U can use IA and create variations of an image and then you do a composition using photoshop for example and create something new, thats is possible, I dont see not ilegal here. What do you think?
@@LeandroT242 i agree with what you say. The only thing i would add however is not being allowed to just add leonardo style at the end if you want the copy-write. He made his own style and it took him a lifetime so why should we be able to use that and claim its ours just because AI can copy it? In my opinion there should be a combination of styles with some aspects of Leonardo for example where the AI artist put in a lot of work to make the work unique instead of a straight copy of Leonardo.
Use your brain to discover ideas. Learn how to write. Learn how to draw. RUclips has all the resources you need to do that. Then you become an enriched person who has achieved and own everything every time. Problem solved.
@@doctormelancholy3042 It depends if he stole that blueprint or not, from hardworking blueprint draftsman and did they even ever gave him permission to use said blueprint or not. Though I doubt it because he’d be too cheap to pay with time and effort.
@@doctormelancholy3042 There's nothing wrong when they just build the cabin for themselves, there is something wrong when they start to sell it on an industrial scale and the design isn't theirs.
Hmm. Little late to the vid. But those first images from midjourney, when it started, were eye opening. It made me think even more creatively. To see how a machine puts images together was thought provolking.
The fundamental problem of generative artificial intelligences lies in the initial violation of the rights of the authors of the material used, in its "training" since in reality the images now encoded remain in its model, (this explains why companies offer the possibility of extracting them if you have the copyright) Therefore any image generated by the AI is always based on previously loaded images, if it does not have them it cannot generate them, therefore it does not learn copy. In short, the authorship of a user in the generation is minimal, only the idea, since with the same propmt the system generates very different images, like a lottery, and the ideas cannot be registered by themselves, they must belong to a work, and here the work is not human.
It isn't any different if it's AI or a human. If an "artist"" believes their copyrights have been violated, they can make a claim to the kangaroo court system to see if they have enough to payoff the judge. Once you publish your work for others to see, in particular online, you take a chance. There is also "fair use" and anyone can trace an outline of the picture you took of your rear end and posted online and add to it and color it and then even you might not see that they borrowed your stuff-If you catch them, call a lawyer. Otherwise F**k off!
@@marcelootero6866 So you are calling video game developers mediocre people without talent? All while your avatar screams "mediocre and without talent"?
@@doctormelancholy3042 Its completely different if you knew anything about how AI works. A human does not dissasemble 20 million pictures into pixels and t heir location and then reasembles avaraged out aproximations of where the pixels were. You are full of shit m8..
@@doctormelancholy3042 Yes overwhelmingly, why do you worship game developers when the majority of industry struggles to make anything interesting since years? Also nice personal insult of the guys face, really shows you are the smart one...
It helps to call what is generated by AI differently from what is created by cameras, artists. I use the word 'synthogram' as that means the output from a synthesis of elements (tensors from the learned image manifold).
@@BenCaesarIf you're not creating anything de nuovo, AI enhancement is what I'd talk about (auto colour is not necessarily AI. If it's adaptive and responds differently to different types of scenes, maybe). Synthograms come from AI models responding to prompts.
@@TomAngPhoto thanks for the clarification! Lightroom has some preset suggestions according to the image this might be more the Ai you're talking about. It's adaptive than the preset library.
Thank you for explaining! I can imagine how complex this topic will be in the future, especially when text-to-video is so improved that it can be used for actual movies.
Thank you for this nuanced discussion. This nuance is what is usually missing from the conversation but it's everything. AI in the hands of an artist is a tool and the end result is usually the last step of a thoughtful multi-step process.
So in other words, u want to kill those ai. Machine learning needs examples to work from, and I hate in when people say it’s theft. It’s like going to a museum and looking at art, making notes then going back, every day until eventually u can make something similar. That’s not theft, and neither is ai
Interesting. Recently, I plugged a few of my photos into Midjourney to create similars. I wonder how the copyright office would land on the output. I can bring them into PS and use its tools to do pretty much the same thing, albeit with more control. That said, if I wanted to use the new PS generative tool to change elements of the photo, still a clear copyrightable work? Because of my photo + PS? It’s all so clear now!
I was thinking along the same line. l use my cartoons as the basis for AI art. It enhances as I could in graphics programs. If I copyright my original image and load it and I'm giving permission to AI to work on my work is the outcome still under my copyright?
Someone told me they did this and showed her process to the copyright office and they approved the copyright but she did it with her digital paintings. I don't know her w it will work with photography.
It's an easy thing... If you draw your OC Character, then put it onto AI Art. You still own the Copyright for BOTH because your OC is grandfathered in as that Copyright IP existed before you put it into AI Tech... ... ... Spider Man by Sony is a good example. If you did 50/50 mix of Human and AI generated art etc. Then you may copyright the entire thing. Since all AI Art is case by case and not 100% Public Domain. If you did small edits but nothing major to alter the raw output of AI Generated Art, then you don't get a copyright for it because you're not the author in the legal side.
20 years ago we were talking about having to think about human rights for AI sometime in the future Now we are debating how to defend our own human rights against AI
This is interesting, and clearly just the beginning. I think "human authorship" is too vague. When a movie is produced, the copyright to that movie typically belongs to the movie's producer or production company. The producer is usually the individual or entity responsible for overseeing the creation of the film, managing the financial aspects, and making key decisions during its production. It's the same for images. AI is just a tool of creative expression. Maybe more powerful than other traditional means, but still it's just an advanced tool. AI generated movies will exist, but the copyright will belong to the production company/producer. This discussion is far from over.
Very helpful. I appreciate your stating legalities without “choosing sides.” I’ve been involved in photography, painting, and sketching. When computers came along, I learned how to use but-mapped program, Freehand, Photoshop, fractal programs. Most of the later work was in Photoshop, but I ventured into Deep Dream Generator. Then I hear about AI (DALL-E and Mid Journey). MidJourney can create awesome works but the artists’ intention plays a big role. I can make dozens of prompts in search of something close to what I imagine. I’ve created many images in which I combined 3 different AI images in Photoshop and then did a lot of work in Photoshop. It’s a journey. Sometimes AI is pretty dumb...7 fingers and 3 arms, limited ability to portray complex emotion. In terms of copyright on occasion I make a very simple prompt because if I don’t use a long prompt, it allows MidJourney to be its most creative. With those I definitely could not claim copyright. I did submit AI art to a magazine. They told me they had not worked out a policy yet about publishing AI art, and I understand that.
copyrighting is not to protect you. its to push the industry forward. the whole concept of an idea belonging to you without pushing the industry forward is imoral to me.
I think protecting the human contribution is fine. AI only creates images or text; it's humans that place those things into artistic contexts from which a human audience derives meaning. The images themselves are now fungible. There's not even really a need to copyright the image; someone else can just use the same prompt and generate their own, none of the artistic value is created by the specificity of the image.
Cope more. It is stolen images. Based on actual drawing artists. All you ai boys do is type some words. I tried stable dif. It's NOT THAT HARD. Try drawing it from scratch ones in your life.
What if you train the AI using your own artwork and original characters? In this case, the AI is not producing images from scratch based on verbal prompts. It has been trained on visual materials, which are the artwork and references that the artist themself created. There is significant input from the artist, and the AI is just reproducing the artist’s art & original characters by putting them in different scenes.
@@ellenripley4837 can't say for sure honestly, even though the input is yours it's running through a dataset of other artists work. Copyright office may not issue a full copyright. If the dataset is trained on your work I think copyright will be easy
I could use your insight to help clear up my understanding on how to properly release an audio/visual project that I have just completed. I wrote, recorded and mastered 9 entirely original synth-wave songs. I used OpenAI to generate and edit 30 minutes of prompt-based visuals that I lined up to the beat of my original music. I'd like to release this as a single audio/visual, copyrighted piece of work on RUclips. Do you have advice on what I might need to consider in terms of copyrighting and monetizing this project? This is my first major release. Thank you Samson. -Carlos in Colorado
That song is copyright able in America because you wrote the skeleton for the music etc. Make sure to send a Copyright Document (depending on your State) to the Copyright Office for legal securities since RUclips is known to be risky with fraudulent DMCAs. So you probably want to publish your music after the legal copyright documents are filled.
With a camera you have much greater control over the creative outcome, which seems to be the aspect they focus on. But I agree that there are somewhat arbitrary lines drawn. What if you use a self-shot photo as a reference for the AI for example? It's way too complex and they don't seem to grasp all the possibilities.
@@missoats8731but does one own the tree or the grass? Does one put the creative elements into the scene if it is instead Camara and software that does the vast majority of the work? More importantly, arent there other things like Disney's decades long copyright stuff that might be worth reviewing. Or old games that are no longer being produced or sold still out of public domain for decades?
@@missoats8731 Questionable actually, since with a camera you only have choice/control over the immediate area in your physical presence, whereas with an AI you can specifically control the subject of your output to be anything you can imagine. The amount of difference in human control here is orders of magnitude different. Similarly, with photography, you get lots of unusable shots, and make choices as to which of the shots you photographed constitute a final artwork, much as an AI artist picks and chooses which generation out of potentially infinite choices they consider to be the final artistic product. This new "law" or precedent is flawed in many ways, and I expect they will have to amend it over time simply because they are under valuing both the level of human choice and control in prompting, and the future capabilities of ai tools to give the user greater control than many of the current technologies provide. There is a stable diffusion tool called control-net which allows the user to pose a subject control facial expression, layout the composition very specifically. These are even greater levels of human input being put in, and should not require the artist to go in and manually edit the final work just to consider it original human created content.
@@toren9120 You're taking this argument in a direction that isn't at all based in the copyright precedent set by these legal findings. The only thing that is being considered for copyrightable artwork according to the video is how much control did the human have in the process. I argue in the case of AI art that it's actually quite a lot more, or at best nearly equal to that of a photographer, and it's a sliding scale like any process, you tend to get more out the more you put in. The difference in the technology that is responsible for getting a photograph onto film stock or into a digital file, compared to having a text to image tool produce an image is significant, but the process that a photographer takes, in choosing a subject, framing a shot, determining various settings, and selecting what shots constitute a finished product are so similar that the differences are mostly symantics. I'm confident that many copyright worthy photographs would constitute far less effort or forethought than a lot of the AI art that I've produced. Your argument above is identical to the argument that a painter or sculptor could make towards any photographer in fact. And yet the law does not preclude works of photography from copyright. The original comment made by Christian Pedersen, is right on the money. If you want to ignore that, disregard, or whatever (and I'm sure you and many others biased against ai art and artists will), that's fine, but the law set out according to the video is wrong headed, unless they want to apply it equally to other forms of art, photography specifically for this argument, rendering all such works no longer able to be copyrighted (and that aint happening).
It is important to assume that those making laws know less about these materials than you do. Remember, our nation is primarily ran by lawyers not nerds. So with the information provided, is it possible that one can assume the copyright of the given prompts since those are authored by the artists? How about if I drew something very crappily, being the treditional element of the authorship, and I improved it via AI artwork? Is it mine still?
This is a very important point! I suppose you would just need to hide your trail and you'd be fine. There is going to be a big problem with this in a lot of fields
Firstly I think this whole law is silly and uninformed But perhaps designers would be expected to present a proof of their work process. It would be relatively easy to implement a sort of timelapse feature as for example adobe fresco has, that shows your entire work process from start to finish.
@@aisamsonreal as someone who utilises AI images, there are consistent elements you'd notice if you looked closely for them... but who's to say if they'll still be there in a newer version, or, like I do, I retouch images to remove traces of the imperfections - so then if there was a bot that found AI generated images, would they be able to pick up retouched AI images?
What about people that throw paint into a machine that splatters it around the canvas and then sell it without any further steps? Do they not have copyright anymore?
@@Bartetmedia I love that people keep saying this because it instantly shows they have no idea how the AI actually works. The AI is nothing more than a mathematical function with billions of variables that have been tweaked based off of billions of inputs. When an AI is "trained" on someone else's art, it doesn't copy it or store it or whatever you think it does, it simply slightly changes one variable by +0.001, another by -0.0004, another by .009 and so on. It's literally just a big equation. It's no more "stealing" it's training data than a photoshop artist is "stealing" the colors from an image by using the same exact RGB value as someone else. It really is incredible how so many people have no idea how the technology works and yet are so sure that it's unethical.
Thanks for watching! What topics on AI would you like to see me cover next? And if you have any thoughts on how to make my videos even better for you, please let me know, I'm all ears. :)
Thank you for the depth explanation! I suggest for the next AI topic is concerning the machine learning and the relation to the current lawsuit copyright infringement. :D
copyrighting is not to protect you. its to push the industry forward. the whole concept of an idea belonging to you without pushing the industry forward is imoral to me.
Important distinction to note, is that your video would only apply to those who are based in the US. US copyright laws do not extend full copyright protection for computer-generated work, unlike other countries where they do. That, ultimately, is why a lot of these platforms still make reference to commercial rights and ownership to prospective users, when using their tool to create art.
The funny thing is, as an artist with nearly 20 years of experience in comics, I don't believe that Zarya resembles Zendaya. Zendaya's eyes, eyebrows, and forehead have distinct differences compared to that fictional character. For artists skilled in capturing likenesses, we understand the significant role that eyes, eyebrows, and forehead play in drawing. Although the images may not be of high quality, it's worth noting that Zendaya's chin is more prominent. Furthermore, aside from their shared Afro heritage, they do not bear a resemblance to each other. Not to my artist's eyes. Yes, as an artist, I utilize sketches and establish the composition before finalizing pieces with Stable Diffusion + ControlNet. While it may not be suitable for various tasks and lacks the necessary AI capabilities for serious commercial projects, it does provide valuable insights into areas I can enhance in my final artwork.
I got a question. Under these rules how exactly is a photograph a copyrighted work since all the "artist" has to do is press a button and the camera does all the work? where is the human authorship there ?
