how bad AI hands are' which was partially solved in January and is pretty much entirely solved today. Unless of course they just use DALL-E's crappy image generator to gen images without looking into all the other options. I assume that's exactly what they did.
Yeah... I was like... "Michael, what the f**k are you talking about?"
When you don't know something... just shut up and go do some good research first.
The best dall-e art I made was typing in my friends name (calder white) in an empty parking lot at night and it was honestly so eerie and beautiful and so stupid - imagine a large piece of paper being stretched in many crossed directions, in an empty parking lot
AI won't replace artists , conceptualism started at least a hundred years ago I think artists have been ready for this 🙆♂️🦔
@5:15 Zdzisław Beksiński - the greatest artist from the 20th century (for me).
Banksy. His name has probably already been said but his is the first name that came to mind when speaking of artist's that helped you (me) see things differently.
Great call.
Finally, a good take on AI art
This is one of the smartest ai art videos I’ve seen, kudos!
What are your thoughts about artists’ work being uploaded to AI data sets and used without the artists’ consent for generating AI art. Is that machine “inspiration” or theft? Given that artists are going out of their way to corrupt the images’ code (through things like Glaze), isn’t there a question to be had on the ethics of consent, ownership, fair use, inspiration and theft. Does AI’s inability to generate from its own imagination or heart and reliance on the work of others with no critical thinking or purpose result in enough difference that the rules for human inspiration should not apply? Should AI companies be paying royalties or purchasing images that are uploaded into data sets?
My guess: we'll basically get some interesting developments in high art, using AI as sketch or background tools, prompts, scenery, or in surrealist art - and a huge loss of jobs in graphic design and applied art, where one person will be able to do the job of dozens much faster.
Midi can sound like anything nowadays. It is just a way to compose with digital instruments (VSTi's). If you use midi to play a modular synth with a waveform that is exactly right (and you let it decay just right) you can mimic anything really. Layer enough sine waves together and you can get anything. That's how digital sound works. All a bunch of sine waves (in correct respective frequency and phase, and modulating just correctly over time).
if you put enough into your prompt, the output can truely say something about your intent on creating that art. in a way you could say these are just another artists tool, lowering the bar to physical skills and raising the bar on digital skills. when you combine it with graphics software like Adobe Photoshop and modifying existing images you or someone else has made, you may have stumbled on making art with more meaning or something completely different.
An artist who has changed how I see the world is Mark Rothko
Nice.
you want my favorite prompt? i had alot of fun with this one (even if the in-joke is lost for others) "Greggggggg, living spell of summon natures ally, the mass of a planet, creature made of one million animals fused together and all the animals are trying to escape, as card art for Magic the Gathering illustrated by Robert Bliss and Junji Ito"
One thing that annoys me when people say that AI can't do this, is that they never say what we're all thinking: They can't do this "YET". I always emphasize the "YET", because it's not a matter of if but when the AI will be able to do it. I have no doubts that AI art will keep improving to the point that it will mop the floor with human artists in terms of technical quality. The one aspect that it won't be able to reproduce is the intention behind the art. That's the only value that will always belong to human artists, that is unless AI manages to develop sentience somehow.
What's a "traditionan artist" anyway?
Somewbody who paints oil on canvas, I guess.
If you opened Photoshop or after FX in the last 10 years you would know how many shortcuts exist that already divide a user from a "traditional artist".
A pro working for Disney took a look at the AI capabilites and said "we use to rotoscope by hand, it took weeks, then we did it with content aware software, it takes days, with AI you can do it in hours, and that's a good thing".
The "art" is what you want to communicate with your creation, it's not the tools you use to get there.
Correct. Every single day it is reported that AI can do at least one new thing
@@ChristianIce Sorry, I chose the wrong term. When I said traditional artist I meant the human artist. Also, I agree with your last point.
If it comes about what art pieces I like... Still remember, though it was years ago, how great impact had some paintings of Odilon Redon, or Witkacy, at me.
These vids are getting better each time.
I don't think it's going to get much better. The era of easy money is over and the machine learning software companies are rushing their flawed and buggy products to the market in the hopes of and getting some quick cash by selling shares of their companies. Now that they've scraped the web for all our personal images just to make software that creates silly pictures, they'll try to pivot to creating products for governments and militaries. The question is, are we OK with our visual data being used in such a way, because it's not just artists who had their images scraped, it's everyone.
