I work in the Major Motion Picture industry up in Toronto. The profits of the Producers is constantly increasing by picking away at the income of the lowest paid people. Please be empathetic to your fellow humans.
It’s the same i many industries. I’m in the restaurant industry and as prices rise hourly pay decreases and tips are being spread out to more positions. All the while productivity has increased. If we don’t rise up as a working class to change things they will continue to nickel and dime us into abject poverty while we work ourselves to death for them.
Ever consider why those individuals were paid so little? Because the tasks they did were considered relatively small. And if they do get jobs again, no guarantee that they will, they will more likely than not be better paid, and they will definitely be free from the tedium of such tasks. I personally would almost rather die than be forced to perform an unnecessary task. And I really don’t care how the task became unnecessary in the first place.
I don't get why people who are against AI art see this only through the eyes of a corporate employee or gig worker. Get some friends, make a studio and leverage AI to compete with the industry you worked for. The same tool that allows them to cut costs, is the tool that lowers the barrier to entry.
Long time follower from Gold Shaw and excited to see you make this other channel. I appreciate the points you make. I hope you don't mind, these are just some thoughts as I watch, being an artist all my life, knowing lots of artists, and having once tried to make a living out of it. It is a TOUGH field to work in, it requires a lot more time, skill, and energy than I think ""non-artists"" appreciate most of the time. It is NOT actually easy to find work across the world as a painter, and it is not actually easy to work as one in a professional setting. One of the biggest issues I have noticed is people misunderstanding how long something should take and expecting artists/animators/etc to work "ghost hours" which further obscure the actual effort that went into a task, and asking for massive last minute edits that require way more work than the boss or commissioner realizes. There's a long almost "cultural" history of suffering for our art, and well known stories like the harsh working conditions of artists in the Spiderverse 2 movie, the health problems faced by 2D animators at Disney, the complexities Michelangelo faced painting the Sistine Chapel, Van Gogh's struggle with mental health. But despite how intensive the process can be, many self-employed artists are lucky to make minimum wage. Artists generally take pride in what they create, which kind of makes up for what a difficult field it is, and now feel that their time and effort is being entirely erased and made meaningless. While there's been uproars in the past, and still remains some bad blood between traditional and digital painters, most nowadays can agree that being a good artist in a digital program like Photoshop STILL takes a lot of skill and knowledge of artistic fundamentals. It requires you to think. Times absolutely change, and I don't expect AI to go away, but I am painfully sympathetic to what being able to produce art-competition winning creations with the input of a handful of words into a machine, is going to mean for artists and the future of art as a profession. Nobody (sane) considers the 6 year old tracing Sunday comics stealing because the 6 year old is doing something personal for fun and education, the same way people get by with including "master studies" in their portfolios and Ao3 can legally protect all the fanfiction it archives. They're not making profit or impacting the profit of the original creators. The AI models, meanwhile, have already begun causing problems in traditionally artist-focused spaces like Magic: The Gathering. Even a recent Wacom drawing tablet ad ran afoul when AI art was used in it, which was sold on the Adobe Stock website as original art without being properly labelled as AI generated. It may not be a copyright violation as the law is written, but as you noted, it is a very real threat to the livelihoods of artists who have already been struggling for a while, with their work being (in their eyes) already severely undervalued. When I was in gradeschool, we read a story about a shoemaker at the advent of factory-produced clothing, and it followed him through the years as what had once been a respected and valued profession became harder and harder for him to survive doing. His custom-made shoes were unique and lasted years, their quality far surpassed the factory products. But they were expensive, took longer to produce, and didn't need to be bought often, because of the professionalism that went into it. Ultimately customers would end up favoring the cheaper, standard fit storebought shoes, that only lasted a fraction of the time, and didn't fit as well. He had no other skills, he was proud of his work, and ultimately the old man died in poverty. It's hard to argue that most factory made clothes are BETTER than those made by cobblers, tailors, etc. They're just different, and represent changing times. I would not call something that an AI machine put out "art" anymore than I would call a T-shirt I bought on Amazon that was made ultra-cheap in China "fine tailoring." And people with THOSE kinds of creative skills are now much rarer. I'd argue most people in the US don't have the tools or even know how to repair or modify their own clothes anymore, they just go buy another cheap shirt off the rack at Walmart if something wears out. Those are dying arts, and I forsee visual art eventually going the same way and skilled artists becoming a rare thing. You say that using this tool gives you "exactly" what you had in mind by entering a prompt, and that's fine if you were looking for something very generalized like "family of farmers" - but modern famous human art is often more nuanced than that. You wouldn't call Simba just a lion, or Iron Man just a superhero, or Hobbes just a tiger - they're recognizable characters, both in design and the style with which they're produced. The reason they look the way they do echoes the tools available to the artist, their experiences as an individual, the emotions they were trying to convey. A skilled artist had to put time and effort into refining those individual characters before AI could ever mimic them. In addition to devaluing an already devalued skill, AI is factory produced art, it's got no heart. It's creations are just telling A story, not a SPECIFIC story, and our creative outlets will not be the same if that becomes the dominant method of content creation. Inevitable or not, I find that very tragic. So yes, you're right. Times are changing, and AI art isn't going away. It has the potential to be very useful as a tool, and obviously not everyone using it instead of paying an artist WOULD have paid an artist if they didn't have AI to use. But I wish that there was more widespread sympathy for how FAST this field is changing, and the consequences of favoring convenience and profit over genuine human connection, skill, and originality.