@@toren9120 and what about the people who do not chose whatever lenses you mentioned and just pull up their Iphone from their pocket and take a photo. They don't need to choose a perspective or anything. Is no different. You could even set a timer when the photo can be taken and leave the phone. That is still their photo and copyrighted. So all your points you mentioned are not needed in order to have photography copyrighted. Also give prompting a go, and see how 'easy' it is. It will take you at least a good few days until you learn to prompt like an amateur, and longer to actually get something good.
If you are specific enough in the prompt where you already have the idea in your head and are simply directing the AI to do your bidding, then I think you can claim ownership.
It's because it takes skill to take a photo that looks good. Why do you think photography is a career. People get paid to take photos with their equipment for a client. It is utter lunacy to expect the same treatment when all you do is type in a prompt. I've used Dream AI in the past for legitimate reasons, mostly for fun. I'm also an artist and I never thought highly of those generated images, just that they looked cool and maybe I could take inspiration from one of them cus it was really unintentionally abstract. Instead of typing words into a prompt. Why don't you: Type words into a story do some doodles play an instrument program something cool Unless youre literally incapable of writing more than one sentence or you have no hands or the motor control of a dude with oven mits on, there is no excuse as to why you cant find something creative or fun to do or pay someone to do it for you. End of
@@Lambsauce10 " because it takes skill to take a photo" what about a photo that i just press the button without using any "skills". A bad photo at that. "all you do is type in a prompt" , "all you do is press a button" Don't kid yourself.
That means anyone can appropriate and sell the non-copyrightable images on tee shirts, cups, posters etc. In addition, the tech overlords can still take all your “iterations” to train AI and reuse the AI generated images the person “thinks” they have creatd
Sounds good in theory but how you will be able to verify who is telling the truth will be a headache. Unless they start requiring proofs but then the process will be incredibly long.
This is basically the law you have right now with public domain images. You don't own the rights unless you change them considerably to make them your own.
I`m curious about possibility of "style" protections in any kind art form . Is it possible ? How many actually new styles we ever see or hear right now ? Without getting inspiration from older arts ? Say some artist who can manually draw in canvas or using mouse and stylus but the final output could be have same style with another artist , how the law will applied ? Is that also means a song cannot copyrighted because drum or guitar sounds same with another songs although they playing different notes ? How about guitar playing styles which generate using same technique with another famous guitarist ? Or how the law can forbid for any musician using same presets and sounds ? Or we`ll having different laws between any form of arts ? Because in some art form like music , Styles could be outputting same sounds because they might be using same techniques
there's a freaking lot of new and unique styles, you just need to search for artists lol, anime and pixar is not the only art that exists, I think the law must be regulate the consent and the right for someone to just don't be part for a model...
@@morizanova wtf.. man, style is not "something that comes from nothing" style is the way you make something, take this simple example: if you feed AI with just the works of "samdoesart" AI will just make derivations of his works, but if you show the same works to a human artist, he will put it's own thoughts and intention, his own unique perspective of what he sees on those works, creating something new with his own authenthicity, every style born that way... AI can't make it because is designed to replicate and recreate the information of it's database, so it can just create derivations. did you know that the "anime style" was a reinterpretation of Disney's eyes and faces? you think that AI could create the "Anime Style" if you feed a model just with disney pictures? No!, it will just create more disney pictures lol. Ai is a doppelganger, the concept of "generator" and "discriminator" is literally tryharding to recreate an actual picture, but it can't so it spits something that the discriminator thinks it looks the same. That's why AI can't make something different of what's on it's database, just derivations.
@@Mente_Fugaz we talked about difficulty to copyright styles. Because you can easily trace the origins if you have vast of knowledge and origins Learn to read first before replying with same old AI cry babies songs which everyone already knew But let's ignore that , And talks about your arguments TBH ,I doubt you even ever doing deep dive trial error in AI tools , because your long points still not quite right . Right now You can easily mix some Disney with luis royo and Mobius art style and if not enough there are already thousand of specific artist style which build as mashup tools Standard fan art artist will make their copy as it is but creative ones will mashup those character fan art with contrasting style so people who have limited knowledges will thinking that new styles And right now AI with LoRA etc can do that . Only newbies AI users will keep pumping one art style . The experiences ones know how create good blends
@@morizanova Copyright styles: you can't copyright styles, you can copyright works of art... and if you use AI, you are like a client asking for a comission to an artist, but the picture wasn't made by you, so you can't copyright it, despite how much specifications or sketches you shared to the artist in order to him to understand what do you want. Also a work that is not yours, you can't use it for anything you want, you need the consent of that artist to use it, you can just consume it, but when you use art that doesn't belongs to you, to train an algorythm to replicate derivations of it, that's simply unethical... I don't know how there's people that can't see the obvious thing yet. about merging styles, when you merge 2 styles, you think it's something new, but those merges have an origin on the dataset of the original stable diffusion models, if you try to merge styles, it won't have any direction, there are just glitched derivations. you sound like those who think that AI is magic or something... man, AI doesn't understand what is doing unless is on it's training data, the only way that creating styles with AI could be something, is in the case you train AI with those glitched derivations, tag them, and then you try to work with them. you could make a lot of interesting abstract art. But let's be honest here, what you are calling here " create new styles" merging algorythimcally a bunch of pictures, is just using something similar on the database that looks like a merge, even actual pictures merging those styles etc, and creating a derivation of it. if you see a "new style" as an output of AI , that looks good, is because is not a "new style" it's just using stuff on the dataset that you think it doesn't exists lol you sound like those who think that AI is magic or something... it's also hilarious to use the term "newbies AI users" when this technology don't have a single year since it was released lol people can become expert in 1 month or less, you just need a good pc and a expensive Nvidia graphic card... and being suscribed to a lot of twitter and reddit accounts that post new models and AI tools. Because the only real way to improve your art on AI is to wait until new tools and models arrive
Where is the copyright when I generate images and use them in a video like on RUclips? I upscale the images after generating, find a fitting music and create a slideshow, sometimes with storytelling elements 😅 If I would use only copyright free images (not AI generated) in the video, the copyright for the full video would still be mine? Whats your opinion on this? 😅🎉
In that example... In America, you own the Copyright of your collage of Public Domain Works but not the individual Public Domain Works used in your collage. You can copyright protect you Collage but not the individual Public Domain materials.
As a creative director, I expect my final pieces of film to be considered one artistic piece. Using AI just means that I can do the jobs of many in the course of days rather than weeks or months. Writing high quality prompts isn't something everyone can do. Fug the courts!
So, i can use Ai generated images to create my RUclips videos and i won't get copyright claims for using them. That's what i want to understand. I use Copilot to create my images but i'm still worries cos i dont know if i'm free to use them on RUclips or not. Someone put me through please
Thanks Samson. This is one of the topics that I -- despite keeping up with the latest -- have so many questions about... many of which you addressed here.
This is a tedious subject that I'm always exasperated with before I even begin addressing it. 1. I'm one of the few people that create and train my own stable diffusion models on my own artistic works (photography, graphite, procreate and krita) and merge them (or LoRa them) with other, larger models that are themselves trained on only royalty free content. Continuously creating, recreating and merging my work and these models together in different ways is like creating a damascus steel of your own creativity, continuously folding it in on itself over and over until you've created something that's so uniquely yours, but almost alive in a way. It is AI, but it's your AI, your creation, creating new things based on your visual and textual inputs. Inpainting, masking, outpainting, img2img, controlnet, ic-light, multimodel, prompt and seed migrations during animations etc etc... it's limitless. It's difficult to describe how satisfying, enthralling and, contrary to what some (understandable) comments here claim, how *genuinely human* this kind of work is. As you can imagine, from my cold dead hands would anyone get these models from me. They're everything. My inference machine is literally air-gapped from the internet. 2. Digital laundering of intellectual property into neural networks is exceedingly easy (as midjourney has so red-handedly demonstrated). This is the major crux of the issue: current methods of assessment by governing bodies rely on nothing but honesty. If I was to hypothetically create new work using the work of other artists, and I didn't want you to know; then you wouldn't. The ONLY way to genuinely prove that this work is 100% my own and legal is to provide the models, the exact repo of Kohya SS I used+it's training values, exact copies of software tools I use for inference, and most importantly - the datasets. Only by providing this kind of prepared ready-to-run data&software package (of my entire artistic identity and secret training sauce? 🤨not a chance) to a governing body, can they then launch Kohya SS, load my training json, train/merge/lora the model themselves and compare the generations using the same webui, settings, prompts, seeds etc.. to confirm authenticity. Only then would giving copyright to an artist using AI like me be truly fair and ethical. Which is a tough thing to admit to myself, but if our goal is to be intellectually honest and responsible in a way that secures the intellectual property of future artists, then sadly that road isn't an easy one. As usual. 🙄It's a mess. This all happened way too fast for us. But just like the internet, here I am in the barrel of the wave. Only this time around I've no naive delusions that this will make the world better. What an adorable mind I had back then.
“I’ll show you what you can do to Protect your work from unauthorized use and avoid infringing on other people's copyrights” As someone with student debt from art school this from the description is absurdly hilarious.
What if I got the image from the internet, modify it using AI, add more elements, filters and other modified AI images into one frame which makes it a completely different and unique, could it still be copyrighted? I hope someone could answer this. Thank you!
lol, what about my creativity in posting cool AI art that i saw someone made, and them posting it like i made it xD It requires creativity, because you need to know where you can find something you like
QUESTION: if someone takes an image and it is altered significantly by AI does it not cease to be their likeness? I believe it was found that if it is altered significantly or is considered sufficiently "transformative" it can be used or am I wrong? As an example: If I make an AI influencer on instragram based on let's say actress Emma Watson, but with a different name and an AI has generated an avatar inspired by Emma Watson... with significant alterations: changes to the nose, different eye color, changes to jawline, significant body changes, different hair color, etc--though someone may look at it and say it reminds them of Emma Watson or maybe it looks like her, it nevertheless has been altered into something else or transformed at least. And then that person uses the AI to then sell naked photos of this AI influencer...is that a legit cause of action or otherwise something to be sued over? They are not claiming to be her and the differences are significant.
@@CultofThings i dont care what it does, you are not allowed to use it because i didnt allow the use of my work by the AI. You essentially a moral corpse, happy to use a fence for stolen goods and your argument is "the bike is different color and reasembled parts of 10 different bikes so its a different bike now" Police begs to differ, and the entire civilized world and property laws speak against your greedy selfish thieving little excuse.
Why not accepting the fact all existed AI models are stolen and modified. Thanks to Stable diffusion, AI model is being trained updated everyday until now from unknown sources and persons.
the ceo of open AI told that there must be laws for people to have the right to opt out all their content from the databases, even in the socialmedia models, I don't know if with that, he was refering that there could be a way to opt out from the models at this point...
@@byujkt3822 I'll accept that when you accept that all art is borrowed or stolen-However you want to put it! The "greatest comicbook writer ever" Alan Moore came to fortune and fame writing Superman, a character created by Joe Shuster, which was stolen by DC Comics. Moore went on to claim his stuff was stolen by DC Comics as well. Live by the pen and die by the contract. Stolen or not, his Superman and Batman were better than most!
Creating through the symbiosis of artificial intelligence and human imagination yields products that are legally eligible for copyright protection. This emerging paradigm of collaborative creation poses novel challenges to the conventional frameworks of intellectual property rights. From a legal standpoint, the question becomes: how can the copyright office distinguish between the human element and artificial intelligence in the creation process? Indeed, this presents a complex puzzle. Given the inherent anonymity of AI and its inability to possess legal personality, accurately identifying the contributions of both entities in the creative process will be challenging. It's a task that stands to be extraordinarily difficult, and yet, it is a necessary evolution in our understanding and application of copyright laws in the age of artificial intelligence.
@@I-Dophler wait; so if that doesn't work, being that the artwork has a seed-number, & the literature has zero-plagiarism; what givez, bkuz I am LITERALLY in the present to present both ai-art, & lit. as I am writing this #againwhatgivez
@@pay_it_forward_franklin4469 The complexities of copyright law in the context of AI-generated art and literature are indeed challenging. The seed number you mention is a unique identifier for the AI-generated artwork, but it doesn't necessarily confer copyright ownership. Similarly, the fact that AI-generated literature is free of plagiarism doesn't automatically grant it copyright protection. The critical issue here is the concept of 'originality' and 'creative input', central to copyright law. AI is not legally recognised as capable of original thought or creativity, so it cannot hold the copyright. Therefore, the copyright status of AI-generated works is still a grey area in many jurisdictions. It's a rapidly evolving field, and legal frameworks struggle to keep pace with technological advancements. #AIcopyrightissues
@@I-Dophler Yea I understand, but wanting to give to those, whether at leisure, or especially in need; all I want to do is exhibit within the linez with no penalty for. So, for example; I put out ai art with literature for monetary-gain; what restriction(z), &/or penaltiez take place?
AI art is for posers, hacks, the lazy. Those without the mental fortitude or moxy to venture into a lifelong pursuit of passion with all the years of hardship and pain you knowingly sign up for, and still say its worth it. AI Art isnt a tool, it's an exploitation, a shortcut towards what i just described. It's for gamers not livers, the people on the outside of hard won skill that are too afraid to tempt fate for years purely for the sport of human endeavor. Ai art is for the weak. The losers. The excuse makers. People that have the time to sit around for idle hours and come up with well crafted arguments being edgy about it, bc theyre not busy developing a skill instead. Ai art is for the compacent and the unengaged, the lower level of sentience. People without that existential buzz within them that knows you can only live it or youre a sham, that difference between results and the work put in to arrive there. All that's changed is more people will convince themselves theyve achieved something creative when they havent, but deep down theyll never know what it actually feels like to risk it all on that passion. To truly EARN it. Becsuse theyve risked nothing theyve gained nothing. They wont know what it feels like to, despite it all, earn a creative skill and worldview. They will have committed the biggest crime, funnily enough, against their own person, their own humanity, and id say thats a fitting justice. Let these masochists of the soul writhe and die if thats the fate theyve chosen. Meanwhile real artists arent going anywhere. More posers will get weeded out by this and the more dogged, gutsy artists will continue to thrive in whatever they choose to do, bc theyre about that life.
