How does Adam Savage feel about A.I. being used in filmmaking? What's something Adam never noticed about a prop until he saw it in person? What has the biggest impact on portraying emotion in a film: actor, script, score or lighting? In this live stream excerpt Adam answers these questions from Tested members @1J_Dellarosa, @ADHDMedia and @Koppa Kabanna, whom we thank for their support. Join this channel to support Tested and get access to perks, such asking Adam questions during live streams:
/ @tested
Adam's Empathy, Kindness Respect T-shirt: cottonbureau.com/p/2EUGSH/shi...
Subscribe for more videos (and click the bell for notifications): kzhead.info_c...
Tested and Adam Savage Ts, stickers, (de) merit badges and more: tested-store.com
About Tested: www.tested.com/about
Meet Adam in Person: www.tested.com/events
TikTok: / testedcom
Instagram: / testedcom
Twitter: / testedcom
Facebook: / testedcom
Discord: / discord
Amazon Storefront: www.amazon.com/shop/adamsavage...
Intro bumper by Abe Dieckman
Thanks for watching!
With thanks to Tested members @1J_Dellarosa, @ADHDMedia and @Koppa Kabanna for their support. Join this channel to support Tested and get access to perks, such asking Adam questions during live streams: kzhead.info/tools/iDJtJKMICpb9B1qf7qjEOA.htmljoin
“It’s the one that hides its intelligence…..”. 100% correct! Thanks for the video
There's no way to protect against the venality of late stage capitalism because there's no bottom to it. Hahahahahahahaha! I had to write that out so I don't forget it. I LOVE it. One of the funniest and true at the same time statements I've heard lately; just the way it rolls off the tongue. Thanks mister. I'm going to steal it and smile inside every time. Just beautiful and succinct. lol I agree with the 'point of view'. The point of art is to express something about humanity. And like Turing said, to paraphrase, we don't know what or how a machine will think once TRUE AI exists, because it won't be human. It can't be. So how can AI truly do art.
Adam would you please take a more concise + clear stand against "ai" generated images? You have barely said a word about the greatest art theft in the history of mankind and the scraping. The makers you are supposed to be uplifting would thank you in droves if stood up for us like Adam Conover does. We need more friends in the industry if we are going to protect the arts from big tech enshitification.
Truckit
I’m wondering, how political is okay with your Caliber of California? The police state seems very unforgiving for free speech & ideas ~ gold rush notwithstanding.
My issue with AI is it's not AI, it's an algorithmic regurgitation of art and hard work that has been stolen from the artists who created it. What's worse is people are now teaching it to imitate specific artists, which is a direct admission of the true motive - it's not there to create something new, it's there to create something that looks good, quickly, by people with no understanding of, or care for, true creative process.
Truth! I'm surprised this aspect wasn't mentioned. Maybe it's too controversial? I've gotten into arguments with people about this.
unfortunately, unless we have some radical advancement that allows for actual general AI... all you're going to get is "algorithmic regurgitation" as that's all they're capable of, currently. and actual general AI would be a much larger problem than the current spate of offerings. for oh-so-many reasons.
Exactly this. True artists go through years, decades even, of practice and failure, refinement and struggle to find their own style and hone their talent. They put a part of their heart and soul in everything they make. It was hard enough to get people to pay properly for art before this. Artists sometimes have thousands of dollars in fine art school loans to pay off, made all the harder by these programs that allow the random person with no real talent to punch a few keys and generate a piece that would take a real artists weeks to make. “Starving artist” was just starting to become a thing of the past and now, there are going to be even fewer real artists able to make a living than there was before. The average person just doesn’t understand how devastating AI is to the art world.
It seems there's a lawsuit going on against midjourney bc someone found proof the devs were working with DeviantArt to define artists they would steal from. Insane shit honestly. This AI art stuff really just brings so much shit with it
You haven't tried Gemini yet have you?
Okay, that is just messed up. The KZhead algorithm interrupted his story with a commercial about AI replacing your workforce for answering questions, social influencers etc.
Sounds like you need to change or update your browser
Cyberpunk Dystopia without any of the cybernetics.
CONTEXTUALLY RELEVANT (adblock tho!)
@@pyerack A boring dystopia. There's a subreddit for that as well.
@@pyerack Not for you.