Great insight. It is sad to think that artist could become a relic of the past. That art could be considered a niche skill that very few have or attempt. That art sites like deviantart would cease to be a thing as no one attempts to create images with their own hands.
@@iheart801 Ironically, deviantart's ownership has fully embraced AI as "art", and incorporates their own AI program called Dream Up into the site itself. One way the site makes money is by offering additional monthly uses of DreamUp for premium memberships or currency. Deviantart has been flooded with AI images the last few months, and it's even being tailored to mimic high-profile artist's unique styles by recognizing their usernames as prompts. If you wanted your username excluded, you had to submit a special request through a specific form for review, and they explicitly would not guarantee that they'd remove your name if you asked. There was a huge uproar when it was released, not least because deviantart originally tried to opt every artist's galleries into their AI's dataset by default. The blowback was severe enough that they reversed their stance and now you have to manually opt-in to have your art used as learning material for DreamUp, but that's certainly one of the exacerbating frustrations for a lot of artists around the whole AI vs Art situation. I think some (not all, but some) of the hostility might be lessened if there wasn't such a massive push from certain circles to categorize AI images as art and force them into artist spaces.
I'm a Graphic Designer, I'm trying so hard to find a job after being on mat leave. People are asking for one, people are looking for one. But I can't seem to find one with my skills or in my area. I don't want AI to take over my job when I get it
I feel like this video really misses the point of the debate for me... You spend half the video talking about how AI stealing art isn't bad, and half the video talking about how AI replacing artists is inevitable, and you have lots of examples to compare to each of those points, but no examples that apply to both at the same time. And that's my issue with AI. It's not "I have stolen your art", and it's not "you lost your job", it is "I have stolen your art and used it to make you lose your job". Tracing art isn't comparable, because your tracings didn't replace Bill Watterson. Other forms of AI like facial recognition aren't comparable, because there was nothing similar existing before for them to replace. Analog video editing becoming obsolete, and other examples of technological progress, aren't comparable, because they are just built on the technology base of the previous generation, and are not reliant on the works of the individual creators. AI is entirely reliant on the works of artists, and is simultaneously undercutting and hurting those same artists. It may not be an exact copy of those artists' work, but we call it theft because blatant plagiarism, stealing someone else's content to repost and profit off it, is the most accurate comparison to what's happening here.
Thank you for your candid feedback. Let's address your points: AI Stealing Art vs. AI Replacing Artists: You're correct that the video discusses these as separate issues. However, I believe it's crucial to understand them individually to fully grasp their combined effect as well as parse the morality of each issue. Tracing Art and AI: While tracing doesn't replace the original artist, it's used as an analogy to illustrate the basic concept of replication, and building on previous creative content, a fundamental element in AI art creation. The way most AI models work, unless you are directly inputting an image as part of the prompt, it's not directly trying to recreate any of the content it was trained on. That's incredibly important distinction that I think often gets conflated or overlooked. Technological Evolution: The comparison with analog video editing and other tech advancements aims to contextualize AI's role within the broader spectrum of technological progress. It's not a direct parallel, but rather a lens to view AI's impact. AI's Dependency on Existing Art: This is a valid and critical point. The video aims to spark discussion on this unique aspect of AI, which indeed sets it apart from other technologies. Plagiarism vs. Innovation: The video seeks to explore the fine line between using existing art as inspiration and outright plagiarism, a core issue in the debate over AI in art.
seriously please someone explain how making new art based on someone's art is stealing from them. I really really don't understand. blatant plagiarism is a very bad comparison because no ai software copies artwork directly and signs it as made by them or sells it as original. it's not happening. where is the theft?
I can't help notice that discussions like this always come to the conclusion of the technology isn't the problem, it's how capitalism will work toward corrupting it.
If I am going to be completely honest, some of the AI people and images in your videos are quite terrifying/off-putting. Apart from that, your videos are great and I appreciate the amount of work and effort that you put into them. Your work ethic is commendable.