I guess for me is - are you going to copyright the wood and the sand you used from the computer as yours, when you used it as part of your work, or the end overall results itself, a reproduceable outcome that's unique to your creativity?
USA Copyright Law for User Generated Content (UGC) has always judged on the Output of the user. Not the Input they used to create the Output. Services such as AI Generated Art Software DO NOT follow the same legalities. A company or individual can be held liable for copyright infringement, unlike UGC, if and only if they can prove that the Data Base contains their infringing copyrighted material(s)... Which is... Good luck with that against Neural Networks, under American Courts.
The Canadian Intellectual Property Office registration set a precedent that the office accepts AI authorship and has "led to Canada gaining publicity as one of the only jurisdictions in the world recognizing copyright in works ‘authored’ by an AI," the application argues.
What if Ai generated an image and I repainted it in real life using oil paints on canvas, including driving the visual styles of the prompt, will the final painting on canvas in real life be Copyrighted?
Isn't that the same as an artist painting on canvas the natural landscape? Nature is not copyrighted, neither AI. The painter holds the copyright to his work in both cases.
So, can I use AI generated images in Midjourney for commercial use? What about if someone will use my generations for commercial use? Is it legat to use someone's AI-art for commercial? For example if someone has a paid monthly ai art magazine and uses ai-generated images by others?
Basically, depends what you did to alter the original AI image, likely. I'm not a lawyer though. Straight-AI image uncopyrighted-altered images need to take copyright into consideration. If you can take an AI generated painting and turn it into an ink drawing which looks very different from the AI generation, for instance-deserves an artist cioyright.
lets say i make a SD model with my own paintings, and then use image2image with sketches, photos, 3D renders, to make new paintings, is that copyrightable?
great great video, i was talking with a neighbor about this since her husband is a draw artist. Her husband was sad about how easily some things could be done now and claimed as the work of someone else. I was like ok but if you go trough all kind of loops and making combinations of different styles which those artist couldn't even do themselves + some points you made in the video... why couldn't it be copyright? in my opinion it should. Just bluntly type a scene you want and add leonardo style should be not be allowed to be copywriter since you just copy the exact style of him. Also the hard work those artists put in to make they're own style should be respected and protected imo. I felt sort of bad by just testing some of they're styles in midjourney if iam honest haha.
Under USA Copyright Law. Art Styles are not copyright protected. I can easily recreate any Art Style 1 to 1 or have an AI do it for me. It matters not in the law. Only that... If I'm replicating Marvel Comics, that I don't make Iron Man because that would be Copyright Infringement. Most artists don't know a damn about USA Copyright Law in detail. I do, since it's a part of my business to know those details for video game development.
@@absolstoryoffiction6615 so how would you write an midjourney prompt to get features of ironman or a mix with him and another super heroes without getting ironman? Since you can use the art style. Just marvel artstyle?
@@MeesterGgaming Legally speaking... As long as you do not make "Iron Man" by Marvel/Disney when using AI Software or drawing it out yourself. Then it's legal under USA Copyright. Just don't call it "Iron Man" since that's Trademark Infringement and don't create any derivatives of Iron Man. You can have your own "man in a robot suit" even to the complex technological level as Iron Man but avoid doing as I stated above. (This is old legal stuff that AI Software never changed nor added upon, in America.) Art Styles are the same since Art Styles are not copyright protected but the Iron Man IP is copyright protected. You can certainly make your own Marvel / DC Style Comics but avoid using copyrighted characters. Avoid Fair Use too, since that legal defence won't work in this subject, specifically. Unless you're making political commentary etc. where super heroes were used in that manner most often. That's protected under both USA Fair Use Law and USA 1st Amendment.
As a photographer I don't mind at all when my images are used freely for personal use. I think it is rather miserable to hide something on a hard drive that might add to someone else's creativity. On the other hand if someone has a consistant habit of just stealing other people's work for profit with no conscience that should be a crime whether AI generated or not. The lawsuits are interesting. The stock libraries do represent a lot of artists who in general want their work protected but I hope it is overturned. I would not mind if elements of my images were used. AI artistic programs that have good intension should be able to use any element already created. It's a little bit like saying if you go for a walk don't take a photo of a rose in someone's garden. Let creativity grow and thrive!
It doesn't matter. Publication happens because individuals can't make movies. A "good movie a friend saw this weekend" can just be a prompt + seed. Copyrighted works will lose market share fast, because of copyright.
@@beth1979 AI art doesn't "devalue". It had near zero value already. Only human art devalues, until AI is all that's left. (BTW I mean this as a bad thing if that isn't obvious.)
@@jonmichaelgalindo the big problem is that, the devaluation of human art will stagnate everything eventually, because without new artists, the AI can't evolve. And if artists can't make a live with it, there will be less and less artists working full time on it, innovating and moving the culture... Until AI work ethically with licenced and public domain data, art will keep bleeding
@@Mente_Fugaz Public domain data is a hundred years old. It's stuff about what society was like before computers existed, before nukes and the holocaust and socialism. AI trained on public domain data is useless to everyone. And AI trained on licensed data will be controlled by the wealthy elite, who will force artists to pay through the nose for the privilege of using the tools trained on their work. Pay artists pennies to steal their work, then charge them an arm and a leg to use the product. >:-( No. The problem I see is saturation. Even before AI, we've been overloaded these last two decades with worthless, cookie-cutter, derivative wastes of time competing for our attention with the true gems. What we need is a way to identify and amplify the impact of artists blazing new, truly worthwhile trails in art. If AI is truly an all-purpose tool, maybe it can do that. An AI that can look at something and say, "that's not random noise, and it's something truly new and valuable." But then again, maybe not. But saturation is the problem, in my humble opinion.
@@jonmichaelgalindo i think, satuartion is a consequence, not a problem... The problem could be the lack of motivation to innovate... The inspiration to make something unique and fresh. The real saturation comes with AI , not with children trying hard to draw my little pony characters, Because a lot of new talents raise making that stuff, and in the future they develop their own vision creating authenthic stuff
As an artist, I am not the least bit afraid of the unstoppable nature of progress. In fact, I find it utterly fascinating. To me, it's simply another tool in my artistic arsenal. Let's delve deeper: art is essentially a concept expressed through a chosen medium. In the past, artists commonly used water or egg-based paints until Jan van Eyck introduced oil paints, heralding a new era in art. It's an ongoing history in the making. While we may have differing opinions about AI as a creative tool, we can agree that it's becoming an integral part of the artistic landscape, particularly for the younger generation who see it as the future. Is it fair? That's a subjective question. If I were the first to paint with oils and someone surpassed my skill, yes, I might feel disappointed. However, just like the color blue was not always available to artists, norms evolve. Not everyone can master oil, watercolors, or digital software like Adobe. Colors and lines are not limited to a single individual's ownership. They are like pre-mixed pigments, ready to be used, and in that sense, AI can be seen as a pre-mixed set of vectors, expanding the creative possibilities.
It's not a natural process of progress, it's ideologically driven. One rooted in the desire of the lazy to plagiarise the works of others who actually contribute what those lazy want to society.
I am also an artist an I used A.i. in my daily job. Let me tell you this: If I say "hello" to you and instead of hearing your voice I'd hear a generated speech based on generated words, I'd consider this disrespectful at best. Same goes for a.i. art. It's not art if it's not made by human, and if there is more human than a.i. in the work, then it's not an a.i. art anymore. Personally, what a.i. does for me is it reduces me to a retoucher, who paints over a random picture. This is not art. This is garbage. When using a.i. generated images, you are not becoming an artist. You are becoming an appendage to an algorithm. If you want to be an artist, go learn actual art, which involves human skill. And no, it's not the same as using photoshop. Photoshop doesn't create composition for you, it doesn't create your palette for your, it doesn't create your brushwork or your personal style or anything at all. Your analogy is wrong. Art, among many other things, is the way we communicate and share ideas and visions. When I see a.i. generated images, all that is communicated to me is "this person can type words", because I fully understand what part of this "art" was actually created by a person.
This discussion about the arbitrary line of where human involvement starts is very similar to the discussion we are having in education: how much work does the student have to do so it’s not plagiarism. Is it enough to summarise/ slightly change an essay that AI has written?
Don't worry in time people will start to use AI for self learning no need for teachers, Oh hell we're all doomed, RUclips and AI and learning by your mistakes it's the future.
@@celtshaun1427it's the future bro this shit will put humanity in the drain it's like the AI bros only reply this is the future you luddite like do you guys have any other response other than you are obsolete like bro do you have a brain we creators made what you call art work in which you guys mish mashed and stole so don't come here talking to us as if AI is some godly entity when in reality it's just a pattern recognition technology that some corporate Dev and CEO make to steal and profit from us artists
i struggle to understand the difference in definition between that not being human authorship, and other ways humans use machines to create art. i am really against currently copyrighted material being used for this tech but just dont understand why the tech in of itself is being demonized
It is curious how the laws do not limit or regulate the application of robots and artificial intelligence in the productive area (industry, agricultural sector, digital companies) nor do they dare to indicate to an entrepreneur what and how to apply it in their businesses, but they do generate all these derisory observations to the artists.
AI is like any brain memory or a book. AI learns and keeps a memory like any person who feeds the memory with info, knowledge and experience. Learnig is not plagiarizing. Otherwise, all the people, conservatories, universities, books, etc. would be violating copyrights. Plagiarism is a horse of another color and it´s easy to demonstrate.
I don’t think we should protect certain celebrity styles because those who are rich on top and corporate their style and Disney can do the same corporations do the same and limits creativity
As the chess game of legal discourse regarding artificial intelligence continues to unfold in the hallowed halls of jurisprudence, the unrelenting tide of technological evolution pushes forward, undeterred. Lawyers and adjudicators, steeped in precedent and principle, strive to wrestle with the complex connotations AI introduces to our established legal infrastructure. Meanwhile, the sphere of artificial intelligence itself refuses to remain stagnant. Instead, it challenges itself and us, pressing forth at a rapid pace that it's nearly tangible, ceaselessly advancing towards newer, more sophisticated echelons of innovation. Every sunrise heralds new frontiers in our exploration and understanding of AI, thrusting us deeper into an epoch marked by path-breaking revelations and boundless potential. Hence, as legal deliberations rumble on, the tireless march of artificial intelligence toward tomorrow never misses a beat.
Except it actually takes skill not just asking a machine to generate an image the photographer has to edit and find the perfect timing and subject just to find the photo
@@dmreturns6485 photography is made by capturing moments in life in which that could stay in time permanently that is via image photography is beautiful because it captures life as it is whilst ai on the other hand is basically your ordering mcdonalds and those who claim to be an artist in doing should be questioned without a doubt as ai is stealing from artists not making new content
Is language copywritable? Surely the prompt language is the creative part here. And I spend hours tweaking prompts. Tell me that's not a creative endeavour akin to a client describing to an artist
well... you can copyright a book, is not the language itself, i think is more closer to a code, you can't copyright a code in programation. a prompt is basically the same
So.... you have to keep the prompt... especially with a complete, complex prompt providing creating details... (without saying the prompt was made by another AI product.. oops.) ?!?
If an art director prompts me as an illustrator with ideas, then curates my sketches or art, and suggests changes, nonetheless the work was done by me. I would have the copyright unless given away in a contract. AI cannot make a contract, and cannot hold a copyright.
It will largely depend on your local copyright laws. While the US does not extend copyright privileges to computer-generated work, there are other countries that do because it is an element of their copyright laws. For instance, in the UK if you create work that is computer-generated, then you retain copyright ownership of that material because the law permits it. I also think that the argument against AI is not as strong as people make it out to be, though I understand the frustration with it. People are upset because AI was taught by essentially looking at other available art... But that is no different to every art student, every artist, every designer, who at some stage or another, has studied works of art with their own eyes, or who has sat in a classroom learning techniques using other art as a reference point. The only difference is that AI can do the tasks at a much, much faster speed. So when you strip it all down, it's really no different to comparing a human vs a calculator.
This feels like an outdated law before it has even been enacted. With all software implementing AI in one form or another in the next few months it will be impossible for any structure to look at individual copyright claims. The volume will be too large and I feel sure we will have to revert to an assumed copyright position for all creative work unless expressively stated as 100% AI.
Sure, but it can still be illegal and individual rulings on cases can cost those ppl dearly, same as it happened with movie piracy which has declined in countries where persecution has become on this case to case randomized base.
@@LorettaBangBang Hang on we were talking about if images an individual generates can claim copyright. Not be given copyright is not the same as breaking copyright. What litigation would you be liable for if nobody owns the copyright? Movie piracy was clearly a breach of copyright because someone owned it.
It's not a gray area, you can use Ai as a tool, if the final results doesn't require use of other tools or a skill you have it's not your work, be honest what you actually contributed to the final outcome.
At some point AI will let you decide everything that is on a image (light, camera angle, subject, composition) exactly like you want it just like in photography. Then AI will behave exactly like a camera. Then we will have another conversation.
I asked ChatSonic and ChatGPT about how the app's gonna know and control copyright of the music made with this system, they saind me "it's a huge challenge for these companies to do it". That's why I am confused, for example, "I" created almost a hundred pieces of music on an A.I. app, but only 10% are listenable and these 10% is in the middle of trashy music. How companies will know if the person took some of these crappy things and tell the world it's yours? It's a bul*****, because if you pay, you can have the copyright, so the problem here is MONEY, elite is going to absorb and make more money with A.I. , pay for this and nobody is going to know.