As a 30 year member of the entertainment industry, primarily working in visualisation these days, Im not a fan of A.I for final images. Yes, its very useful for non skilled producers to give me a flavour of what they are after, and thats great. The biggest problem it is creating at the moment, though (and no one is really talking about), is that it has created a change in peoples expectations of how much work can be done in a given amount of time. Ive seen a huge shift in people thinking we can do it all much faster. A.I may be fast, but bespoke hand created artwork, whether digital or analogue, still takes time to get just right. Im constantly having to tell producers that a job will take a few days, only for them to say that A.I does it in seconds. This is frustrating.
See, while I think your ability to work under that condition is earnestly impressive, I don’t even agree with this as a use for it because it just tells me a director/producer lacks the ability to verbally articulate to you what they want; so instead they generate an image, and go “like that.”
@@MrMarsFargoBelieve me, as much as I don’t like it, sometimes it’s just easier because some producers are really bad at telling you what’s in their heads
@@annawildsville Absolutely. AI is a massive improvement to workflows. But just look at the creative output of any LLM. Its usually cliched stuff, with barely anything interesting going for it. I ask gpt4 to give me ideas on scripts, and it usually throws the most predictable stuff you'd expect. Its great for mundane tasks, or trying to find multiple angles on an issue, but currently its incapable of replacing anything that will be under heavy creative scrutiny.
In my experience, using ai to get a baseline rather than making it from scratch can get you 80% of the way there, maybe more with serious futzing, very quickly… but unless you’re doing some seriously low quality stuff - work that works probably be complied from stock assets and templates - that’s just not particularly useful. Making something from scratch, that first 80% progress takes like 10% of the time, and the real time sync is in getting things good enough for serious production work. With the exception of the professional tools like Nuke, AI is going to output things to a crappy final delivery format, like compressed video or image, and make waaaaaaay more mistakes doing it. It’s like having the worst intern you’ve met try and make a completed first pass at all of your tasks and being forced to work from it for your final pass.
yeah. _‘ai could do that in a few seconds’_ in these cases usually means _‘ai could amalgamate some prior examples of similar work into an averaged out collage of roughly what i’m going for, which looks impressive on first glance but breaks down on closer inspection’._ to the untrained eye, ai generated images can sometimes seem like production ready final pieces (especially if the person generated it themselves, since that tends to create a sense of ownership over, and subsequently, defensiveness of the image) but we’ve all also watched the fits audiences throw when the VFX in a new blockbuster movie looks wonky, cheap or otherwise half baked. if someone noticed two background extras in a Spider-Man film merge into each other? or the length and colour of the Terminator’s trench coat fluctuating during a shot? the mockery would be _brutal._ it’s the kind of thing that can permanently damage a franchise or studio’s reputation. so it’s left to the artists, craftspeople and technicians to delicately explain to the producer, or whoever, that while yes, their Midjourney png certainly conveys the gist of what they’re going for, and looks very attractive, if they were to do the whole scene the same way it would get derisively memed on so hard even Morbius would wince. filmmaking involves way more than having cool ideas, and coming up with keywords that describe an aesthetic. you actually have to practically, consistently implement those visuals and ideas on every level of the production. that’s the _hard bit._
My favorite statement about A.I. "art" - and I don't know who originated it, but it was in reference to writing - is this: If you couldn't be bothered to write it, why should I bother to read it? That can be generalized to any artform. If you can't be bothered to make it, why should I bother to consume it?
you want me to learn something for 10 years just so I can show you the exact same thing I was gonna make with AI? that makes zezro sense
@@incriptionBut the point is that if you get ai to generate it, you *didn't* make it.
@@incription You don't make "the exact same thing" with AI guaranteed. You prompt and get an approximation (more or mess precise but never exact) of what you want. I had that discussion with a prompter where I showed him a painting of my wife's avatar holding our cat in her arms. The picture was semi parody with her almost smothering the cat while having that "Oooooooh so cute" expression, almost like aggressive cuteness. The dude served me with like 20-30 prompts of a catgirl looking like my wife's character holding some cat looking cute with a blank expression. None of these images told the same story as my stuff, not even close. It all looked like nondescript overly rendered korean MMO shovelware. No story, no expression other than slight smile, the most neutral pose and framing possible. No apparent brush work, mind melting choice of colors (even though it copied my shit, looked like washed up neon). You can prompt a thousand times, you'll never reproduce the same vibe as someone who know how to draw I'm sorry. You'll always make some plastic approximation, it's never going to "say" anything.