You are arguing about how AI aids processes by reducing the effort and expertise required to do said task or the increased availability of that process at a lower price. However, that's just feeding on the late-capitalism trope of convenience and low cost are the goal. It does not address the costs, both the obvious and the unexpected ones, that society will experience. Can AI do amazing things with exorbanant energy consumption? Can it produce significant text, pictures, and sounds. Yes, it can, but have we, as a society, answered the moral and ethical issues that this capability presents? And I acknowledge that you touched on some of these, but your arguments were based on economic impacts upon the consumer. What is the cost to society for large numbers of expert creators not being able to get a living wage and are now living on the streets of Peacham, VT? These AI's are generative, what if the source was in the public domain, can I now use it to make a proprietary product? For example, if I ask an AI to produce a picture of Steamboat Willy but in pychodelic colors and with squigily lines, can I now charge for it? Technology always exposes moral and ethical issues that have never been an issue before. The technology should not exist just because it does something, it should exist because it addresses a problem that society agrees needs to be solved. There will always be grifters (i.e., early adopters) who will jump on a tech and benefit, but we must address the moral and ethical issues when individuals benefit themselves but hurt society as a whole. I am a computer programmer with 40 years of experience. When I started, we did simple things like mail-merge to generate form letters for different customers. The union was rightfully concerned about the reduction in work leading to staffing reductions. Luckily, management recognized a changing market and an opportunity to reduce tedious work and replace it with more interesting, engaging work. No one enjoyed typing 20 copies of the same form letter on a manual typewriter changing only the address and salutation; having the time to investigate and fix a customer's problem was a much better way to spend their time. As an IT organization, we didn't look at staff reduction as a legitimate justification for our time; explaining another project that could be possible with some task automated moved things up the prioritization queue. Mind you, it was the 80's; today's corporate environment sees technology as a way to increase shareholder returns, not improve the business or the customer experience. I would also like to make you aware of Free Software (Free as in Freedom, not as in beer, but often free that way as well). Free Software, and it's corporate evil doppleganger, Open Source, publish computer source code publically so that it can be (0) studied and used for any purpose, (1) modified to correct or enhance it, (2) distributed to others, and (3) distributed to others with modifications. Some are also constrained (with what is called Copyleft) that requires that if you distribute either modified or unmodified versions, you must do so under the original license that you received the software under. This is clearly not a economics-first, make Elon money approach to software, yet it is a highly successful model used for the Linux kernel, Firefox browser, Libre/OpenOffice productivity bundle, and Blender 3D animation/modelling package. There has been a discussion around a Microsoft project, known as Co-Pilot, where an AI will spit out code solutions to prompts presented to it. Co-Pilot uses source code in GitHub as the source material. There was concern that a lot of code in GitHub was Free Software under Copyleft restrictions. If an AI generates a block of code derived from a Copyleft licensed code base could that code be embedded in a proprietary software package? Microsoft has made adjustments to their algorithms so that developers, like myself, can opt-out of having our own code used to train Co-Pilot, or so they claim, but how can we be sure? Morgan, I appreciate you starting the discussion, but I think it should not be about whether AI will happen, but rather what are the morals and ethics of using it when it is. How do we prevent predictable negative consequences for a measurable portion of society, and the inevitable draining of pockets to be delivered to those at the top? There is a lot of sublity and nuance to address and discuss, and we must not skip over those hard conversations just because we can instantly create pictures of Ryan Reynolds in a jock strap at our HS junior prom (no one else, just me...? :/ ). AI poses significant challenges to society as a whole, and despite being able to make things that were hard before, easy, we must pause and answer what is good and acceptable and what we will not accept. Art will come from AI, but it must do so in a way that does not exploit others.
This is a story as old as history. When someone discovered bronze worked better than stone, you can bet people got scared to lose their stone making jobs, and instead of learning to use bronze, they got mad, and tried to outlaw bronze
The biggest difference between the technological advancements of decades and centuries past is that typically it was one big disruptor at a time. People could adapt. They could learn a new skill, a new trade. New technologies also meant new jobs and new types of jobs. It wasn't easy, but we could make the transition. Now, A.I. and huge advancements in robotics are looking as though they could replace whole sectors of jobs - possibly even most jobs. And it could do so essentially all at once or at least in a very short span of time. And as always, the new technologies will be able to do the work of a person far less expensively. But this time, there won't be nearly as many new jobs to train for. We will not only be unemployed, we will be unemployable. If you thought the world was unstable and heading for social and economic disaster before... give it five, possibly ten years. A revolution is coming with A.I. and there will need to be a social revolution to go along with it or it will be anarchy on a global scale.
Great video Morgan! I think one thing to add would be the use of the AI art. Using AI art in a commercial use to me is a form of “theft”. You copying the comics was perfectly ok- it developed your skills. But you didn’t go selling those copies and hinder the original comic. Both coexisted. AI art used for commercial use hinders those it draws from and to me that’s the difference. Would love to discuss further (as friends of course)!