This is a nice and instructive video.(AI and Virtual Reality: Creating Immersive Experiences with Technology) I hope there is a lot to learn from it.,,,,,,
I guess the amount of alteration required to define a work as legitimately different enough from the original output to qualify as copyright could be expressed this way; imagine that someone takes an image you have created - how much alteration would you feel they would need to make to your image in order to claim copyright of that 'new' image for themelves? Most artists would feel that simply taking their work and making minor changes to it would not be enough for the copyright of that image to pass from them to the person making the changes. There would need to be substantial alterations to the orginal work before it was acceptable to define the resulting image as a 'new' work. If we apply this to AI generated images again it would not be enough to make some minor alterations in order to claim copyright- substantial changes would need to be made in order to claim that the resulting image was a genuine new creation that qualified for copyright protection. I do wonder however if the ability of AI's to create iterations of exisiting images renders this whole debate null and void. For example, suppose I see an image you have created and would like to use that image in a book I am publishing- I could perhaps licence that image from you, or I could simply use your image as prompt to generate something similar and use that instead- it's not clear to me that there will even be a market for image licencing in the future so the issue of copyright may already be an anachronism not worth worrying about.
If it's in the internet, its under public domain. It is everyone's interest to protect their intellectual properties, it means do not published it that you're not willing to share it to the world, because you have specific consumers who are willing to give you money.
The logic here is a little funny. So you do a lot of work through ChatGPT so you can stay a small team and dont need to hire other people. But somehow doing the same with CG or visuals and just generating them rather than hiring someone is a huge etical issue? Any team or company can claim they need to save money.
Ai is the future and it is futile to try and stop it with laws from a time when it didn't exist. It is also a tool to express your fantasy and it still has a way to go to be a very good tool. There is a huge amount of work necessary to refine an ai-image and a lot of knowledge in ps and/or other software to do so. If i learn to draw, it is still my fantasy coming to life. Did anybody complain, that digital art is not real art, because you have ctrl+z instead of a manual eraser? It just makes it easier to bring your ideas to life. Your art is not how you do it, but what is on your mind. And that is why it is still not the best tool, because there are limitations to what you can do right now. To get to the expressions of the pictures in your mind is a long and hard process with a lot of redefining and frustration. There are good ai-artists and bad ones and you need to learn how to comunicate with the machine to be good. The demand for this not to be art is just fear of artists to lose their jobs, while in reality they just need to adapt, take this new great tool and to express their minds like they did before. Creativity is a gift not everyone is blessed with. This is what makes an artist, not the tools he is using.
The problem with it is that, in fact, the real huge ammount of work necessary to make AI to look better and evolve, comes from human artists whose works are being used without consent to create derivations of them... It doesn't matter how good you are on AI, if there's no new artists to feed the AI , it will stagnate what it can do eventually, also, you can't build an authenthicity through AI, because is not something you made with your own hands, you can't build your own style, because literally you are using style from human artists on digital art you can build that authenthicity, that's the difference. Also, there's no really need to learn a lot of those softwares, literally we don't have 1 year since it was released, and everyone is expert now lol in 1 month you can learn everything you need and more, the hard part is to have a good Nvidia graphic card, and a good pc
@@Mente_Fugaz it is a tool, which is still in development. To try abd stop its development is backward thinking. New Art and new artstyles are always influenced by existing art. Every artist got influenced by other artists, combined different styles and gave something of their own. This is no different, besides the machine is influenced by a lot more artists than a human could ever be. But say it is like you told me (i talked a lot with friends about that exact topic and i had the same thought) and no new artstyles come from ai: Then there is the future for human artists. Real creativity. And there will always be customers, who'll prefer human art. Also no human artist will have to do jobs which are under their dignity. Bad paid jobs to design a logo or a flyer without any creative freedom. Look at the market and how art and design is paid right now. If you're able as an artist to do these works in about 3 clicks and it pays, why would you complain to have more time for your real art. The problem is not the competition by ai, the problem is that being an artist is an underpaid and unthankful job rn and people are afraid to lose the little bit they got left.
@@Cruentus the machine is not " influenced" man Is just an algorythm, You are using works that you don't own to put them on a blender of data, is literally using your actual work to decompose it's data to make something without your consent, Is the actual work of the artist being used without consent , Like when you through fruits on a blender and then you say " i made a smoothie, this is my smoothie" when you stole the fruits from another person's basket, Humans learn how to cultivate their own fruits, that's the difference, It's simple , if someone has amazing work and an identity, And disney want to work with it, They must pay for the artist and have the consent to use his work to put it on a blender.
@@Mente_Fugaz look, i'm not trying to make you angry, i just want to shift your perspective from the fear to get obsolete by new technology to the massive benefits you'll have from said tech. Like i said, real creativity will come from human artists for a long time. It is futile to try and stop new technology. Say you'll be able to stop it in the western world - then it will be developed in asia. You have to adapt to it and use it before anybody else does. It might seem to be unfair, but in reality not too many people are interested in it right now, so it is a great opportunity to start with it early on. I'm not arguing with you, i want to help you to see it differently and more positive. Also it is a bit more than just a blender. It is a pattern-recognition and -recreation. One could argue, that real intelligence does the same, but with more intentional understanding of the result.
@@Cruentus AI doesn't need to be stopped, If someone wants to make AI models with licenced and public domain data, or your own works, it will be a great tool, and you should be able to copyright your creations with it, but right now, what everyone calls "AI art" is derivations of stolen art, that's the issue here, AI can't evolve without new human pictures to feed it, or at least, it can't evolve into something that looks good (it would just evolve into abstract stuff, because it functioning doesn't follow any principles or fundamentals, it just follow the concept of replica). you need to understand that, if there's no new artists working with fundamentals through their own art, AI will stagnate, and if there's no motivation for new artists to keep learning about how to create pictures because they can't make a live with it anymore, and is more profitable just grab the work of other artists and in seconds make infinite derivations of it, Art itself will stagnate, at least in this specific topic "illustration"
I don't find the chatgpt poem a good example at all. So if I tell it how to rhyme, and keep adding a lot more detail, am I not in a major way leading how the result will end up? As Samson mentions, there is a big difference between saying "make a nice image" and more detailed promts, like "make a pencil drawing of a girl with thin lips, eyes further apart than normal, messy blonde shoulder length hair, freckles and a mole just to the right of her nose-tip". Of course that could be taken further, I could have manually drawn a landscape and use photoshop to insert her into it? I only get to copyright the landscape in that picture, or does it now constitute a completely copyrightable work of art?
This is obviously new to society as a whole, and we have a long run to cross yet in a legal enviroment... but proclaming that just by a machine producing a material isn't copyrightable is just as saying that a photograph made by any camera isn't also copyrightable. BUT IT IS! Any photographer can proclame ownership of any picture that he takes althought every single pixel of it is made exclusively by the 'machine' Camera. And the photographer also (most of the times) didn't do a single bit of a thing onto the scene that he takes a shot on. So why he can copyrigth that?? It is just because he bough an device that helps economy as whole? Instead of the current state of AI Ecosystem where you can generate a ton of images even for free? I believe that a good starting point should start from the companies that provide AI Art generation in making clearer as possible that any image generated in a free plan isn't free for commercial use. But at the moment they just explicited that the images aren't in your 'private channel'. This is what makes the regulation difficult in my opinion, because anyone can copy any image made on those free channels and proclame that it belongs to him/her. And how he/she (the real owner even) could ever prove that? Thats why is important to make it clear how a free or paid plan should work. Then would become much easier to prove authority and ownership of an AI Art generated.
i love your content and feel free to delete this comment, but the technique of slow zooming on every shot is distracting and is drawing attention to you rarely blinking. trying to be constructive not critical
Here Consider this: you have a creative idea, and you take paint, pencil or whatever tools you need to create that art...right. Well now let's look at AI we take a program that learns from a prompt (a different kind of tool like paint or canvas) and takes what you prompt it to do, which is in essence human creating the baseline of what is to be created. So, without the human aspect you would notget the images or video result at all!!! I think in many ways there are different kinds of art, and digital art has been around a long time, this is just a new form of digital art, which in essence creates a new kind of art! You take the prompt from your brain, and AI artist take that same prompt from their brain, and both manifest the image! Don't hate the way art comes to be! The idea of art is creative expression, AI and what you see as the norm in art are one and the same just a different path to accomplish the very same goal...it is still art, no matter how you get it, sometimes a closed mind only sees the negative outcome, try opening your mind. Computers & programs have been creating art for decades, in movies, aand other advertisements, what you see in your everyday life was created by a camera (essencially a small computer) or a computer to generate the effects, which a type of artist created...Do you think that art should stay stagnant while the art world and the way it is created evolves? Art is art and there are many forms of it, to condemn this form of art is simple hypocracy. Get over it, do what you do, and evolve or don't but, not allowing yourself to evolve, makes you blind to the reality that is the world around us!
I think artists should keep their a record of their prompts and the outputs from say mid journey then they’ll be able compare with the final image after all of their tweaking and expression. Maybe this should be part of the registration process. If you think that’s tedious, try registering a collaborative music work with artists who aren’t on the same registry and use different names 😅
This would be task of the platform tbh. Midjourney's user experience is still hideous right now. Scrolling up forever to find something I did in the past? No thanks. Ai summarizing as text whatever I was trying to create that day and collecting it by date? Yes.
That has nothing to do with legalities because AI Art Software is not own by the USA Government. So there isn't a legal obligation with a heavier legal merit to enforce that.
@@absolstoryoffiction6615 good point there isn’t a centralized policy. but the contention is the licensed work the Ai is trained on. The rights are owned by many artists of many different jurisdictions and even after all this time I see some companies outright ban or seriously hinder Ai. Not to mention social media is updating their policies soon. Govt might not be the only regulation at this point
Humans have no more control over the art that they create than an AI does, but y'all ain't ready for that conversation. A good enough FMRI can predict your thoughts well before you are even consciously aware of them, so can you really say that you control your thoughts? The answer is that, no, you don't have any control over your thoughts at all whatsoever. Every single thought you've ever had just appeared in your awareness whether you wanted it or not. Sure, they're based in part on what you see and hear and experience, but that proves even more so that you don't control your thoughts. So if you don't control your thoughts, are you responsible for them? Are you responsible for your ideas? Are artists any more responsible for their creativity than an AI? No, they aren't. So making a distinction between art created by an artist themselves and art created by someone prompting an AI is meaningless. The only distinction between the two is the amount of time taken to create them.
The problem with the copyright office's opinion is a prompt is human input, and a lot of work goes into creating good AI art. It takes hours of input and sorting through images to get something, which itself is creativity
@@jphakphumromsaithong when you pay for an artist to make a comission, the copyright is for the artist, because despite how much you specified what you want to the artist, the result wasn't made by you. You are just the client of the artist, in this case is the same, you are the client of the AI, you paid for the product
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
Check out this video for the most recent AI Law developments:
ruclips.net/video/uyZXxLO7rSI/видео.htmlsi=s9fkGmmjHxCU5-aD
The answer is simple. Use artificial intelligence to explore ideas, and then hire an artist to draw for you. In this way, you preserve your rights
U can use IA and create variations of an image and then you do a composition using photoshop for example and create something new, thats is possible, I dont see not ilegal here. What do you think?
@@LeandroT242 i agree with what you say. The only thing i would add however is not being allowed to just add leonardo style at the end if you want the copy-write. He made his own style and it took him a lifetime so why should we be able to use that and claim its ours just because AI can copy it?
In my opinion there should be a combination of styles with some aspects of Leonardo for example where the AI artist put in a lot of work to make the work unique instead of a straight copy of Leonardo.
Use your brain to discover ideas. Learn how to write. Learn how to draw. RUclips has all the resources you need to do that. Then you become an enriched person who has achieved and own everything every time. Problem solved.
A human artist??? You know how long it takes them to produce art?
AINT NOBODY GOT TIME FO DAT (BRONCHITUS)
@@taiyoctopus2958your comment screams daddy’s money
When you say protect your work, do you mean sitting around and waiting for a machine to do all your work 😂
Its funny how they see machine grifted stuff as "work"
AI gets paid more than humans@@beth1979
Are you saying someone who uses a blueprint and a cookie-cutter frame kit can't build a log cabin-Even if he isn't Amish?
@@doctormelancholy3042 It depends if he stole that blueprint or not, from hardworking blueprint draftsman and did they even ever gave him permission to use said blueprint or not. Though I doubt it because he’d be too cheap to pay with time and effort.
@@doctormelancholy3042 There's nothing wrong when they just build the cabin for themselves, there is something wrong when they start to sell it on an industrial scale and the design isn't theirs.
Hmm. Little late to the vid. But those first images from midjourney, when it started, were eye opening. It made me think even more creatively. To see how a machine puts images together was thought provolking.
The fundamental problem of generative artificial intelligences lies in the initial violation of the rights of the authors of the material used, in its "training" since in reality the images now encoded remain in its model, (this explains why companies offer the possibility of extracting them if you have the copyright) Therefore any image generated by the AI is always based on previously loaded images, if it does not have them it cannot generate them, therefore it does not learn copy. In short, the authorship of a user in the generation is minimal, only the idea, since with the same propmt the system generates very different images, like a lottery, and the ideas cannot be registered by themselves, they must belong to a work, and here the work is not human.
It isn't any different if it's AI or a human. If an "artist"" believes their copyrights have been violated, they can make a claim to the kangaroo court system to see if they have enough to payoff the judge. Once you publish your work for others to see, in particular online, you take a chance. There is also "fair use" and anyone can trace an outline of the picture you took of your rear end and posted online and add to it and color it and then even you might not see that they borrowed your stuff-If you catch them, call a lawyer. Otherwise F**k off!
@@doctormelancholy3042 Only mediocre people without talent use AI, good luck with your mediocrity
@@marcelootero6866 So you are calling video game developers mediocre people without talent? All while your avatar screams "mediocre and without talent"?
@@doctormelancholy3042 Its completely different if you knew anything about how AI works. A human does not dissasemble 20 million pictures into pixels and t heir location and then reasembles avaraged out aproximations of where the pixels were. You are full of shit m8..