@@Alarios711 eh, I disagree, I don’t think there is anything “special” about most real art, nothing that can’t be replicated by a machine anyway. When I’m looking at art most of the time I just want something that looks good, rarely ever do I care about the story or process of making it
@@incription Do you have some business? Here is example from my work. Solar fasade company ask me to make illustrations with company sollar panels. It should be artistic illustration of the house with specific their solar panels. That will be a big picture in the head office. So my question how you will make it with AI. How you will make right panels with right texture of that panel because it is not regular roof solar panels and they should be attached to the house in specific way. So you dont care but many do care. Blizzard are using AI and we can see how Diablo 4 is doing.
The same reason that whenever an artist I like sells a "sketchbook", I buy it right away. I love the process and the beauty in the exploration.
My concern with AI is not that it will ever be able to capture the true essence of the human spirit in art, but that it will approximate it close enough that it can create things that the trend following masses will eagerly consume, making the former obsolete in the market. I don't think AI could write a show as good as something like Deadwood, but it could certainly make the majority of the content on the streaming services today. It will be able to create songs that sound like dead musicians, but never be able capture the depth of soul in the music. My concern is how few seem to find that important anymore.
This. THIS.
don't be silly, it's still created by a human, until AI does stuff unasked, by itself, without a prompt, it is still human effort. i put it this way, we have no second thoughts about leaving ther planet to our children, we assume our kids will do a good job of life, but what is the difference between handing over the planet to our kids and handing over the planet to robots? it's kinda racist to think that robots aren't "as good" as humans, they will be built, designed, programmed by humans, and as long as they don't actively kill everyone, when there is no more biological life, who is to say they aren't just the next generation - unless you plan to live forever, what do you care?
It can copy/steal what's already been created. But it can not create new ideas.
Doomer
Humans are already doing this for profits sake i.e radio music. "My concern is how few seem to find that important anymore." that is indeed a problem that AI doesn't exacerbate, and might actually solve.
I find miyazaki summed ai up beautifully when he said, "i strongly feel this is an insult to life itself.i I feel like we are nearing the ends of time. We humans are losing faith in ourselves" The best thing about art is that a human made that. You look at this beautifully crafted master peice and are awed that someone had spent yests perfecting their craft to make this. Ai art is as interesting and valuable as a blank piece of paper.
You're taking an idiotic old man out of context there. But that aside, the demo(that was not really AI in any way) they presented to him was made by people who wanted the kind of text-to-img AIs we have today, and someone made that technology. Now what kind of person who argue that the people who created such technology lost faith in themselves? They made something utterly amazing and impressive. The person doing that is the one who lacks faith and perception. Humans will progress, we'll keep innovating. Being traditional is cool and all but you can say that it's also allegorical to being a dullard. Like why do you value traditional artists more than technical geniuses that make the process of creating art near infinitely easier? It's almost like valuing an angry chihuahua over an apache. Do you think novel AI generated images in Studio Ghibli's style is all garbage? Even if you cannot tell the difference between it and the real thing?
"Idiotic" are the people blindly supporting and/or accepting AI, dismissing the blatant unethical results and giving greedy corporations and individuals the nth "free pass" to screw all our lives more and more... Money is truly evil and people never learn...
I was trained in the British Film Industry in Art Direction back in 2003. My Teacher was Terry Ackland Snow. Art Director on Batman, Aliens, labrynth, Dark Crystal. I was just one of many students he has taught over the years. I feel proud to have been taught by him. I never had a career in art though, its more of a hobby now. Anyway, it breaks my heart to think that all these art jobs will be lost to AI art. A few words into a seach bar and you get hundreds of designs in seconds. How can people in artistic industried even compete with that? People say its just a tool, yes for now. However i feel AI will eventually totally replace the illustrator consigning the art to just a hobby. Thing is people need a sense of purpose and AI on many levels will take theat sence of purpose away not just in the creative industries.
You can have 1000 replications of _Guernica,_ but only one of them speaks out about a Spanish genocide.
@@wcjerkyartists cant make things like guernica if they are starving and unemployed
Artists can hurt people, but they are chose not to. They could just trace and photoshop existing arts, but they know better to not. Film maker can just spam C Reel Footage but that's just a filler and not the content of a film. Sculptures can just buy 3D printers and print whatever but everyone knows its the skill that matters. AI brought the worst of people upfront and watch as our kids grown up messed up watching directionless unsubstantiated recycled illegal copies made by scammers and thieves.