I'm a full time professional cat artist. Been one for 11 years. (Yeah, I know! You "hate" cats.) One thing AI is doing, is it's pushing artists away from digital back to traditional mediums. This doesn't solve the problem though, because the vast majority of traditional artists make most their money through prints and merchandise. Not the sales of original art. For example, coffee mugs with my traditionally painted cats on them are now competing with a thousand AI generated coffee mugs. People are no longer buying my mugs because they are the coolest thing out there, but rather because they have a connection to me as a person. Which means far fewer people are buying my stuff. There is an artist purge happening right now. Art subs and forums are full of artists crying over lost income. They are giving up on art. How will this go down? Well. AI is over-saturating the internet. It's starting to train on itself. Self-implode. So it's possible AI can't get any better than it is because of this. Who knows? Meanwhile, it takes 10 years to make a good artist. Same as a doctor, or vet, or lawyer. So there may be ramifications down the road. I just don't know what they are yet. This is for sure. Because of the modern entertainment industry, art is more utilized than it ever was in the history of humanity. How will the beast be fed if there aren't enough artists and AI still can't draw hands? As for myself, I'm just trying to figure out how to survive this purge. And it's not just AI I'm dealing with but globalization as well. Artists from rich countries have had a major disadvantage compared to artists from poor countries. So there's already been a purge of artists in rich countries for some years now. I feel like there's AI on one side, and globalization on the other, and I'm being squeezed to death between the two. I honestly don't know if I'm gonna make it. But I'll try.
When it comes down to it technology is evolving at a rate that exceeds our capability to fully understand its consequences. When the lines begin to blur between what humans are capable of and what a machine is capable of how will we be able to differentiate between fact and fantasy. An analogy I would put forth is when GTA came out and there was an uproar over children becoming desensitized to violence and possibly committing crimes in the game in real life. Now adults are facing the same dilemma, not from a video game, but from real people using AI to create false narratives with real consequences. These false narratives could be described as a form of content creation. They are posted on social media and generating revenue, so technically speaking it's the same thing. As you say, the question is where to draw a line, but who decides where the line is and how do you keep it from being crossed? If I had to put a fine point to it, my question would be whether humans are evolved enough to reel in the consequences of the technology that seems to be quickly out-evolving them..?
This is a form of Sorites Paradox with no definitive or agreed upon solution. I have a few questions after this: Is Gold Shaw Farm (or any form of the name) trademarked? I’m sure you probably have an LLC as a business, but curious if you’ve trademarked any portion of you, your content, etc. What are your thoughts if someone were to use your likeness, your stories, etc and undercut you? Say if someone were to use AI art to create Gold Shaw Farm T-Shirts with likenesses of you and sell it at a fraction that you sell it? What if someone were to use your literature to create their own children’s book mimicking all aspects of your farm and sell it for a fraction of the cost of your book?
I do hold a hand a handful of trademarks. Like I said in the video, impersonation, trademark theft, forgery, etc. aren't OK. But I actually see a big distinction, between those actions and data scraping for learning. Think of it as the difference between a person reading your book and reusing ideas for something new that they create vs. making a forgery of it. One is OK, one isn't.
@@morgangold I think it’s a case of “you can’t have your cake and eat it too.” On one hand you want to protect yourself individually from infringement, forgery, etc. by way of trademark to protect your IP. On the other hand you’re suggesting leveraging other existing content for your use, whether that’s for art, ideas, inspiration, and so on. I’m not saying you’re right or wrong, as always I think context should always be accounted for, but for a topic such as this once Pandora’s box is opened we can’t pick and choose what we find appropriate.
The learning of AI doesn't parallel an actual human's, because AI doesn't really come out with its own style; it's just a jumbled up energy of all the styles it studied. The real originality comes from a particular individual's own experiences, thoughts, etc. They can have a bit of inspiration from other artists, but what makes their unique style is what makes it original.
Because a machine, that can analyse a million times faster than me and outperform me in everyway, and can develop faster than me. That allows some middleschool dropout who's stoned """makes""" a prestine piece that many could only dream to make. That's just a bad play on my side. Like it seems like the biggest fuck you to artists. It's not just a matter of amazing technology and meeting profit margins. It's actually fucked and it makes a lot of new artists nit want to even try. Also. "Ai artist" makes me want to vomit. Like that one ai "artist" that tried to diss a one piece animator. Also. Most advancements of this scale are meant to adress issues thst society wants fixed. But the only problem ai art fixes is cost for big companies. Wow, thank god we can make life easier for them and cut out the reason for them to hire us so we can all not be forced to either be mechanics and software engineers ior unemployed. You talk about the economic impact. But the ethical impact is just as important. I feel like ai supporters forget they are humans, and forget what community is.
With all the kindness and respect I have for you, I can't watch this one. As a digital artist who has worked so hard to be able to support myself, being heavily disabled, it pains me seeing people I respect go the AI art route for things that make money. So many sites and things artists have used all our lives at this point if you're in the 30-ish millenial bracket like Deviantart, it just AUTO Opted people and their galleries into AI learning. It lost SO much art to deletion, people protecting their work.. a moral dillema for SO MANY, whether they want their art stolen, or removing it from the internet. I've used AI art to make memes and stupid little things with friends, I think that sort of thing is a fun use, but the moment it goes into anything past a meme.... ethically, it just sucks. The IDEA of Ai art.... SO COOL. The way it's been implimented, and how it hurts people now.... oof. I hope shit gets better, the way people are so comfortable using other people's work now is hard. It's not hurting people who already commissioned us, it just normalizes stealing and disprespecting artist's work on a grander scale that... probably isn't going anywhere GOOD.