@@doctormelancholy3042 Yes overwhelmingly, why do you worship game developers when the majority of industry struggles to make anything interesting since years? Also nice personal insult of the guys face, really shows you are the smart one...
It helps to call what is generated by AI differently from what is created by cameras, artists. I use the word 'synthogram' as that means the output from a synthesis of elements (tensors from the learned image manifold).
Agree 👏🏾I'm personally needing a distinction between generated (prompt to outputs) and assisted (which can simply be auto correct or auto color)
@@BenCaesarIf you're not creating anything de nuovo, AI enhancement is what I'd talk about (auto colour is not necessarily AI. If it's adaptive and responds differently to different types of scenes, maybe).
Synthograms come from AI models responding to prompts.
@@TomAngPhoto thanks for the clarification! Lightroom has some preset suggestions according to the image this might be more the Ai you're talking about. It's adaptive than the preset library.
Thank you for explaining! I can imagine how complex this topic will be in the future, especially when text-to-video is so improved that it can be used for actual movies.
Thank you for this nuanced discussion. This nuance is what is usually missing from the conversation but it's everything. AI in the hands of an artist is a tool and the end result is usually the last step of a thoughtful multi-step process.
Ai in the hands of an artist is a tool? That one might be the new low.
The next stage is to get Stability and Midjourney to completely wipe their datasets and start from sets that are only sourced from Artists who opt in.
Or they just keep their outputs and purge their data sets and then you'd never even know wtf it's using lol
> *only sourced from Artists who opt in.*
That would kill using public domain works, which seems avoidable on top of needless.
@@gondoravalon7540 Then tough titties for people who steal
@@gondoravalon7540
Under USA Copyright Law... Public Domain, Non Copyright, Licencing/Contracts, TOS/EULA etc... Are all legal to use for AI Art.
So in other words, u want to kill those ai. Machine learning needs examples to work from, and I hate in when people say it’s theft. It’s like going to a museum and looking at art, making notes then going back, every day until eventually u can make something similar. That’s not theft, and neither is ai
Interesting. Recently, I plugged a few of my photos into Midjourney to create similars. I wonder how the copyright office would land on the output. I can bring them into PS and use its tools to do pretty much the same thing, albeit with more control. That said, if I wanted to use the new PS generative tool to change elements of the photo, still a clear copyrightable work? Because of my photo + PS? It’s all so clear now!
I was thinking along the same line. l use my cartoons as the basis for AI art. It enhances as I could in graphics programs. If I copyright my original image and load it and I'm giving permission to AI to work on my work is the outcome still under my copyright?
Someone told me they did this and showed her process to the copyright office and they approved the copyright but she did it with her digital paintings. I don't know her w it will work with photography.
@@ellenripley4837 Great to know. Thank you!
It's an easy thing...
If you draw your OC Character, then put it onto AI Art. You still own the Copyright for BOTH because your OC is grandfathered in as that Copyright IP existed before you put it into AI Tech... ... ... Spider Man by Sony is a good example.
If you did 50/50 mix of Human and AI generated art etc. Then you may copyright the entire thing. Since all AI Art is case by case and not 100% Public Domain.
If you did small edits but nothing major to alter the raw output of AI Generated Art, then you don't get a copyright for it because you're not the author in the legal side.
@@OzVegan
Yes... In America... Yes... Because USA corporations can do that.
20 years ago we were talking about having to think about human rights for AI sometime in the future
Now we are debating how to defend our own human rights against AI
Humans are their own evil...
Such an ironic extinction...
This is interesting, and clearly just the beginning. I think "human authorship" is too vague.
When a movie is produced, the copyright to that movie typically belongs to the movie's producer or production company. The producer is usually the individual or entity responsible for overseeing the creation of the film, managing the financial aspects, and making key decisions during its production.
It's the same for images. AI is just a tool of creative expression. Maybe more powerful than other traditional means, but still it's just an advanced tool.
AI generated movies will exist, but the copyright will belong to the production company/producer.
This discussion is far from over.
Very helpful. I appreciate your stating legalities without “choosing sides.” I’ve been involved in photography, painting, and sketching. When computers came along, I learned how to use but-mapped program, Freehand, Photoshop, fractal programs. Most of the later work was in Photoshop, but I ventured into Deep Dream Generator. Then I hear about AI (DALL-E and Mid Journey). MidJourney can create awesome works but the artists’ intention plays a big role. I can make dozens of prompts in search of something close to what I imagine. I’ve created many images in which I combined 3 different AI images in Photoshop and then did a lot of work in Photoshop. It’s a journey. Sometimes AI is pretty dumb...7 fingers and 3 arms, limited ability to portray complex emotion. In terms of copyright on occasion I make a very simple prompt because if I don’t use a long prompt, it allows MidJourney to be its most creative. With those I definitely could not claim copyright. I did submit AI art to a magazine. They told me they had not worked out a policy yet about publishing AI art, and I understand that.
copyrighting is not to protect you. its to push the industry forward. the whole concept of an idea belonging to you without pushing the industry forward is imoral to me.
I think protecting the human contribution is fine. AI only creates images or text; it's humans that place those things into artistic contexts from which a human audience derives meaning. The images themselves are now fungible. There's not even really a need to copyright the image; someone else can just use the same prompt and generate their own, none of the artistic value is created by the specificity of the image.
Humanity was created... If only humans understood their irony.
Cope more. It is stolen images. Based on actual drawing artists. All you ai boys do is type some words. I tried stable dif. It's NOT THAT HARD. Try drawing it from scratch ones in your life.
Does this mean that I can use these Ai generated images to make my RUclips videos and not get copyright claims on them?
@@iAmNothingness Try making a video game or film using AI art that doesn't suck!
What if you train the AI using your own artwork and original characters? In this case, the AI is not producing images from scratch based on verbal prompts. It has been trained on visual materials, which are the artwork and references that the artist themself created. There is significant input from the artist, and the AI is just reproducing the artist’s art & original characters by putting them in different scenes.
Too bad, its no longer yours.
This is future I'd like to see supercharged
@@beth1979 here lies the rub 🫠
You can probably copyright those images since the source is yours.
@@ellenripley4837 can't say for sure honestly, even though the input is yours it's running through a dataset of other artists work. Copyright office may not issue a full copyright. If the dataset is trained on your work I think copyright will be easy
I could use your insight to help clear up my understanding on how to properly release an audio/visual project that I have just completed. I wrote, recorded and mastered 9 entirely original synth-wave songs. I used OpenAI to generate and edit 30 minutes of prompt-based visuals that I lined up to the beat of my original music. I'd like to release this as a single audio/visual, copyrighted piece of work on RUclips. Do you have advice on what I might need to consider in terms of copyrighting and monetizing this project? This is my first major release. Thank you Samson. -Carlos in Colorado
That song is copyright able in America because you wrote the skeleton for the music etc. Make sure to send a Copyright Document (depending on your State) to the Copyright Office for legal securities since RUclips is known to be risky with fraudulent DMCAs.
So you probably want to publish your music after the legal copyright documents are filled.
What about camera generated?
The human only select an object, and press a button, using auto mode.
With a camera you have much greater control over the creative outcome, which seems to be the aspect they focus on. But I agree that there are somewhat arbitrary lines drawn. What if you use a self-shot photo as a reference for the AI for example? It's way too complex and they don't seem to grasp all the possibilities.
@@missoats8731but does one own the tree or the grass? Does one put the creative elements into the scene if it is instead Camara and software that does the vast majority of the work?
More importantly, arent there other things like Disney's decades long copyright stuff that might be worth reviewing. Or old games that are no longer being produced or sold still out of public domain for decades?
@@missoats8731 I will just start taking photo's of my AI works and take the piss that way lol
@@missoats8731 Questionable actually, since with a camera you only have choice/control over the immediate area in your physical presence, whereas with an AI you can specifically control the subject of your output to be anything you can imagine. The amount of difference in human control here is orders of magnitude different. Similarly, with photography, you get lots of unusable shots, and make choices as to which of the shots you photographed constitute a final artwork, much as an AI artist picks and chooses which generation out of potentially infinite choices they consider to be the final artistic product. This new "law" or precedent is flawed in many ways, and I expect they will have to amend it over time simply because they are under valuing both the level of human choice and control in prompting, and the future capabilities of ai tools to give the user greater control than many of the current technologies provide. There is a stable diffusion tool called control-net which allows the user to pose a subject control facial expression, layout the composition very specifically. These are even greater levels of human input being put in, and should not require the artist to go in and manually edit the final work just to consider it original human created content.
@@toren9120 You're taking this argument in a direction that isn't at all based in the copyright precedent set by these legal findings. The only thing that is being considered for copyrightable artwork according to the video is how much control did the human have in the process. I argue in the case of AI art that it's actually quite a lot more, or at best nearly equal to that of a photographer, and it's a sliding scale like any process, you tend to get more out the more you put in. The difference in the technology that is responsible for getting a photograph onto film stock or into a digital file, compared to having a text to image tool produce an image is significant, but the process that a photographer takes, in choosing a subject, framing a shot, determining various settings, and selecting what shots constitute a finished product are so similar that the differences are mostly symantics. I'm confident that many copyright worthy photographs would constitute far less effort or forethought than a lot of the AI art that I've produced.
Your argument above is identical to the argument that a painter or sculptor could make towards any photographer in fact. And yet the law does not preclude works of photography from copyright. The original comment made by Christian Pedersen, is right on the money. If you want to ignore that, disregard, or whatever (and I'm sure you and many others biased against ai art and artists will), that's fine, but the law set out according to the video is wrong headed, unless they want to apply it equally to other forms of art, photography specifically for this argument, rendering all such works no longer able to be copyrighted (and that aint happening).
It is important to assume that those making laws know less about these materials than you do. Remember, our nation is primarily ran by lawyers not nerds. So with the information provided, is it possible that one can assume the copyright of the given prompts since those are authored by the artists? How about if I drew something very crappily, being the treditional element of the authorship, and I improved it via AI artwork? Is it mine still?
It's actually run by the rich who want to stay rich, just like the police is the government's mafia, they use then as a threat of violence
How would anyone prove that the image was Ai generated if you tried to pass it off as your own creation.
This is a very important point! I suppose you would just need to hide your trail and you'd be fine. There is going to be a big problem with this in a lot of fields
Firstly I think this whole law is silly and uninformed
But perhaps designers would be expected to present a proof of their work process.
It would be relatively easy to implement a sort of timelapse feature as for example adobe fresco has, that shows your entire work process from start to finish.
If you use services like MJ, it is logged and public
@@stephanreiken9912
Logged, yes
But for it to be accessible to the public then one would need to know the username of the person making it
@@aisamsonreal as someone who utilises AI images, there are consistent elements you'd notice if you looked closely for them... but who's to say if they'll still be there in a newer version, or, like I do, I retouch images to remove traces of the imperfections - so then if there was a bot that found AI generated images, would they be able to pick up retouched AI images?
What about people that throw paint into a machine that splatters it around the canvas and then sell it without any further steps? Do they not have copyright anymore?
The paint being used is not someone elses art, AI Art is... duh!
@@Bartetmedia I love that people keep saying this because it instantly shows they have no idea how the AI actually works. The AI is nothing more than a mathematical function with billions of variables that have been tweaked based off of billions of inputs. When an AI is "trained" on someone else's art, it doesn't copy it or store it or whatever you think it does, it simply slightly changes one variable by +0.001, another by -0.0004, another by .009 and so on. It's literally just a big equation. It's no more "stealing" it's training data than a photoshop artist is "stealing" the colors from an image by using the same exact RGB value as someone else. It really is incredible how so many people have no idea how the technology works and yet are so sure that it's unethical.
Thanks for watching!
What topics on AI would you like to see me cover next?
And if you have any thoughts on how to make my videos even better for you, please let me know, I'm all ears.
:)
Thank you for the depth explanation! I suggest for the next AI topic is concerning the machine learning and the relation to the current lawsuit copyright infringement. :D
copyrighting is not to protect you. its to push the industry forward. the whole concept of an idea belonging to you without pushing the industry forward is imoral to me.
Important distinction to note, is that your video would only apply to those who are based in the US. US copyright laws do not extend full copyright protection for computer-generated work, unlike other countries where they do.
That, ultimately, is why a lot of these platforms still make reference to commercial rights and ownership to prospective users, when using their tool to create art.
The funny thing is, as an artist with nearly 20 years of experience in comics, I don't believe that Zarya resembles Zendaya. Zendaya's eyes, eyebrows, and forehead have distinct differences compared to that fictional character. For artists skilled in capturing likenesses, we understand the significant role that eyes, eyebrows, and forehead play in drawing. Although the images may not be of high quality, it's worth noting that Zendaya's chin is more prominent. Furthermore, aside from their shared Afro heritage, they do not bear a resemblance to each other. Not to my artist's eyes.
Yes, as an artist, I utilize sketches and establish the composition before finalizing pieces with Stable Diffusion + ControlNet. While it may not be suitable for various tasks and lacks the necessary AI capabilities for serious commercial projects, it does provide valuable insights into areas I can enhance in my final artwork.
The link is not opening for me....
Which link?
I got a question. Under these rules how exactly is a photograph a copyrighted work since all the "artist" has to do is press a button and the camera does all the work? where is the human authorship there ?
@@toren9120 and what about the people who do not chose whatever lenses you mentioned and just pull up their Iphone from their pocket and take a photo. They don't need to choose a perspective or anything. Is no different. You could even set a timer when the photo can be taken and leave the phone. That is still their photo and copyrighted. So all your points you mentioned are not needed in order to have photography copyrighted.
Also give prompting a go, and see how 'easy' it is. It will take you at least a good few days until you learn to prompt like an amateur, and longer to actually get something good.
If you are specific enough in the prompt where you already have the idea in your head and are simply directing the AI to do your bidding, then I think you can claim ownership.
@@posthawk1393 I agree. In UK you can anyway. It's more relaxed than USA.