@@ilovepatlabor2Please elaborate on how employment status is relevant to making art. EDIT: Second question: how is employment status relevant to replicated art's inability to have a message; that is, to hold purpose?
@@wcjerky i think he it meant as: there will be no artists if they can't make a living.
It good to hear you pointing up the labor / commerce side - because tech folks often don't talk about it. So much of the buy-in and investment around this tech is with the end-goal of just reducing labor costs. As huge conglomerates like Amazon just run out of ways to scale, the only route to increasing profits every quarter is to make the workforce as cheap as possible. It used to be outsourcing call centers overseas - now they want to outsource it to machines for even cheaper.
AI is coming for all desk jobs. Eventually anyone working behind a computer will be replaced. Learn to swing a hammer folks, blue collar will inherit the world 😁
And not a single person will benefit from those increasing profits aside from the people at the top. The companies used our wages increasing to justify their increased prices to consumers. But those prices will not come back down after massive layoffs.
This was always the logical conclusion of capitalism. The US is appearing to do fine because of their neocolonial foreign policies.
'Neocolonial foreign policies' - what does this mean? You could have described the individual policies referred to, but instead we have this very tired pop-sociology catch-all term.
@@jaredwilliams1031 it’s more of a pattern than an on-paper policy. America finds an “enemy”, America sends troops to “liberate” that country, that country now is required to re-establish themselves under the US’s terms (including selling land cheaply to American owned businesses), the US let’s private businesses build factories there for cheap and unregulated labour, the businesses harvest natural resources and refines them to various degrees before sending them back to America. This is one of the reasons why Guatemala is in such a bad state right now. Brazil was left under a military dictatorship supported by America for several years. The assassination and coup in Vietnam (1963) committed by the CIA. They offered support in the Rwandan genocide (as in, they aided in a genocide that has been compared to the holocaust). And they are currently supplying Israel with weapons of indiscriminate killing. And I haven’t even scratched the surface. Does that answer your question?
"lacks a point of view" the best summation i've heard so far. But also: AI never makes the subconscious connection between seeming unconnected images/ideas which is fundamental to most of the art i enjoy.
Oh it does, and so do the people that use it. And don't worry, it'll get infinitely better very rapidly.
@@Ilamarea Peoplee using it, sure, but the "AI" is just a bunch of math, there exists nothing even resembling a conciousness. People are hallucitating about what current-day "AI" is far more so that the technology itself does. It's all machine learning, something we've done for _decades_
@@Koushakurhow would you know? We can't even clearly point to the process that creates our "conscious" experience yet. The concept and idea behind neural nets is to mimic the process that works in our brain. Obviously they are different, but it isn't just a bunch of "math" in a vacuum.
@@Ilamarea I have yet to see it happen. as of now it's mostly the art of the common denominator. i guess it also depends on what you consider "art".
@@Koushakur So is the human brain
My biggest gripe with ai art, is the fact that no artist was informed their art was going to be used to train the ai that will ultimately replace them. There was no expressed permission and I bet the majority of the artists online whose art was scraped, when they put creative commons on their work, the intention was for other human artists to take inspiration. The big issue is there was no protection for these artists. Another element of this to consider: Peoples voices and likeness aka (faces) are also not protected, and there is nothing stopping these people from taking your face or voice and using it to create faces and voices. I believe there should be protections automatically, in the case of someones likeness or voice, from the moment of birth, and anyone who wishes to use someones likeness, or voice, should have to ask permission before their allowed to be used in training these Neural network models.
But I mean, that logic doesn't work. That's like saying "human artists" must disclose every piece of art they've ever laid their eyes on. Every musician must list every single song he/she has ever heard on the album art. Etc.
@@user-lv6rn9cf8m is that what I'm say? Their is a difference between taking inspiration and tracing someone's work, the art community as a whole has that as one of their unspoken rules, there is a community there that understands each other and collectively have decided what is permissible in the context of artistic inspiration. the training of neural networks on mass, was not something that that community was prepared for or understood, it just happened, I believe the overall consensus is that it was not permissible, and to not even engage with that community, before scraping their works of the internet, is morally bankrupt. I don't believe that training a neural network and the process of being inspired by other artists is even the same thing, people seem to think they are, but where is the research and evidence to back that up? I mean we like to describe it like it's a system like the neural pathways of the brain, but perhaps we are merely choosing that language to justify, what amounts to stealing.