If you traced someone else's art and made minor tweaks to it for personal use, that's not theft. If you sell that art without license from the artist, it's theft. If you truly transform the work, it's not theft. But Lensa doesn't always really "transform" work - sometimes I even recognize the underlying source image. Singular. That feels more like theft. You as a person can make a judgement about what is "transformative enough" to no longer be stealing. That judgement so far isn't being programmed into the AI generators.
Could you explain how that's theft, when no physical property has been taken from the artist without consent? And what would the implications of that standard be for the internet as a whole, given that all information seen on it is a relatively free copy of information, including any copyrighted works one views?
Also this is the perfect channel for me as I love hearing people talking about things as I find it interesting to see how people think 🤔 also I am autistic so it helps me communicate with people so I am becoming less stressed
Decades ago when the first home computers came out, my dad's boss said, "This is going to enable a lot of people to create a lot of sh*t". He was right. A lot of the learning vs. theft debate boils down to "how many angels can dance on the head of a pin?"-type questions. Legislation and court cases can set some guardrails, such as in copyright disputes, while the world adjusts to a new inevitable AI-driven reality. But those guardrails won't hold forever and in most cases they'll be shutting the barn door after the horses have long gone. We're already seeing that in other controversies surrounding AI use. For example, when an AI commits a war crime, who do you drag before the Hague?
The makers of art AI’s use the artists art for free, to train their AI, so they have an art AI they can then make money with. Without the work of the artists used to train the AI the AI could not exist, so taking their content/labor for free, then profiting from it. Which to me personally feels kind of like theft. What is keeping people from paying AI a few dollars to do artwork in the style of their favorite artist, instead of using artists? I know you said you don’t condone that, but with this tech out here what’s preventing that? As corporations continue to try and squeeze more profit, with this tool available, what would be stopping them from this? I think that’s why people are concerned. I think there are so many good uses for AI, generative fill like you showed is great. I’m also not against individuals using AI to help express themselves, I think that’s great! My problem lies mostly with feeling like artists should be able to opt in and be somewhat compensated for their art being used to train AI, without their content it wouldn’t exist after all. It’s the corporation profiting from the work of the everyday man again
that can't be generalized. unless you know how the AI was trained you also don't know if the imagery were in some way illegally obtained. Most images are on google nowadays. Everyone can see them and imitate them. I know it is harsh but that is reality. You wouldn't complain if a human learned to draw an artstyle by googlesearching.
@moujayay Being available on the internet or viewable through a Google search doesn't implicitly give a company the right to use those images for AI training. Licensing and copyright still exists and there's a reason OpenAI is being sued for using licensed content to train its large language model. And no, you're right, the average person wouldn't complain about a human using licensed images to learn an art style, but they're also not doing it at a scale where they can process several thousand images per minute. Humans also have the ability to modify that art style in creative ways, which AI isn't capable of. There's no creativity there, just remixing on steroids. Morgan brings up the point of democratization of content creation tools, but AI in its current form is only going to consolidate the wealth available for content creation to those that already have the means to train, iterate, and deliver on their AI models, which to me is the opposite of democratization.
But scale doesn't necessarily define copyright law as it exists today. I see the distinction you're trying to create, but that's not how laws define things today.
@@morgangold even if we think worldwide there are great differences in defining copyright. as someone from a EU country it is pretty clearly labled for us but not nearly as strict as in Japan as example. Other countries don't even care about copyright like China. Especially seeing this as a global evolution thanks to the internet if can't simply be defined by US Laws. Yes, there need to be some rules to follow and often it is the country's rules where either the culpid lives or the demander sues from. But that is often quite hard to get through with. And THAT is clearly not for us to decide upon and has way higher consequences.
The artist of Deviant Art were informed that all of our art on it's sight was being used in AI to give it it's database of art. There was no way to opt out of the theft what so ever. It is theft. @@moujayay
LLM “Large Language Models” is the core of Machine Learning reaching to the Artificial Intelligence era. By its very nature it uses large data input to structure thesis and create an output. Like any tool in art or anything else that is using AI which is used heavily by the financial sector could be used for good and bad, hence a safe railguard and frameworks are required so we won’t end in a “Skynet” situation.
I don't think your statement reflects the nuance of US copyright law. Just because a person bought a newspaper, it doesn't give greater rights or protection vs looking at something online.