It's because it takes skill to take a photo that looks good. Why do you think photography is a career. People get paid to take photos with their equipment for a client. It is utter lunacy to expect the same treatment when all you do is type in a prompt. I've used Dream AI in the past for legitimate reasons, mostly for fun. I'm also an artist and I never thought highly of those generated images, just that they looked cool and maybe I could take inspiration from one of them cus it was really unintentionally abstract. Instead of typing words into a prompt.
Why don't you:
Type words into a story
do some doodles
play an instrument
program something cool
Unless youre literally incapable of writing more than one sentence or you have no hands or the motor control of a dude with oven mits on, there is no excuse as to why you cant find something creative or fun to do or pay someone to do it for you. End of
@@Lambsauce10 " because it takes skill to take a photo" what about a photo that i just press the button without using any "skills". A bad photo at that.
"all you do is type in a prompt" , "all you do is press a button" Don't kid yourself.
That means anyone can appropriate and sell the non-copyrightable images on tee shirts, cups, posters etc. In addition, the tech overlords can still take all your “iterations” to train AI and reuse the AI generated images the person “thinks” they have creatd
Yes and no... The USA Copyright Office made it clear that not all AI Art is in the Public Domain / Non Copyright Able. It's all case by case.
Sounds good in theory but how you will be able to verify who is telling the truth will be a headache. Unless they start requiring proofs but then the process will be incredibly long.
This is basically the law you have right now with public domain images. You don't own the rights unless you change them considerably to make them your own.
in the announcement of the copyright law, they said that in case you modify the picture,
what you own is just the modifications of it
@@Mente_Fugaz
Yes and no... It depends on the 50/50 mix. Do it well, then you own the Copyright of the entire IP.
I`m curious about possibility of "style" protections in any kind art form . Is it possible ? How many actually new styles we ever see or hear right now ? Without getting inspiration from older arts ? Say some artist who can manually draw in canvas or using mouse and stylus but the final output could be have same style with another artist , how the law will applied ?
Is that also means a song cannot copyrighted because drum or guitar sounds same with another songs although they playing different notes ? How about guitar playing styles which generate using same technique with another famous guitarist ? Or how the law can forbid for any musician using same presets and sounds ?
Or we`ll having different laws between any form of arts ? Because in some art form like music , Styles could be outputting same sounds because they might be using same techniques
there's a freaking lot of new and unique styles, you just need to search for artists lol,
anime and pixar is not the only art that exists,
I think the law must be regulate the consent and the right for someone to just don't be part for a model...
@@Mente_Fugaz show me examples so I can know for sure that those new styles not just iteration or modification for old ones
@@morizanova wtf.. man, style is not "something that comes from nothing"
style is the way you make something,
take this simple example:
if you feed AI with just the works of "samdoesart"
AI will just make derivations of his works,
but if you show the same works to a human artist,
he will put it's own thoughts and intention, his own unique perspective of what he sees on those works,
creating something new with his own authenthicity, every style born that way...
AI can't make it because is designed to replicate and recreate the information of it's database, so it can just create derivations.
did you know that the "anime style" was a reinterpretation of Disney's eyes and faces?
you think that AI could create the "Anime Style" if you feed a model just with disney pictures?
No!, it will just create more disney pictures lol.
Ai is a doppelganger,
the concept of "generator" and "discriminator" is literally tryharding to recreate an actual picture, but it can't so it spits something that the discriminator thinks it looks the same.
That's why AI can't make something different of what's on it's database, just derivations.
@@Mente_Fugaz we talked about difficulty to copyright styles. Because you can easily trace the origins if you have vast of knowledge and origins
Learn to read first before replying with same old AI cry babies songs which everyone already knew
But let's ignore that , And talks about your arguments
TBH ,I doubt you even ever doing deep dive trial error in AI tools , because your long points still not quite right .
Right now You can easily mix some Disney with luis royo and Mobius art style and if not enough there are already thousand of specific artist style which build as mashup tools
Standard fan art artist will make their copy as it is but creative ones will mashup those character fan art with contrasting style so people who have limited knowledges will thinking that new styles
And right now AI with LoRA etc can do that . Only newbies AI users will keep pumping one art style . The experiences ones know how create good blends
@@morizanova
Copyright styles:
you can't copyright styles,
you can copyright works of art...
and if you use AI, you are like a client asking for a comission to an artist,
but the picture wasn't made by you, so you can't copyright it, despite how much specifications or sketches you shared to the artist in order to him to understand what do you want.
Also a work that is not yours, you can't use it for anything you want,
you need the consent of that artist to use it,
you can just consume it,
but when you use art that doesn't belongs to you, to train an algorythm to replicate derivations of it, that's simply unethical... I don't know how there's people that can't see the obvious thing yet.
about merging styles,
when you merge 2 styles, you think it's something new, but those merges have an origin on the dataset of the original stable diffusion models, if you try to merge styles, it won't have any direction, there are just glitched derivations. you sound like those who think that AI is magic or something...
man, AI doesn't understand what is doing unless is on it's training data,
the only way that creating styles with AI could be something, is in the case you train AI with those glitched derivations, tag them, and then you try to work with them. you could make a lot of interesting abstract art.
But let's be honest here,
what you are calling here " create new styles" merging algorythimcally a bunch of pictures, is just using something similar on the database that looks like a merge, even actual pictures merging those styles etc, and creating a derivation of it.
if you see a "new style" as an output of AI , that looks good,
is because is not a "new style" it's just using stuff on the dataset that you think it doesn't exists lol
you sound like those who think that AI is magic or something...
it's also hilarious to use the term "newbies AI users" when this technology don't have a single year since it was released lol
people can become expert in 1 month or less,
you just need a good pc and a expensive Nvidia graphic card...
and being suscribed to a lot of twitter and reddit accounts that post new models and AI tools.
Because the only real way to improve your art on AI is to wait until new tools and models arrive
Where is the copyright when I generate images and use them in a video like on RUclips? I upscale the images after generating, find a fitting music and create a slideshow, sometimes with storytelling elements 😅 If I would use only copyright free images (not AI generated) in the video, the copyright for the full video would still be mine? Whats your opinion on this? 😅🎉
In that example... In America, you own the Copyright of your collage of Public Domain Works but not the individual Public Domain Works used in your collage.
You can copyright protect you Collage but not the individual Public Domain materials.
As a creative director, I expect my final pieces of film to be considered one artistic piece. Using AI just means that I can do the jobs of many in the course of days rather than weeks or months. Writing high quality prompts isn't something everyone can do. Fug the courts!
So, i can use Ai generated images to create my RUclips videos and i won't get copyright claims for using them. That's what i want to understand. I use Copilot to create my images but i'm still worries cos i dont know if i'm free to use them on RUclips or not. Someone put me through please
But the AI generator states you have fll copyright of the images produced
Thanks Samson. This is one of the topics that I -- despite keeping up with the latest -- have so many questions about... many of which you addressed here.
This is a tedious subject that I'm always exasperated with before I even begin addressing it.
1. I'm one of the few people that create and train my own stable diffusion models on my own artistic works (photography, graphite, procreate and krita) and merge them (or LoRa them) with other, larger models that are themselves trained on only royalty free content. Continuously creating, recreating and merging my work and these models together in different ways is like creating a damascus steel of your own creativity, continuously folding it in on itself over and over until you've created something that's so uniquely yours, but almost alive in a way. It is AI, but it's your AI, your creation, creating new things based on your visual and textual inputs. Inpainting, masking, outpainting, img2img, controlnet, ic-light, multimodel, prompt and seed migrations during animations etc etc... it's limitless. It's difficult to describe how satisfying, enthralling and, contrary to what some (understandable) comments here claim, how *genuinely human* this kind of work is. As you can imagine, from my cold dead hands would anyone get these models from me. They're everything. My inference machine is literally air-gapped from the internet.
2. Digital laundering of intellectual property into neural networks is exceedingly easy (as midjourney has so red-handedly demonstrated). This is the major crux of the issue: current methods of assessment by governing bodies rely on nothing but honesty. If I was to hypothetically create new work using the work of other artists, and I didn't want you to know; then you wouldn't. The ONLY way to genuinely prove that this work is 100% my own and legal is to provide the models, the exact repo of Kohya SS I used+it's training values, exact copies of software tools I use for inference, and most importantly - the datasets. Only by providing this kind of prepared ready-to-run data&software package (of my entire artistic identity and secret training sauce? 🤨not a chance) to a governing body, can they then launch Kohya SS, load my training json, train/merge/lora the model themselves and compare the generations using the same webui, settings, prompts, seeds etc.. to confirm authenticity. Only then would giving copyright to an artist using AI like me be truly fair and ethical.
Which is a tough thing to admit to myself, but if our goal is to be intellectually honest and responsible in a way that secures the intellectual property of future artists, then sadly that road isn't an easy one. As usual. 🙄It's a mess. This all happened way too fast for us. But just like the internet, here I am in the barrel of the wave. Only this time around I've no naive delusions that this will make the world better. What an adorable mind I had back then.
“I’ll show you what you can do to Protect your work from unauthorized use and avoid infringing on other people's copyrights”
As someone with student debt from art school this from the description is absurdly hilarious.
What if I got the image from the internet, modify it using AI, add more elements, filters and other modified AI images into one frame which makes it a completely different and unique, could it still be copyrighted? I hope someone could answer this. Thank you!
So what happens if you just say your AI generated art is your art that you painted?
All seems very honor system to me, same with licenses.
Prompting is human input
but you are not trying to copyright the prompt,
you want to copyright the picture lol,
What about my creativity in copying a prompt ?
lol,
what about my creativity in posting cool AI art that i saw someone made,
and them posting it like i made it xD
It requires creativity, because you need to know where you can find something you like
The fuck man
Really interesting, thanks for keeping us all up to date!
Any time!
QUESTION: if someone takes an image and it is altered significantly by AI does it not cease to be their likeness? I believe it was found that if it is altered significantly or is considered sufficiently "transformative" it can be used or am I wrong? As an example: If I make an AI influencer on instragram based on let's say actress Emma Watson, but with a different name and an AI has generated an avatar inspired by Emma Watson... with significant alterations: changes to the nose, different eye color, changes to jawline, significant body changes, different hair color, etc--though someone may look at it and say it reminds them of Emma Watson or maybe it looks like her, it nevertheless has been altered into something else or transformed at least. And then that person uses the AI to then sell naked photos of this AI influencer...is that a legit cause of action or otherwise something to be sued over? They are not claiming to be her and the differences are significant.
Good video. Great examples that clearyfies the otherwise quiet abstract content!
Its absolutely not a tool and the nature of it replaces said artists.
Or it just changes art.
@@CultofThings Made by an algorithm sewage.
@@CultofThings i dont care what it does, you are not allowed to use it because i didnt allow the use of my work by the AI. You essentially a moral corpse, happy to use a fence for stolen goods and your argument is "the bike is different color and reasembled parts of 10 different bikes so its a different bike now" Police begs to differ, and the entire civilized world and property laws speak against your greedy selfish thieving little excuse.
Why not accepting the fact all existed AI models are stolen and modified. Thanks to Stable diffusion, AI model is being trained updated everyday until now from unknown sources and persons.
the ceo of open AI told that there must be laws for people to have the right to opt out all their content from the databases, even in the socialmedia models,
I don't know if with that, he was refering that there could be a way to opt out from the models at this point...
@@byujkt3822 I'll accept that when you accept that all art is borrowed or stolen-However you want to put it! The "greatest comicbook writer ever" Alan Moore came to fortune and fame writing Superman, a character created by Joe Shuster, which was stolen by DC Comics. Moore went on to claim his stuff was stolen by DC Comics as well. Live by the pen and die by the contract. Stolen or not, his Superman and Batman were better than most!
Creating through the symbiosis of artificial intelligence and human imagination yields products that are legally eligible for copyright protection. This emerging paradigm of collaborative creation poses novel challenges to the conventional frameworks of intellectual property rights. From a legal standpoint, the question becomes: how can the copyright office distinguish between the human element and artificial intelligence in the creation process? Indeed, this presents a complex puzzle. Given the inherent anonymity of AI and its inability to possess legal personality, accurately identifying the contributions of both entities in the creative process will be challenging. It's a task that stands to be extraordinarily difficult, and yet, it is a necessary evolution in our understanding and application of copyright laws in the age of artificial intelligence.
hey, what if you turn your ai art, & literature into an NFT; do you own it then?
@@pay_it_forward_franklin4469 Good idea, but I've tried that already.
@@I-Dophler wait; so if that doesn't work, being that the artwork has a seed-number, & the literature has zero-plagiarism; what givez, bkuz I am LITERALLY in the present to present both ai-art, & lit. as I am writing this #againwhatgivez
@@pay_it_forward_franklin4469 The complexities of copyright law in the context of AI-generated art and literature are indeed challenging. The seed number you mention is a unique identifier for the AI-generated artwork, but it doesn't necessarily confer copyright ownership. Similarly, the fact that AI-generated literature is free of plagiarism doesn't automatically grant it copyright protection. The critical issue here is the concept of 'originality' and 'creative input', central to copyright law. AI is not legally recognised as capable of original thought or creativity, so it cannot hold the copyright. Therefore, the copyright status of AI-generated works is still a grey area in many jurisdictions. It's a rapidly evolving field, and legal frameworks struggle to keep pace with technological advancements. #AIcopyrightissues
@@I-Dophler Yea I understand, but wanting to give to those, whether at leisure, or especially in need; all I want to do is exhibit within the linez with no penalty for. So, for example; I put out ai art with literature for monetary-gain; what restriction(z), &/or penaltiez take place?
AI art is for posers, hacks, the lazy. Those without the mental fortitude or moxy to venture into a lifelong pursuit of passion with all the years of hardship and pain you knowingly sign up for, and still say its worth it. AI Art isnt a tool, it's an exploitation, a shortcut towards what i just described. It's for gamers not livers, the people on the outside of hard won skill that are too afraid to tempt fate for years purely for the sport of human endeavor.