@@vlander1992ableYou don't seem to understand what ai is or how it generates art. It's the same thing. I am aware that huge portions of the art community have a really poor understanding of ai art. Where is the research to support what you are saying? There's plenty of evidence and research to support what I'm saying. You can for yourself determine that it's not stealing, it's creating new material.
@@user-lv6rn9cf8m So If I was to download or record every movie that is available online, and used a neural network, trained it with those movies, video and audio, to generate a movie, it's morally permissible? Honestly I don't even care about the specifics I only really care about the moral argument and you haven't even addressed it. This is the same thing that's happening to every artist online, but the difference is they probably aren't protected legally like most movies are. So where do we draw the line, does copyright protection protect art from being used to train ai? Does every artist that no longer wants their work stolen, now have to jump though that hoop. The real problem is the law hasn't caught up to it yet, so until we can figure out some way to exclude peoples art from being used in training models, artist's really don't have a choice whether their work is used in this capacity.
@@vlander1992ableAbsolutely. Why wouldn't it be? No one is having their work "stolen". You seem to not understand the technology. I don't understand what you are saying even. Why would we exclude people's art from being used in training models? That's like saying people can't look at famous paintings because they'll learn something from them that they might incorporate in their own artworks. Which yes, is happening and is not seen as an issue. We're not talking about copying or plagiarizing here. AI isn't stealing their art, it creates something new out of what it has learned. Artists doesn't have a choice in whether their art is being "used" to train other artists either. Same thing. Or me, I've listened to music before. Am I now not allowed before to produce any music because I'll be influenced by what I've heard before? Should everything just pop into existence from nothing? I mean, that would be kind of interesting. But that's not how humanity deals with art.
"The computer can't tell you the emotional story. It can give you the exact mathematical design, but what's missing is the eyebrows"....Frank Zappa.
Frank was right back when he made his comment, but his comment no longer applies now. Just check out what is possible with Suno (generative music service). It has become extraordinarily sophisticated in a very short amount of time.
I mean, gotta love Frank, but that's only true until it isn't. Also, the analogy falls apart once you realize that eyebrows are hardly necessary for the emotional story. Ask NoHo Hank.
@@minhuang8848 ... Well, one can analyze the humor out of anything I guess.
@@buddystewart2020 If you come to a discussion and your contribution is a joke, then it should hardly be surprising that people focus on the meat of it rather than just saying "haha, it succeeded as a joke because it's obviously silly".
they are really claiming eyebrows aren't a good analogy for emotional expression 😮
Revelation! I so relate to your father observing his art through a mirror. As a wedding and formal wear alterationist, I use the mirror to observe any fitting inconsistencies that somehow don't appear when I look directly at the garment. It make me see the garment as a whole, not in fragmented details.
AI as a final product is for people who think themselves artists but don't want to take the time to do the work, like CEOs who think they're engineers because the people under them solve the hard problems.
Very good point in that statement
People that talk about how they have to craft the right prompt and it's hard have clearly never worked with an artist to commission a piece of art before. It's basically the same thing. You're telling the AI (artist) what you want (prompt) and communicating with the AI (artist) until you get what you want. You're not an artist because you can talk
@@xXESproductionsXxThe meme that was going around to this point said "you can't make a prompt in 10 minutes" and "you can't master this in just an hour" - which suggests they have never engaged with artists or art before. How much experience do you really have with a new craft if you think it takes a little over an hour to master?
I recently saw somebody put it as that AI is for ideas people. Anybody can be an ideas person, but they don't want to do nor respect the work it takes to turn an idea into a reality.
@@Khotetsu 👏👏👏
@3:19 I work in telephony, this reminds me of a project many years ago with an early voice recognition system, based on phonetics, nothing to do with AI. All through the design/planning stages, line staff and low-level management to a person brought up the fear of job replacement. Every time, the project manager and sales person assured, "This is a tool to make everyone’s job easier, those answering and transferring the calls today, will be freed up do more important tasks." Late in the project, in case there were technical questions, I was included a meeting with some director level person in the organization. It was very depressing and angry-making, uninterested in the details, his only question was when the head count reductions he’d been promised would finally happen...🤬
I presume he was very upset with head count actually rising? Technology is a great thing that can make less people make more, but to scale profitably it will usually require more people than before.