@morgangold Morgan, thanks for the reply. Sadly, I understand both patent and copyright law more than I care to. Also, my point is with the "greyish" area of using copyrighted work to learn from, and not just the kaw, but the ethics around it. You practiced your art on artwork that the artist was compensated for, even if minimally for any day's comic. Generative AI, on the other hand, generally just scrapes the internet without any compensation at all. What you did was perfectly legal. Especially as you didn't try to profit from the artwork you traced out, or the ones you drew when you improved. Generative AI systems absolutely profit from the result of their scraping. THAT is the 1st big issue, and I think there could/should be legislation about that. The second issue is that fee-for-service artwork has dropped in value. Why pay an Etsy artist $50 or $100 when I can just spend an hour or two and get something "good enough", especially when it's free or nearly so. I know a couple creators personally that are feeling the pinch. Consider me "Syndrome" and the actual artists as Dash Incredible. When I'm able to do "special" work, then those that truly ARE special, are no longer special anymore. I do chemistry in a lab, professionally. I deal with patents regularly, and thanks to other work, I've had dealings with copyright. Both systems are broken, but I believe at their hearts, both systems are necessary, and corporate profitization of them is killing the little guy, (heck, even "little guys" that make and sell $5MM annually are functionally small when compared to the Apples, GMs and ExxonMobils of the world) Watch Kyle Hill's video about AI scam science channels on RU-vid, and tell me that pandora's box isn't a massive threat to real creators like yourself. Anyhow, cheers!
@@jmacd8817 It seems like your understanding of copyright is largely emotionally biased in favor of artists. Generative AI scraping the internet is only possible because the art is available for free, that's how the internet works. Everything seen on the internet is just a copy of information, including scraped artwork, and explicitly copyrighted works. Google Images would be a far more massive plagiarist than any AI art platform ever made if the same logic were applied. Generative AI does often profit, but keep the data storage and outputs in mind. It stores almost none of the original information of images, to the point of being unrecognizable. All of its outputs are also unique, and fall under being transformative. Morgan mentioned the impacts of new technologies on people's work, that's in the video. Of course it'll impact people, but that's not going to stop it. He also mentioned he'll have to compete with channels that are largely AI generated, and that he is concerned by it. Did you even watch the full video, or did you just comment very early? Also, to use your own standards where the artist was paid with the newspaper comic example, what if he gave it to a friend for free, and his friend got to trace for free? Is the friend suddenly a thief while Morgan isn't? If not, take compensation out of the equation, because it's irrelevant. If the friend is a thief, then congratulations, we're all thieves, including Morgan since he didn't pay for the newspaper, his parents did.
I am very late, but in defense of AI art, it's innocent until proven guilty. Lots of accusations are thrown its way, but very little argument and evidence is provided as to why it's guilty while humans aren't under similar circumstances. The thing is, if AI art is theft, so is a 5 year old drawing Spider-Man, but I don't see anyone wanting to throw 5 year olds in jail over drawing Spider-Man, and your tracing example also works. Many artists apply a double standard to AI art to deem it guilty until proven innocent so it can be torn to shreds and rebuilt from the ground up with "ethically sourced" training data, which is often just them wanting money for art they've made publicly available. 20:18 "I believe if it's ok to use chainsaws even if you're opposed to the actions of Leatherface." This is such a wonderful analogy lmao.
I'm a retired graphic designer. When I started, my armament included a drawing table, a T square, magic markers, a photostat camera, rubber cement, and the all-important X-acto knife. I made regular use of typesetters, studio photographers, and a pre-press house. When I retired, most of those support jobs were gone. I was sorry to give up using Adobe Creative Suite programs all day every day. The future you outline is inevitable. Jobs disappear. People will find new ways to be creative for a living if that's what they're good at.
My line is posting full AI images and not marking them as such. I comprehend why you'd use them, if you just really really want to (I cant stop you). But they are problematic for actual artists, and steal from said artists as well. Tools are fine, and I'm personally even fine with AI being used for certain effects (animated lips) or like you showed, to edit actual images. But when the entire image is AI generated (and data scrapes from legit artists), and meant to replace human made art...it doesn't sit right. Never will. I like your videos, I usually like your opinions too, but on this topic? AI "art"/images in the way you often use them, especially when left unmarked (not including a note that they are AI generated), is problematic and frankly sad :( from an artists perspective. Not to mention they look uncanny and strange most of the time. Always tipped off by just that soulless, weird look the images have to them. You can overlook it at first, but most of the time it becomes obvious upon a slightly closer glance. Convenience seems nice at first. I think you head down a dangerous road with this one in particular, though. I know that sounds empty, but I usually am very much open to new technology and such. When it comes to AI images claiming to be art, and stealing from artists though...no. I don't support it. And I'm surprised you would, considering your opinions on other topics. And it comes down to it's easier for you. Which...if it being easier makes it worth it to you, so be it.