Ai art is for the weak. The losers. The excuse makers. People that have the time to sit around for idle hours and come up with well crafted arguments being edgy about it, bc theyre not busy developing a skill instead.
Ai art is for the compacent and the unengaged, the lower level of sentience. People without that existential buzz within them that knows you can only live it or youre a sham, that difference between results and the work put in to arrive there.
All that's changed is more people will convince themselves theyve achieved something creative when they havent, but deep down theyll never know what it actually feels like to risk it all on that passion. To truly EARN it. Becsuse theyve risked nothing theyve gained nothing. They wont know what it feels like to, despite it all, earn a creative skill and worldview. They will have committed the biggest crime, funnily enough, against their own person, their own humanity, and id say thats a fitting justice. Let these masochists of the soul writhe and die if thats the fate theyve chosen.
Meanwhile real artists arent going anywhere. More posers will get weeded out by this and the more dogged, gutsy artists will continue to thrive in whatever they choose to do, bc theyre about that life.
I guess for me is - are you going to copyright the wood and the sand you used from the computer as yours, when you used it as part of your work, or the end overall results itself, a reproduceable outcome that's unique to your creativity?
USA Copyright Law for User Generated Content (UGC) has always judged on the Output of the user. Not the Input they used to create the Output.
Services such as AI Generated Art Software DO NOT follow the same legalities. A company or individual can be held liable for copyright infringement, unlike UGC, if and only if they can prove that the Data Base contains their infringing copyrighted material(s)... Which is... Good luck with that against Neural Networks, under American Courts.
The Canadian Intellectual Property Office registration set a precedent that the office accepts AI authorship and has "led to Canada gaining publicity as one of the only jurisdictions in the world recognizing copyright in works ‘authored’ by an AI," the application argues.
Ain't no way you're saying ai "artist"
Get with the times
Implying that any one of us can define "art" objectively
@@aisamsonrealai art isn't art by definition. poser
What if Ai generated an image and I repainted it in real life using oil paints on canvas, including driving the visual styles of the prompt, will the final painting on canvas in real life be Copyrighted?
Isn't that the same as an artist painting on canvas the natural landscape? Nature is not copyrighted, neither AI. The painter holds the copyright to his work in both cases.
So, can I use AI generated images in Midjourney for commercial use?
What about if someone will use my generations for commercial use?
Is it legat to use someone's AI-art for commercial? For example if someone has a paid monthly ai art magazine and uses ai-generated images by others?
Basically, depends what you did to alter the original AI image, likely. I'm not a lawyer though. Straight-AI image uncopyrighted-altered images need to take copyright into consideration. If you can take an AI generated painting and turn it into an ink drawing which looks very different from the AI generation, for instance-deserves an artist cioyright.
lets say i make a SD model with my own paintings, and then use image2image with sketches, photos, 3D renders, to make new paintings, is that copyrightable?
great great video, i was talking with a neighbor about this since her husband is a draw artist. Her husband was sad about how easily some things could be done now and claimed as the work of someone else.
I was like ok but if you go trough all kind of loops and making combinations of different styles which those artist couldn't even do themselves + some points you made in the video... why couldn't it be copyright? in my opinion it should.
Just bluntly type a scene you want and add leonardo style should be not be allowed to be copywriter since you just copy the exact style of him. Also the hard work those artists put in to make they're own style should be respected and protected imo.
I felt sort of bad by just testing some of they're styles in midjourney if iam honest haha.
Under USA Copyright Law. Art Styles are not copyright protected.
I can easily recreate any Art Style 1 to 1 or have an AI do it for me. It matters not in the law.
Only that... If I'm replicating Marvel Comics, that I don't make Iron Man because that would be Copyright Infringement.
Most artists don't know a damn about USA Copyright Law in detail. I do, since it's a part of my business to know those details for video game development.
@@absolstoryoffiction6615 so how would you write an midjourney prompt to get features of ironman or a mix with him and another super heroes without getting ironman? Since you can use the art style. Just marvel artstyle?
@@MeesterGgaming
Legally speaking... As long as you do not make "Iron Man" by Marvel/Disney when using AI Software or drawing it out yourself. Then it's legal under USA Copyright.
Just don't call it "Iron Man" since that's Trademark Infringement and don't create any derivatives of Iron Man.
You can have your own "man in a robot suit" even to the complex technological level as Iron Man but avoid doing as I stated above.
(This is old legal stuff that AI Software never changed nor added upon, in America.)
Art Styles are the same since Art Styles are not copyright protected but the Iron Man IP is copyright protected.
You can certainly make your own Marvel / DC Style Comics but avoid using copyrighted characters.
Avoid Fair Use too, since that legal defence won't work in this subject, specifically. Unless you're making political commentary etc. where super heroes were used in that manner most often. That's protected under both USA Fair Use Law and USA 1st Amendment.
As a photographer I don't mind at all when my images are used freely for personal use. I think it is rather miserable to hide something on a hard drive that might add to someone else's creativity. On the other hand if someone has a consistant habit of just stealing other people's work for profit with no conscience that should be a crime whether AI generated or not.
The lawsuits are interesting. The stock libraries do represent a lot of artists who in general want their work protected but I hope it is overturned. I would not mind if elements of my images were used. AI artistic programs that have good intension should be able to use any element already created. It's a little bit like saying if you go for a walk don't take a photo of a rose in someone's garden. Let creativity grow and thrive!
It doesn't matter. Publication happens because individuals can't make movies. A "good movie a friend saw this weekend" can just be a prompt + seed. Copyrighted works will lose market share fast, because of copyright.
No, the market is quickly becoming saturated with AI stuff, market saturation devalues everything.
@@beth1979 AI art doesn't "devalue". It had near zero value already. Only human art devalues, until AI is all that's left. (BTW I mean this as a bad thing if that isn't obvious.)
@@jonmichaelgalindo the big problem is that, the devaluation of human art will stagnate everything eventually, because without new artists, the AI can't evolve.
And if artists can't make a live with it, there will be less and less artists working full time on it, innovating and moving the culture...
Until AI work ethically with licenced and public domain data, art will keep bleeding
@@Mente_Fugaz Public domain data is a hundred years old. It's stuff about what society was like before computers existed, before nukes and the holocaust and socialism. AI trained on public domain data is useless to everyone. And AI trained on licensed data will be controlled by the wealthy elite, who will force artists to pay through the nose for the privilege of using the tools trained on their work. Pay artists pennies to steal their work, then charge them an arm and a leg to use the product. >:-(
No. The problem I see is saturation. Even before AI, we've been overloaded these last two decades with worthless, cookie-cutter, derivative wastes of time competing for our attention with the true gems.
What we need is a way to identify and amplify the impact of artists blazing new, truly worthwhile trails in art.
If AI is truly an all-purpose tool, maybe it can do that. An AI that can look at something and say, "that's not random noise, and it's something truly new and valuable." But then again, maybe not. But saturation is the problem, in my humble opinion.
@@jonmichaelgalindo i think, satuartion is a consequence, not a problem...
The problem could be the lack of motivation to innovate...
The inspiration to make something unique and fresh.
The real saturation comes with AI , not with children trying hard to draw my little pony characters,
Because a lot of new talents raise making that stuff, and in the future they develop their own vision creating authenthic stuff
Thanks for the update! Appreciate that, really.
As an artist, I am not the least bit afraid of the unstoppable nature of progress. In fact, I find it utterly fascinating. To me, it's simply another tool in my artistic arsenal. Let's delve deeper: art is essentially a concept expressed through a chosen medium. In the past, artists commonly used water or egg-based paints until Jan van Eyck introduced oil paints, heralding a new era in art. It's an ongoing history in the making. While we may have differing opinions about AI as a creative tool, we can agree that it's becoming an integral part of the artistic landscape, particularly for the younger generation who see it as the future. Is it fair? That's a subjective question. If I were the first to paint with oils and someone surpassed my skill, yes, I might feel disappointed. However, just like the color blue was not always available to artists, norms evolve. Not everyone can master oil, watercolors, or digital software like Adobe. Colors and lines are not limited to a single individual's ownership. They are like pre-mixed pigments, ready to be used, and in that sense, AI can be seen as a pre-mixed set of vectors, expanding the creative possibilities.
Interesting point!
It's not a natural process of progress, it's ideologically driven. One rooted in the desire of the lazy to plagiarise the works of others who actually contribute what those lazy want to society.
Bullshit, son
The only ai I'll use is when I tell gimp to show the grid for my drawing's boundaries.
I am also an artist an I used A.i. in my daily job. Let me tell you this: If I say "hello" to you and instead of hearing your voice I'd hear a generated speech based on generated words, I'd consider this disrespectful at best. Same goes for a.i. art. It's not art if it's not made by human, and if there is more human than a.i. in the work, then it's not an a.i. art anymore. Personally, what a.i. does for me is it reduces me to a retoucher, who paints over a random picture. This is not art. This is garbage. When using a.i. generated images, you are not becoming an artist. You are becoming an appendage to an algorithm. If you want to be an artist, go learn actual art, which involves human skill. And no, it's not the same as using photoshop. Photoshop doesn't create composition for you, it doesn't create your palette for your, it doesn't create your brushwork or your personal style or anything at all. Your analogy is wrong.
Art, among many other things, is the way we communicate and share ideas and visions. When I see a.i. generated images, all that is communicated to me is "this person can type words", because I fully understand what part of this "art" was actually created by a person.
This discussion about the arbitrary line of where human involvement starts is very similar to the discussion we are having in education: how much work does the student have to do so it’s not plagiarism. Is it enough to summarise/ slightly change an essay that AI has written?
Don't worry in time people will start to use AI for self learning no need for teachers, Oh hell we're all doomed, RUclips and AI and learning by your mistakes it's the future.
@@celtshaun1427 “no need for teachers” would be bad news for me…
@@celtshaun1427 with AI there won't be any mistakes to learn from.
@@celtshaun1427it's the future bro this shit will put humanity in the drain it's like the AI bros only reply this is the future you luddite like do you guys have any other response other than you are obsolete like bro do you have a brain we creators made what you call art work in which you guys mish mashed and stole so don't come here talking to us as if AI is some godly entity when in reality it's just a pattern recognition technology that some corporate Dev and CEO make to steal and profit from us artists
It all starts somewhere...
These are the first steps into a larger world
Nope... Given cybernetics and true AI... You're rather too far behind, Humanity.
i struggle to understand the difference in definition between that not being human authorship, and other ways humans use machines to create art. i am really against currently copyrighted material being used for this tech but just dont understand why the tech in of itself is being demonized
Such great, technical detail, thank you.
Glad you liked it!
It is curious how the laws do not limit or regulate the application of robots and artificial intelligence in the productive area (industry, agricultural sector, digital companies) nor do they dare to indicate to an entrepreneur what and how to apply it in their businesses, but they do generate all these derisory observations to the artists.
AI is like any brain memory or a book. AI learns and keeps a memory like any person who feeds the memory with info, knowledge and experience. Learnig is not plagiarizing. Otherwise, all the people, conservatories, universities, books, etc. would be violating copyrights. Plagiarism is a horse of another color and it´s easy to demonstrate.
I don’t think we should protect certain celebrity styles because those who are rich on top and corporate their style and Disney can do the same corporations do the same and limits creativity
As the chess game of legal discourse regarding artificial intelligence continues to unfold in the hallowed halls of jurisprudence, the unrelenting tide of technological evolution pushes forward, undeterred. Lawyers and adjudicators, steeped in precedent and principle, strive to wrestle with the complex connotations AI introduces to our established legal infrastructure.
Meanwhile, the sphere of artificial intelligence itself refuses to remain stagnant. Instead, it challenges itself and us, pressing forth at a rapid pace that it's nearly tangible, ceaselessly advancing towards newer, more sophisticated echelons of innovation.
Every sunrise heralds new frontiers in our exploration and understanding of AI, thrusting us deeper into an epoch marked by path-breaking revelations and boundless potential. Hence, as legal deliberations rumble on, the tireless march of artificial intelligence toward tomorrow never misses a beat.
hey, what if you turn your ai art, & literature into an NFT; do you own it then?
Photography? Produced by a machine!
You can press a SINGLE button and get a copyrightable image from a photography "machine".
Bing Bang BOOOOOM
Beyond that camera traps don't even have a human present to capture the image. Wildlife photography etc.
Except it actually takes skill not just asking a machine to generate an image the photographer has to edit and find the perfect timing and subject just to find the photo
@@paulaumentado1588 Yeah this is the point. Both AI art and Photography can be done with or without skill. (this actually applies to all art forms)
@@dmreturns6485 photography is made by capturing moments in life in which that could stay in time permanently that is via image photography is beautiful because it captures life as it is whilst ai on the other hand is basically your ordering mcdonalds and those who claim to be an artist in doing should be questioned without a doubt as ai is stealing from artists not making new content
Fantastic video Sam, great job.
Is language copywritable? Surely the prompt language is the creative part here. And I spend hours tweaking prompts. Tell me that's not a creative endeavour akin to a client describing to an artist
well... you can copyright a book,
is not the language itself,
i think is more closer to a code,
you can't copyright a code in programation.
a prompt is basically the same
Its by no means on part with what wou would need to do to achieve the result withou the AI. So you cant even say you ate 50/50
So.... you have to keep the prompt... especially with a complete, complex prompt providing creating details... (without saying the prompt was made by another AI product.. oops.) ?!?
If an art director prompts me as an illustrator with ideas, then curates my sketches or art, and suggests changes, nonetheless the work was done by me. I would have the copyright unless given away in a contract. AI cannot make a contract, and cannot hold a copyright.
It will largely depend on your local copyright laws. While the US does not extend copyright privileges to computer-generated work, there are other countries that do because it is an element of their copyright laws.
For instance, in the UK if you create work that is computer-generated, then you retain copyright ownership of that material because the law permits it.