@ShadeAKAhayate but it will requires a different kind of skill set. The industrial revolution is kind of proof of that.
@@NotTheStinkyCheese It always does. The only question is how different this skill set would be. Usually, it requires deeper understanding of the subject -- like, driving and maintaining a horse-powered wagon and a horse differs from driving and maintaining a truck.
What about the simple fact that companies like OpenAI flagrantly stole copyrighted materials to train their LLM's and are now directly profiting from that? I mean, that's the Elephant in the room with all AI tech is it not? Just because something is "on the internet" does not mean it's public domain.
factually, murky at best. The De Facto captum manu flammae offensae for anything out in the public, and from a, frankly, common sense standpoint, if you don't want it mimici'd don't have it out in the open. you can't steal a perfect photo copy. For one it's a , murky, and futile at best moral and legal area. And ultimately futile to winge that Johny Mc Clickbait heavily borrowed from anime move cover 9999. You can, however, _potentially_ make a some money from the right wheel works: one of a kind physical, or else it's just futility at its best. OpenAI being trained or using meta data (the terrible rough discrption) of your work very well (" ") is, bluntly, the samething a human might do...at speeds we can't achieve, not yet anyway. So it's either be flattered, or tilt a few windmills yelling to the heathen gods: wtf!
Because copyright up until today only protected your right to your piece of art being used directly by another entity public facing. Everyone has the right to download any piece of work you have put on the internet. That is not copyright infringement. There have already been several cases where companies have been sued for scraping images off the internet and using them in models and those companies won every time. copyright protects your art from being resold face value. If that art is meaningfully changed then if become a fair use. It will be VERY difficult to say that AI output is significantly similar to any art piece to constitute copyright infringement. The distance between the trained artwork and final output are just too far removed. Now that is not including accusations of companies using pirated materials to train(and they should have the book thrown at them for it), but downloading something on the open internet is not pirating or protected by copyright from being downloaded and used personally.
@@miclowgunman1987 "Used personally" is the operative word, though, isn't it? It doesn't matter if the training art isn't being resold directly -- it was still used *for profit* by those companies because they wouldn't have a product otherwise. That's why I think the fair use argument falls flat.
@@jamesprayyou can legal use a copyright work for profit if the final product is transformative. I think it's pretty safe to say that an AI model (the model itself, which was created by training on the images) is significantly different from any of the pictures used to train it (Its not even a picture, its a statistical model). I will admit the output is being litigated still, but so far everything remotely close has been thrown out and that anyone arguing otherwise has a huge mountain to climb to change past precedent. But I follow image generation a lot closer than i do LLMs so there might be something to recent ChatGPT lawsuits.
At this time you cannot copyright a style. There is no way to put the toothpaste back in the tube for AI. We need to adapt.
AI is currently being used to frighten the working class into accepting low wages, poor treatment, exploitation. "We can always replace you so you better accept this job while you have it." It's completely anti human in every sense of the word. So it's incumbent for all workers to stand up against AI and against our collective exploitation. Only through struggle can we achieve dignity.
Nah lol. That's just deferring the problem, what really matters is grasping the real underlying issues by the stones and getting it over with. The businesses that just do it for the sake of arbitrary workforce reduction when it isn't even remotely feasible are going to just vanish because of their lack of business savvy. The businesses where it actually works and leads to orders of magnitudes more revenue (which just means jobs change, not that people get laid off, mind you) are right in doing so, like the translation industry. Which, in case people have been living under rocks, has been pretty much demolished overnight, with a huge amount of (freelance) workers having been delegated to quality control work long before LLMs actually hit the scene. We need to make sure that dignity is preserved by properly allocating resources, of which there are plenty, to the hundreds of millions of impending jobless, who will still be able (and willing!) to then continue working in any capacity THEY want. Work isn't ever going to die, none of it. We'll just hopefully end up at a point where nobody has to worry about not getting to sit (or, lmao x2, stand like in countries) behind a register and slowly killing their mind and body in the process. Most of this is amazing news, people just haven't come to terms with that a complete revolution is inevitably going to follow.