I think what gets lost is that AI isn't creating anything, it's just grabbing information from a database and then presenting them in a way it thinks the end user wants. This isn't a bad thing. It's not going to replace the high end creators like singer song writers, film makers and actors making master pieces. It may replace a lot of the filler content that we put on in the background. Like bad reality TV or daytime soap style dramas. It won't make the next goodfellas because AI has nothing new to say. I've also seen AI used in very creative fashions. People using it in ways I wouldn't have imagined or known how to do. it's not as simple as entering a few commands and it coming out with something like this video for example. It is killing a lot of entry level jobs. There are people who have Etsy stores selling wedding invites that are entirely run by AI. 3 years ago that would have been a student fresh out of college doing it as a side hustle or even their main income. The main problem with AI is that the people who created the content that fill the databases that AI draws from aren't being compensated. For years Facebook and the Bird Site have used for free the content created by traditional media to drive engagement that fuels their ad revenue without having to compensate traditional media. The unions picked the wrong line in the sand to fight at. They should have been fighting to ensure the huge corporations that sell these AI services have to credit and compensate the owners of the content they draw from.
I love the farm videos but this one breaks my heart. The "democratization of art" you speak of rings hollow to me when so many people who have trained their whole lives to produce art that was used to train these models without their consent or compensation struggle to find employment, or have to settle for even lower wages than what they already get using a program that probably stole their work. As shown by the recent writers' strikes, studios were already gunning to cut costs by generating first drafts with AI then hiring writers for a fraction of their wages to rewrite the generic schlop into something usable. How is it more "democratic" when using such software requires so much monetary investment or when the people using it disrespect the people whose work its derived from? How is it democratic to require artists to bend the knee to using this software to earn a living, especially when "premium" is behind a paywall? How is it democratic to make original works less accessible because people are removing it from the web in an attempt to guard their livlihood? How is it "democritizing" to make original work and resources harder to find as search engines become flooded with derivative AI garbage? Countless, COUNTLESS tutorials are available for free online for any number of mediums, traditional or digital, along with low-cost or FREE software and other materials. These generative models flatten the creative process. In creating an original image, traditional or digital, you have to - to some extent - understand why something looks the way it does. You absorb references and learn to appreciate the details of everything you look at and train your hand to delicately recreate them in some way or another. When studying your favorite artists, you come to understand their process and it becomes part of you. With generated images, you don't have to understand any of that - you don't even have to understand the model. You don't learn anything or grow creatively because you're off-loading that labor to the model. An apt analogy for generative art is like placing your order at a restaurant. You can read your order off of a script and send it off to the kitchen, and as you say, the more generic at first the better your return and maybe you add some specific instructions when you order again later on. YOU didn't make the burger. The chefs in the kitchen did. YOU didn't make the recipes that lead to your meal; the kitchen took your order and interpreted it with the recipes it had available. People claiming generative images are "art" are like someone ordering a delicious burger from the drive thru and claiming they made it because they placed the order. I'm all for goofy little non-commercial uses or making tedious repetitive tasks easier, but so many people using AI for "creative" purposes seem to think that the creation part itself is the tedious thing that needs replacing. It makes me feel sad and misunderstood as an artist to be reminded how much people take for granted all the work that goes into creative works.
Given that AI art will make the creation of art easier than ever, it will actually democratize art. Artists who've spent their entire lives getting better will still make art, and if they adopt AI then they will be the ones at the top of the AI art food chain, as their artistic skills can be translated to better products. Also, some of those "How is it democratizing" questions are just category errors that aren't related to democratization, like artists choosing to remove their works from the internet. With AI models and the creative process, while there's far less to learn with traditional art theory, there's still plenty of room for growth, learning, and alteration of the model and individual images, including using digital art and/or photoshop to alter pieces further to achieve a desired result. I think the "Who gets the credit?" question is tricky, but humans can be used by other humans and largely lack credit. A counter analogy to your burger analogy is Terminator 2 being a James Cameron film, despite so many people being involved with its creation because it was James Cameron's creative vision being realized. While other people get credit, it's still considered a James Cameron film. The creation part itself of art is tedious, and if it can be replaced, it simply will. Lots of work goes into art, but AI means that is no longer necessary when it's not wanted. While many people take it for granted, it somewhat is when so many people do it, and for free, and when people don't want the backstory and simply want nice visuals.
Morgan please can you bring back farm crimes but this channel as the algorithm doesn't like people to change a lot but I did enjoy the Oprah one but it is an extremely unusual crime
I think some of the examples you give are tools to help you modify your own works and the other (AI realm) takes someone else's work and it being modified by a computer for your own use. It's one thing to use AI products to supplement your own works, and it is an entirely different thing to have AI use other peoples works to make something you had no physical or digital hand in creating. You typing words to generate an AI image is not the same as actually creating something that you will then modify using AI.
So I’ll often use image generation to create an image that I then include in one of my videos. I think that straddles the line of what you’re talking about.
@morgangold If you use an image generator that only uses Public domain images for it source then I see no problem using image generators (not sure there are any out there that does this). The problem is when it uses any and every image it find regardless of copyright or if it should even have access to them.
Thanks Morgan for making us think on this new channel. My motto has always been to take it all in & throw out the garbage. It’s helped me learn & grow for these 77 yrs.