One of the first weird prompts I gave Dall-E was "post-apocalyptic dildo." Terrifying.
talken about midi music.... while listening to FF VI ost.... while FF VI flashes across the screen... (chefs kiss)
I think we will have AGI in the next 5~ years, ai absolulty can be creative and do any other thing a human can, just not yet.
Total side note, that backdrop is amazing
Lots of good points made here!! I think machine learning will soon be able to make plausible or believable images and be honed in its craft. But I don't think it will ever make art that will be able to reach out to viewers centuries later with the same wonder that artwork from the Italian Renaissance or the Impressionists have. I don't think artists (and hopefully graphic designers like myself) will ever go away. Humans value the human hand at work and I think there will always be a desire for handmade artwork. Cheers!!
thats not really what ai generated content is for anyway. A single AI image is never gonna be impressive. But a game built with AI generated assets? definetly could be. Imagine if notch had these tools when he was developing minecraft, it would likely look completly different, as an example.
Id also argue that no modern paintings are ever going to have the same impact as the ancients famos ones, simply because they did it first and with far less resources and tech. That makes everything after inherently less impressive.Keep in mind AI right now can make better paintings than most of those painters and in the same style. But they obviously wont be held up to the same worship.
For a split second, I was left wondering what "mayonnaise expressionism" looked like.
I mean, Manet just basically threw condiments at a canvas, didn't he?...
9:11 I want one of these gibberish “live laugh love” pieces in my house lol
I think also in regards to the reference of a quality of space, the quality isn't really where it is but how it is in space. A photo/reproduction doesn't actually LOOK or feel the same in person. So people who haven't seen much art in person don't get that there's such a HUGE difference in the physicality of a work, especially with painting. I've very often thought that work that looks amazing in a photo, just felt so flat and lifeless in real life-but also visa versa, the works that look kinda flat and lifeless in a photo sometimes take on a quality you never would have noticed if you hadn't seen it in person! (example: picasso and monet- kinda disappointing, Suprisingly better than photos-cy twombly, basquiat, and da vinci) I think it just emphasizes the idea of AI being a tool, rather than a replacement.
Best pictures I got was from.
“Taco Bell flaming hands Doritos, 90s commercial”
I think you are missing the point. 1) It is still humans who prompt the AI to generate the art. It is those humans who provide the higher level stuff, such as providing a perspective of the world. 2) Most art doesn't have to wow. I suspect 99% of all art goes on magazine covers, one off ad campaigns, etc.
AI will favor artists with a strong brand or who developed a distinctive style (as long as they can defend their IP). But it will really hurt creatives who are not very creative.
All my creative impulses have been channeled into computer coding for a while, so it is nice to have AI art tools like Stable Diffusion invading my space and engaging my impulse to make visual art. The experience of making images with AI does definitely remove my sense of authorship, making it more like an exercise in curation or editorializing. Some images come out wholly complete and I don't want to alter them, which is more like stumbling on a beautiful corner of the latent space. Other times I try to obtain a very particular result and struggle to achieve it, almost an exercise in futility. I think the best results will come when I can draw something "detailed enough" on a tablet and prompt the AI to fill in the details without having to switch modes too much. The flow of creativity is important to bring out something more personal. It's the very earliest days for these new tools, and I have no doubt it will only get better, or at least more interesting.
This hits the nail on the head. Well done.
M C Escher is one of my favorite artists
I'm a little disappointed that you didn't mention "ready-mades" or "found art". And I'm very disappointed that you only had one quote by an artist, an offhand one at that. I just have a BFA, so take this with a grain of salt, but for the next video dealing with Art, be more balanced between artists & philosophers. Yes, I know, philosophy channel, but if philosophy is thinking about thinking, who thinks more about how we think about Art than artists?
BTW, 4 of my fave visual artists are Magrette, Giger, Fairey, & Waterson (the Calvin & Hobbes guy)
Actually, would love a video on the philosophy of Calvin & Hobbes/Bill Waterson!
Fair points Matthew! Appreciate the thoughtful feedback.
@@WisecrackEDU I do actually quite enjoy this channel, have since 8-bit Philosophy, would always take any video over none.
Question to you all: How can we make people value the authenticity of actual human work (not just in arts) more?