Cause business wasn't already exploiting the working class and wouldn't have done it if it weren't for the invention of AI. AI is a tool. Arguing against AI is like arguing photoshop would put illustrators out of a job 30 years ago. The issue isn't AI. The issue is the suits who will use anything they can to increase their profits at the expensive of those who do the work.
your opinion is wrong and wildly shortsighted. AI can be used to replace human jobs, so HUMANS DONT HAVE TO WORK. if ai bots take nearly all human jobs and humans are destitute and poor, that is a failing of capitalism, oligarchies, and human greed. AI is like startrek technologies where all human basic needs can be met far easier than any point in history. BTW what, you want to ban ai? lol. ok good luck convincing china, russia, northkorea and multinational corporations to all agree.
open borders are hurting the working class way more than A.I. A.I. is harming the office worker middle class.
Workers need to stand up against their exploitation, like they always have. But really, AI will be a boon in the long term. But society needs to evolve significantly before the introduction of something like AGI would be a blessing not a bane.
"the thing itself is not as interesting to me as the mind and heart about that thing" Completely agree, and this is always where AI can fall short. AI is just inherently disposable. Seeing art in context is incredibly important. For example, if a wooden box was made on a factory floor, all CNC and robotic assembly and you got it for very cheap, you'd probably use it for simple storage and not think twice. If a wooden box was made by a professional woodworker by hand, you would inspect and care for it, look at how it was made and probably use it to store something much more important, as well as display it and not just stash it away Imperfection in industry is a manufacturing error. Imperfection in art is what makes it human. Knowing the love and care put in by someone to make something, especially if it's someone you know making something specifically for you, is incredible. If there is artwork that captures me in a gallery, I read the notes on the painting to get more insight on the artist. I'll study the brushstrokes or linework, and see the human aspect. AI will get to a point where it can mimic art perfectly, but context matters in art as much as the work itself. It's why the Klein Blue painting and his blue era has such impact when seen in person. Looking up pictures is pointless to the work because it's just a blue rectangle. Seeing it in context, seeing that specific colour that was handmade and reading more about it makes it matter, and makes it an international art piece, even if you dislike the painting/modern art in general
that is what people don't get about art, specially modern art, the importance of context. the history and context of an art piece, and the person that created it, is just as important as the art itself, in some cases you can argue that the name of the piece is the art, not the piece itself, like "Who's Afraid of Red, Yellow and Blue", the piece wouldn't be nearly as interesting without the name and cultural context. But AI don't have context, its just a semi-random spread of pixels that happens to look like something because we trained it to. you could say that the prompter has the context, but they are just the commisioner, when you look at "The Last Supper" you don't prise Ludovico Sforza for the idea, you praise the actual creator of the piece, da Vinci.
But I mean, the way I'm looking at it - ai is a tool like every other. You're talking as if it's agi art. It's not. The human is still the artist, just that instead of a brush you use words.
@@user-lv6rn9cf8m the prompt writer is no more an "artist" than a person commissioning an artist to draw something for them is an "artist." Paint splatters on a canvas made by a chimp or elephant are more real art than any AI-generated image.
@@user-lv6rn9cf8mYou're describing commissioning. I've never thought of a commissioner as 'an artist using a tool'
@@ChezzyKnyttNo, I'm not.
As an old art student when I now build models or I am creating a diorama, there’s nothing like reference, I use my phone or Pad to always check for tones & detail. So you using your father as a reference to detail, brings me back to my old art teachers/professors. Great video.
A month or two ago I saw a video where they did several versions of a single scene and in each version all they changed was the score. It was amazing how much you could change the mood and even the meaning by just changing the music. I was honestly very surprised when I saw it because I never would have thought it would be such a dramatic difference.
The problem with AI is highlighted perfectly by that director describing how much it helps them iterate. Mustache twirling capitalists aren't the only ones driving automation; it's legions of people who aren't looking beyond their immediate wants. The implications for creative fields include the decimation of mentorship, collaboration and diversity of ideas. Output becomes increasingly insular and derivative and opportunities scarce. It's long been a problem in some industries but AI is pushing it into overdrive.