I also think that the argument against AI is not as strong as people make it out to be, though I understand the frustration with it. People are upset because AI was taught by essentially looking at other available art... But that is no different to every art student, every artist, every designer, who at some stage or another, has studied works of art with their own eyes, or who has sat in a classroom learning techniques using other art as a reference point.
The only difference is that AI can do the tasks at a much, much faster speed. So when you strip it all down, it's really no different to comparing a human vs a calculator.
This feels like an outdated law before it has even been enacted. With all software implementing AI in one form or another in the next few months it will be impossible for any structure to look at individual copyright claims. The volume will be too large and I feel sure we will have to revert to an assumed copyright position for all creative work unless expressively stated as 100% AI.
Sure, but it can still be illegal and individual rulings on cases can cost those ppl dearly, same as it happened with movie piracy which has declined in countries where persecution has become on this case to case randomized base.
@@LorettaBangBang Hang on we were talking about if images an individual generates can claim copyright. Not be given copyright is not the same as breaking copyright. What litigation would you be liable for if nobody owns the copyright? Movie piracy was clearly a breach of copyright because someone owned it.
It's not a gray area, you can use Ai as a tool, if the final results doesn't require use of other tools or a skill you have it's not your work, be honest what you actually contributed to the final outcome.
At some point AI will let you decide everything that is on a image (light, camera angle, subject, composition) exactly like you want it just like in photography. Then AI will behave exactly like a camera. Then we will have another conversation.
I asked ChatSonic and ChatGPT about how the app's gonna know and control copyright of the music made with this system, they saind me "it's a huge challenge for these companies to do it". That's why I am confused, for example, "I" created almost a hundred pieces of music on an A.I. app, but only 10% are listenable and these 10% is in the middle of trashy music. How companies will know if the person took some of these crappy things and tell the world it's yours? It's a bul*****, because if you pay, you can have the copyright, so the problem here is MONEY, elite is going to absorb and make more money with A.I. , pay for this and nobody is going to know.
'Prompt Artist', a common 2024 job title
Nop
AI is but a tool that depends on human input much like any other tool such as synthesizers for modern music which can be copyrighted.
Couldn't you put AI art under an IP?
Controlnet would override that if the user came up with the condition themselves.
This is a nice and instructive video.(AI and Virtual Reality: Creating Immersive Experiences with Technology) I hope there is a lot to learn from it.,,,,,,
I hope so too!
so its still ok to use is not just for a tool
“Actually formed” the traditional elements of authorship… isn’t that what a camera does?
I just wonder how the hell they plan to enforce this?
in the future you will make a decision to buy machine or handmade product.
I guess the amount of alteration required to define a work as legitimately different enough from the original output to qualify as copyright could be expressed this way; imagine that someone takes an image you have created - how much alteration would you feel they would need to make to your image in order to claim copyright of that 'new' image for themelves?
Most artists would feel that simply taking their work and making minor changes to it would not be enough for the copyright of that image to pass from them to the person making the changes. There would need to be substantial alterations to the orginal work before it was acceptable to define the resulting image as a 'new' work.
If we apply this to AI generated images again it would not be enough to make some minor alterations in order to claim copyright- substantial changes would need to be made in order to claim that the resulting image was a genuine new creation that qualified for copyright protection.
I do wonder however if the ability of AI's to create iterations of exisiting images renders this whole debate null and void. For example, suppose I see an image you have created and would like to use that image in a book I am publishing- I could perhaps licence that image from you, or I could simply use your image as prompt to generate something similar and use that instead- it's not clear to me that there will even be a market for image licencing in the future so the issue of copyright may already be an anachronism not worth worrying about.
the copyright on that case only protects the modifications that you made, not the whole picture
Thank you for this video! Help me a lot
If it's in the internet, its under public domain. It is everyone's interest to protect their intellectual properties, it means do not published it that you're not willing to share it to the world, because you have specific consumers who are willing to give you money.
Amazing, This is just a modern Luddite revolution.
The logic here is a little funny. So you do a lot of work through ChatGPT so you can stay a small team and dont need to hire other people. But somehow doing the same with CG or visuals and just generating them rather than hiring someone is a huge etical issue? Any team or company can claim they need to save money.
Ai is the future and it is futile to try and stop it with laws from a time when it didn't exist.
It is also a tool to express your fantasy and it still has a way to go to be a very good tool. There is a huge amount of work necessary to refine an ai-image and a lot of knowledge in ps and/or other software to do so. If i learn to draw, it is still my fantasy coming to life.
Did anybody complain, that digital art is not real art, because you have ctrl+z instead of a manual eraser?
It just makes it easier to bring your ideas to life. Your art is not how you do it, but what is on your mind.
And that is why it is still not the best tool, because there are limitations to what you can do right now. To get to the expressions of the pictures in your mind is a long and hard process with a lot of redefining and frustration.
There are good ai-artists and bad ones and you need to learn how to comunicate with the machine to be good.
The demand for this not to be art is just fear of artists to lose their jobs, while in reality they just need to adapt, take this new great tool and to express their minds like they did before.
Creativity is a gift not everyone is blessed with. This is what makes an artist, not the tools he is using.
The problem with it is that, in fact, the real huge ammount of work necessary to make AI to look better and evolve, comes from human artists whose works are being used without consent to create derivations of them...
It doesn't matter how good you are on AI,
if there's no new artists to feed the AI , it will stagnate what it can do eventually,
also, you can't build an authenthicity through AI,
because is not something you made with your own hands, you can't build your own style, because literally you are using style from human artists
on digital art you can build that authenthicity, that's the difference.
Also, there's no really need to learn a lot of those softwares,
literally we don't have 1 year since it was released, and everyone is expert now lol
in 1 month you can learn everything you need and more,
the hard part is to have a good Nvidia graphic card, and a good pc
@@Mente_Fugaz it is a tool, which is still in development. To try abd stop its development is backward thinking.
New Art and new artstyles are always influenced by existing art. Every artist got influenced by other artists, combined different styles and gave something of their own. This is no different, besides the machine is influenced by a lot more artists than a human could ever be.
But say it is like you told me (i talked a lot with friends about that exact topic and i had the same thought) and no new artstyles come from ai: Then there is the future for human artists. Real creativity.
And there will always be customers, who'll prefer human art.
Also no human artist will have to do jobs which are under their dignity. Bad paid jobs to design a logo or a flyer without any creative freedom.
Look at the market and how art and design is paid right now. If you're able as an artist to do these works in about 3 clicks and it pays, why would you complain to have more time for your real art.
The problem is not the competition by ai, the problem is that being an artist is an underpaid and unthankful job rn and people are afraid to lose the little bit they got left.
@@Cruentus the machine is not
" influenced" man
Is just an algorythm,
You are using works that you don't own to put them on a blender of data, is literally using your actual work to decompose it's data to make something without your consent,
Is the actual work of the artist being used without consent ,
Like when you through fruits on a blender and then you say " i made a smoothie, this is my smoothie" when you stole the fruits from another person's basket,
Humans learn how to cultivate their own fruits, that's the difference,
It's simple , if someone has amazing work and an identity,
And disney want to work with it,
They must pay for the artist and have the consent to use his work to put it on a blender.
@@Mente_Fugaz look, i'm not trying to make you angry, i just want to shift your perspective from the fear to get obsolete by new technology to the massive benefits you'll have from said tech.
Like i said, real creativity will come from human artists for a long time. It is futile to try and stop new technology. Say you'll be able to stop it in the western world - then it will be developed in asia.
You have to adapt to it and use it before anybody else does. It might seem to be unfair, but in reality not too many people are interested in it right now, so it is a great opportunity to start with it early on.
I'm not arguing with you, i want to help you to see it differently and more positive.
Also it is a bit more than just a blender. It is a pattern-recognition and -recreation. One could argue, that real intelligence does the same, but with more intentional understanding of the result.
@@Cruentus AI doesn't need to be stopped,
If someone wants to make AI models with licenced and public domain data, or your own works,
it will be a great tool,
and you should be able to copyright your creations with it,
but right now, what everyone calls "AI art" is derivations of stolen art, that's the issue here,
AI can't evolve without new human pictures to feed it, or at least, it can't evolve into something that looks good (it would just evolve into abstract stuff, because it functioning doesn't follow any principles or fundamentals, it just follow the concept of replica).
you need to understand that, if there's no new artists working with fundamentals through their own art,
AI will stagnate,
and if there's no motivation for new artists to keep learning about how to create pictures because they can't make a live with it anymore, and is more profitable just grab the work of other artists and in seconds make infinite derivations of it,
Art itself will stagnate, at least in this specific topic "illustration"
I don't find the chatgpt poem a good example at all. So if I tell it how to rhyme, and keep adding a lot more detail, am I not in a major way leading how the result will end up? As Samson mentions, there is a big difference between saying "make a nice image" and more detailed promts, like "make a pencil drawing of a girl with thin lips, eyes further apart than normal, messy blonde shoulder length hair, freckles and a mole just to the right of her nose-tip". Of course that could be taken further, I could have manually drawn a landscape and use photoshop to insert her into it? I only get to copyright the landscape in that picture, or does it now constitute a completely copyrightable work of art?
This is obviously new to society as a whole, and we have a long run to cross yet in a legal enviroment... but proclaming that just by a machine producing a material isn't copyrightable is just as saying that a photograph made by any camera isn't also copyrightable. BUT IT IS!
Any photographer can proclame ownership of any picture that he takes althought every single pixel of it is made exclusively by the 'machine' Camera. And the photographer also (most of the times) didn't do a single bit of a thing onto the scene that he takes a shot on.
So why he can copyrigth that??
It is just because he bough an device that helps economy as whole? Instead of the current state of AI Ecosystem where you can generate a ton of images even for free?
I believe that a good starting point should start from the companies that provide AI Art generation in making clearer as possible that any image generated in a free plan isn't free for commercial use.
But at the moment they just explicited that the images aren't in your 'private channel'.
This is what makes the regulation difficult in my opinion, because anyone can copy any image made on those free channels and proclame that it belongs to him/her.
And how he/she (the real owner even) could ever prove that?
Thats why is important to make it clear how a free or paid plan should work. Then would become much easier to prove authority and ownership of an AI Art generated.
You don't own it, the AI does.
i love your content and feel free to delete this comment, but the technique of slow zooming on every shot is distracting and is drawing attention to you rarely blinking. trying to be constructive not critical
Here Consider this: you have a creative idea, and you take paint, pencil or whatever tools you need to create that art...right. Well now let's look at AI we take a program that learns from a prompt (a different kind of tool like paint or canvas) and takes what you prompt it to do, which is in essence human creating the baseline of what is to be created. So, without the human aspect you would notget the images or video result at all!!! I think in many ways there are different kinds of art, and digital art has been around a long time, this is just a new form of digital art, which in essence creates a new kind of art! You take the prompt from your brain, and AI artist take that same prompt from their brain, and both manifest the image! Don't hate the way art comes to be! The idea of art is creative expression, AI and what you see as the norm in art are one and the same just a different path to accomplish the very same goal...it is still art, no matter how you get it, sometimes a closed mind only sees the negative outcome, try opening your mind. Computers & programs have been creating art for decades, in movies, aand other advertisements, what you see in your everyday life was created by a camera (essencially a small computer) or a computer to generate the effects, which a type of artist created...Do you think that art should stay stagnant while the art world and the way it is created evolves? Art is art and there are many forms of it, to condemn this form of art is simple hypocracy. Get over it, do what you do, and evolve or don't but, not allowing yourself to evolve, makes you blind to the reality that is the world around us!
I think artists should keep their a record of their prompts and the outputs from say mid journey then they’ll be able compare with the final image after all of their tweaking and expression. Maybe this should be part of the registration process.
If you think that’s tedious, try registering a collaborative music work with artists who aren’t on the same registry and use different names 😅
This would be task of the platform tbh. Midjourney's user experience is still hideous right now. Scrolling up forever to find something I did in the past? No thanks.
Ai summarizing as text whatever I was trying to create that day and collecting it by date? Yes.
I`m not sure you can replicate any results even if you used same prompts
You get the exact same image if you generate two images with the exact same parameters(including seed) in Stable diffusion.
That has nothing to do with legalities because AI Art Software is not own by the USA Government. So there isn't a legal obligation with a heavier legal merit to enforce that.
@@absolstoryoffiction6615 good point there isn’t a centralized policy. but the contention is the licensed work the Ai is trained on. The rights are owned by many artists of many different jurisdictions and even after all this time I see some companies outright ban or seriously hinder Ai.
Not to mention social media is updating their policies soon.
Govt might not be the only regulation at this point
Humans have no more control over the art that they create than an AI does, but y'all ain't ready for that conversation. A good enough FMRI can predict your thoughts well before you are even consciously aware of them, so can you really say that you control your thoughts? The answer is that, no, you don't have any control over your thoughts at all whatsoever. Every single thought you've ever had just appeared in your awareness whether you wanted it or not. Sure, they're based in part on what you see and hear and experience, but that proves even more so that you don't control your thoughts. So if you don't control your thoughts, are you responsible for them? Are you responsible for your ideas? Are artists any more responsible for their creativity than an AI? No, they aren't. So making a distinction between art created by an artist themselves and art created by someone prompting an AI is meaningless. The only distinction between the two is the amount of time taken to create them.
The problem with the copyright office's opinion is a prompt is human input, and a lot of work goes into creating good AI art. It takes hours of input and sorting through images to get something, which itself is creativity
We pay for it. And then we waste time to write a right prompt. Booom… It’s AI’s coppyright😂😂😂 how come😂😂😂😂😂
@@jphakphumromsaithong when you pay for an artist to make a comission,
the copyright is for the artist, because despite how much you specified what you want to the artist,
the result wasn't made by you.
You are just the client of the artist,
in this case is the same,
you are the client of the AI, you paid for the product
please make a video about wonderstudio vfx ai
They could ask people to donate photos and art to train with
Nice explanation!
Glad it was helpful!