I think most people don't actually care. And I feel sorry for them, for having less and less true "human experiences" of connection and shared feelings.
Did you like Jurassic Park?
Did the fact that some of it was artificial take anything away from the experience?
People should care, yes, they should care about rejecting shitty products, doesn't matter the means used to put them together.
There's no CGI in "The Room", it's all human work, no AI, no special FX... but it sucks big time, doesn't it?
People have different interest and not everyone have an eye for quality. The good thing is that the overflowing amount of AI art, about to get produced, will make us shift toward expressions that AI can not so easily copy. Something real and rough. Something you need actual emotions and body movement to make.
It is like all other trends. Nothing interesting about the stuff everybody have. The stuff that is everywhere is not worth anything.
It is up to human artists to make something AI can not reproduce. And honestly my favorite art styles are not easily replicated by AI, because AI lacks the how and the why of a human brain.
@@ChristianIce I think this misses the point. The creative part in Jurassic Park is still human-made.
But let me give you an example: When you see a cool image completely created by an AI, knowing that it the create might just have used 3 (!) words (!) as prompt… can you still be wowed by this? Can you become a fan of the creator?
In the age of generative AI, I can be impressed by technology and show praise to the creators of that technology in the comment section, …, join a community of fans of that company. That is something, I admit!
But is this… enough? Is this sustainable for you as a social being?
In this new era… who can you aim emotions like thankfulness, closeness, or - in case of "The Room" disgust and disappointment at?
No one! There is no one… except for the anonymous developers and faceless corporation which created the AI.
Generative AI is a soulless tool by itself, isn't it?
Hence, if (!) there is no significant (!) involvement of a human being in it, it's output is soulless by definition.
@@svanemy Great point: "Nothing is interesting about the stuff everybody have. The stuff that is everywhere is not worth anything."
It's not what I was referring to, but it still ties into it. Soulless products, without any heart.
Just today I have prompted an image creator AI to create birthday greeting image to send to a friend. It looked awesome, truly great! -
But: I almost felt embarrassed sending it! Because there was no true craft, no love and energy put into it, not even money (to pay an artist).
And the recipient didn't even reply to the graphic - because nothing creative is worth anything anymore.*
*(Overexaggerating a bit here to bring my point and my frustration accross. :D)
@@Nasarbajew84
AI is just a tool, and a tool in the hands of a pro compared to the same tool in the hands of a random person is not the same thing.
Generating a random image doesn't make anybody an artist. If you have nothing to say you can generate 1000 of those and nobody would care.
Nobody should be “anti-robots” or “anti-ai,” but anti-capitalist
ZZzzzzzz......
True
Capitalism made west strong, and strength is justice.
Exactly. No one would be worried about “robots taking their jobs” If they weren’t terrified that they’d be homeless and hungry
@@Fang_Zheng Made thanks to the suffering of the rest of the world
No matter what writing task I give ChatGPT while playing with it ("write a poem about..." "write a flash fiction piece about...") it veers into kind of cringe, ad-y language. It's definitely not at the point where it can replace writers, even for good marketing copy; good marketing copy ideally wants to sound natural and original. But it could give a boost to a copywriter in a rut that day?
"Arnold Schwarzenegger riding a tricycle" produced some pretty funny stuff, as did "The Witch-King at the waterpark", "Elmo and Tom Bombadil smoking weed", "Bernie Sanders with shotgun asking for money", "Lao Tzu fighting a T-rex", and something about a capybara suing Elon Musk. Unfortunately, with "Kermit shoots Sauron" , "Bigfoot and Littlefoot buying shoes", "Jar Jar Binks did Pearl Harbor", "Bud Spencer punching Mussolini", and "Kaiser Franz Josef shaking hands with a Schnitzel", the prompt was way funnier than the actual art produced.
Ähm did anybody from this channel saw images from the current version of Midjourney? They can be photorealistic and nearly unidentifiable from real photographs.
The showed pictures are from a version one year back maybe....
It may be a tool now, but this video didn't really dive into the possibility of a future AI eventually reaching a point of being identical to humans, capable of the same level of art. I've yet to see evidence to suggest why this wouldn't be an eventual reality. Even an advanced AI fed a "simulated" life should theoretically be able to produce art as if it had actually been alive as a human.
AI is just a tool.