@24:35 What does it mean to get the exact image that you had inside your brain? Do you really have that image inside your brain? The prompt seems super broad. Was it really a blue background with people wearing blue clothing in your head? I feel like this is one of the biggest lies of the whole AI experience. People keep saying this, but it doesn't ring true at all.
I realize later why some people would think this. AI image generators produce roughly what you expect if you prompt that model with that text. It is not that it produces what is in your brain. Your brain thinks of a image that you can produce with an image generator. A human wouldn't be confused by prompt weight on blue and then make the cloths, eyes, and background blue. It is almost certainly not what you had in mind, unless you were restricting yourself to what the model produces.
I start by imagining creepy family in a Sears portrait from the early 80s. I keep making requests and variations until I get there. I've found that broader prompts and narrowing in is a way more effective than trying to be hyper-specific out of the gate. The image generators can anchor too quickly on specific words and that makes it harder to get what you're looking for.
Awesome video! I learned a long time ago to not bet against technological progress. Everyone needs to be able to adapt and overcome...including artists.
I'm sorry Morgan, I love you and respect you, but generally speaking, AI use by a company or celebrity rather than paying an artist only makes them look both morally and financially bankrupt.
I have great sympathy for artists who's art got illegitimately got used but as soon as an image lands in the google library it is more or less lost nowadays. I know the training of AIs is a bit shady for most people that don't have the insight but it is happening and we won't be able to stop it. In my opinion the using of a fully trained AI is also an art in itself as it is not as easy as some might think. Knowing which keywords lead to what and using different AIs for specific jobs (hands, feet, faces and so on) requires some amazing imaginary skills and fantasy to put all that together in a way to get the result you want. I both fear and await the growing of AIs but I don't think art or original artists will die out by it. either it shifts of they learn skills that then are getting even more unique and rare to have so they are more sought after.
It varies from person to person. I've seen people who'd be ok with it if it didn't impact people's jobs, and I've seen people entirely against it for various reasons that apply regardless of money. Using it for references seems fair game to me, though I am biased as I do that.
You are correct. Argument of morality of AI learning through art is obviously void. People are just worried about their jobs, because unless they adapt, they will be replaced.
It’s really how the Terminator world began. It’s too late we have let the genie out of bottle and nothing can be done to stop it. Kid’s spending major bucks for college then four years later those jobs are obsolete. Unhappiness from losing the ability for people to work will cause chaos but this train has left the building.
I am loving ALL your videos!!! I usually just listen to just music but watch yours too - regular GSF and now these! What about AI with HR and talent/recruiting??? A PERSON used to manually look through resumes... now these systems AUTOMATICALLY look at the same resumes! Not art, but I feel like this totally relates! ALSO I have been saying for the last few years - MOVIES - no one has NEW ideas, just do sequels or prequels or redo! It's ALL CRAZY!!!
Interesting, thought provoking and controversial. There is a huge cultural and legal lag with AI in the US. Also, Morgan, you had me at Walter Benjamin. I miss the thinking and learning part of academic life. My 1988 grad school video thesis was titled "Post Modernist Thought and Professional Wrestling."
This was definitely interesting, I definitely respect your opinions about this. I've definitely heard of AI before, it's good to learn more about it. I've never used it though. Btw it's nice learning more things about you.
I use a particular one that takes pictures I add in, and it turns them into anime style versions of the real thing. I never use other’s pictures, but rather use my own stuff.
As a complete moron about this stuff, can someone answer how ‘people’ get away with using ‘Joe Blo’s’ image and voice in yt shorts, without them getting his permission? I felt dumb not knowing it was a computer image, then when I heard it mispronounce a simple word, I looked further into it to discover it was a computer generated audio/video. How?
There are several different laws that give a framework for how to legally show someone's work. For instance; parody, using less than 8 seconds of audio or video, and news reporting. Parody gives people a lot of leverage to make funny takes on someone's content such as Weird al Yankovic and SNL.
@@DollyDagger0519 thank you. But, say in the instance of a famous ‘Joe…R@gan’, that’s using his artificial image and voice, spewing bs…. Can they do that? Or is it up to the ‘Joe’ to stop it/sue them?
The makers of art AI’s use the artists art for free, to train their AI, so they have an art AI they can then make money with. Without the work of the artists used to train the AI the AI could not exist, so taking their content/labor for free, then profiting from it. Which to me personally feels kind of like theft. What is keeping people from paying AI a few dollars to do artwork in the style of their favorite artist, instead of using artists? I know you said you don’t condone that, but with this tech out here what’s preventing that? As corporations continue to try and squeeze more profit, with this tool available, what would be stopping them from this? I think that’s why people are concerned. I think there are so many good uses for AI, generative fill like you showed is great. I’m also not against individuals using AI to help express themselves, I think that’s great! My problem lies mostly with feeling like artists should be able to opt in and be somewhat compensated for their art being used to train AI, without their content it wouldn’t exist after all. It’s the corporation profiting from the work of the everyday man again