If Adobe Firefly was out before they released this video, I guess some adjustments would have be done.
But, considering they think AI still produces the wrong number of fingers (which was true 3 weeks ago), I think this is weeks old planned video, sadly in a time when every two days there's a revolution.
Max Ernst's art opened my mind to the beauty within the bizarre.
There's a lot of f*ck, lately in wisecrack... i think i love it.
As soon as this piece was shot, AI rendered many of the criticisms null and void. MidJourney V5 cleaned up eyes, faces, teeth and fingers. Exponential growth is… exponential
There's a guy on tiktok who uses a photo to generate a description from an AI and then gets another AI to generate an image from that description. And then he does that a few times over and the results are fascinating! And I think the guy who does it is doing art, it's a creative and transgressive idea and I would call that art but the actual image that the AI creates isn't art.
An interesting video. Have you used midjourney version 5 yet?
I do think digital art is dead or devalued but traditional handmade original art will thrive. People will long for handmade things outside the digital world.
I hadn't really thought of the MIDI comparison. You mentioned DEVO but MIDI is extremely far-reaching in terms of its influence on popular culture by way of a pretty humble thing, the Roland TR-909. We would probably not have modern hip-hop, house, or techno without it. Again, that required a lot of advanced knowledge in programming. I was in an industrial band in my late teens and bounced hard off of drum machines and MIDI synths in favor of more (IMO) user-friendly stuff like Fruity Loops, Acid Pro, and Soundforge, so it definitely took a kind of artistry and expertise.
I'm not sure if I buy the idea that someone might be able to replicate that same artistry via training an algorithm. Taking writing or literature as an example, anything that ChatGPT writes inevitably sounds like you're discussing something with a customer service representative or middle manager at a bank. There are some really interesting examples of how AI fails to write compelling fictional dialogue in people building mods for things like Skyrim or Mount & Blade II. In the latter, one of the videos showcasing the AI tech had an NPC (if you're unfamiliar with M&B it's set in a no-magic medieval world with knights and feudalism), saying "Let me know if I can be of further assistance to you, my lord." This is after feeding all of the lore and other dialogue bits into the OpenAI instance that was being used to manage this mod's dialogue.
All that to say: I don't think we can produce human-led artistic intentionality with AI unless someone has full private ownership and trains it without allowing it to utilize the API of other instances of its software to update its algorithm. An AI artist, in order to develop the conditions for art that speaks for them, needs to intentionally feed the AI specific images, books, articles, films, songs, etc. without the AI being able to crawl the Internet on its own and thereby develop influences beyond the artist in question.
Tell an AI to make "AI making art" or "art making art backwards in time" or even "the same exact art over and over" lol that should reveal interesting results 😏😉
dali idea: Henry Winkler winking in a rink.
Ironically enough, if one's intention is to create purposefully unsettling art, A.I. is actually really great for that right now, as unethical as it is. Just try using DallE-2 to draw something in the style of Zdzisław Beksiński and see what happens.
The idea that AI keeps making itself worse by scraping its own artificial creations fascinates me. A copy of a copy of a copy until humans are lost in a morass of unreality
Copy of A by Nine Inch Nails is stuck in your head
This is exactly what I have been thinking! Like the web was not wierd enough already..
AI art as a tool has tremendous limitations, just as each different paintbrush, camera, or word. There is always a limit. Each tool and medium is actually defined by, and celebrated for its limitations and the specific space and time that appears. It is an experience, whether used to create something or as the audience.
4:57 I'll have some fries with a side of mayonnaise impressionism, please.
Trying to define art seems to me like trying to define intelligence... I'm not even sure it exists in the first place!
So this was interesting. I made sure to watch all the way to the end card because I felt like it had to be an april fools joke a week early and someone just posted it on the wrong day and there was a "gotcha! april fools!" at the end. This is so much of a hilariously bad take that it feels like is has to be. It is full of information that is just flat not accurate in the slightest as well as makes some pretty bold assertions on the nature of creativity itself. I am going to spare you a line by line but I am actually a decently successful AI artist. Based on the content here it feels like you can't have possibly actually talked to any. This is just really surreal because the more philsophical takes you guys have are usually very informative and insightful but there is so much of this that is just flat not correct. If you guys want to talk about it you can shoot me a message. But failing that, wow, that is an... interesting approach to the subject.