Opportunities for what? To make the next blockbuster hit? AI has come to destroy the blockbuster, destroy the movie star because we will all be the latter starring in the former.
highly disagreeing with this one, actually. for every director who can already hire artists to make their images, there are a hundred others who want to work like that but can only afford to make machinima videos on youtube. ai, and especially freelance artists working with ai (for which there is already a vibrant market) will enable their voices to shine as well. it is the ai abstentionist side that's promoting a lack of diversity of ideas, by protecting an in-group's relevance at the cost of stigmatizing the out-group. and sure, purely ai-generated content _could_ be less diverse, if we allowed corporations like openai and midjourney keep their ai models under wraps, but we humans are crafty critters and we will use anything for creativity, so when ai models get open sourced there's always a huge uptick in community activities in trying to find creative ways using them. if you think the only thing there is to ai is writing prompts into chatgpt, look up controlnet. it's mindblowing how well an ai can work under your hands these days to help you express your ideas, as opposed to just generate something vaguely similar. it's only the megacorps that want to lock out human creativity because they're afraid that we'd generate something that doesn't fit their "community guidelines"
talking about the threat to the "culture industry" is ? because it only further highlights the labor problems that this new technology is throwing into sharp relief, but so many blame the technology, and not the problem
@@gnarledfork fair, but it's not just that. look at things with explicitly zero money changing hands, like fanfiction writers or fan artists, and you'll see _even more_ vicious hatred for generative ai. there's definitely an undercurrent of people who want to keep enjoying their elevated status in communities because they're the only ones who can create, even if it wasn't relevant to their jobs. even if we lived in a star trek universe people would be fighting this and stigmatizing those who recruit the computer's help to express themselves
@@dustinmcgregor9110 I like movies because of the human aspect, to see a vision made by a director i like, performed by actors i admire that I can talk to my friends about. If everyone's making movies using ai then that ruins the whole point of them.
Looking at the painting through a mirror/camera... THESE kind of tips are priceless!
You might also be interested in a tool I use when doing graphic design work with physical media: a reducing lens. It’s a concave lens that makes things look further away. Aside from giving you a long view, taking in the piece in a different visual context reduces the sort of perceptual ghosting we get from looking at something too long.
Looking into a mirror while painting is more about double checking your proportions. Digital artists flip their paintings in Photoshop too
Old techiques that he even said he doesn't use.
"To a creative person, regardless of the medium, creating art is as much a part of processing the experiences we go through in life as dreaming at night is. -Paraphrased from a video I saw recently
Hey one thing i wanted to say as an artist is that your dad was looking in the mirror to get a new perspective of the painting. Its a common thing we do because your eyes get used to a piece after a while and you can miss things you otherwise wouldnt. Its getting a new set of eyes on it so to speak lol. Its like when people see their own image not reflected they think they look strange because theyre so used to seeing themselves in reverse.
Very true. XD And you look so awesome by the way! The hair color is amazing.
I appreciate Adam living up to his shirt when talking about A.I. being chosen over humans in the creative process and taking a deep breath - twice, I might add - before continuing on in measured tones. A good example of how to not be disagreeable! I also appreciate his statement that the creative process is an alchemy and that there is no formula - so many authors, for example, have used a lot of resources writing books about what helps THEM, under the premise that THEY have stumbled across the "magic formula". I have personally found none of these approaches to be completely applicable to me in my writing, and few, if any, of them have held, in part, bits and pieces of my own, personal method, which I am, to be honest, still in the process of discovering! We may all, as creators, be one day remembered - like Mr. Kubrick - as being some kind of genius in our preferred medium (or media), but for us, we have to find what works FOR OURSELVES, and it's truly a PROCESS in and of itself to simply discover OUR OWN creative processes! I appreciate Adam for being an example for all of us in this respect, pointing us all on the path of discovering our own creative processes by way of these livestream Q&A sessions. Truly inspiring, and truly inspired!
I don't remember where I heard it now, but I remember a piece of advice on creativity, which is that an artist or writer's main responsibility is to stay alert and to wait patiently. When inspiration comes along, you don't need guide books and formulas, you just instinctively go to work. That's when you get that uncanny feeling of a story, poem or song "writing itself" with the words flowing faster than you can write it down. I've never painted a picture or made a film before, but I imagine it's much the same experience.
I am an illustrator/concept artist for large feature films (the batman, guardians of the galaxy 3, etc) and one of the people at risk of being displaced by AI generated images right now. In fact, my union (IATSE, specifically local 800) is negotiating with the producers right now about AI and a bunch of other things. My concerns with AI are many, and I love the way Adam put several of them. 1) you dont know who you are plagiarizing from. companies have been putting in barriers to generating images of mickey mouse for instance, but they are artificial and can be bypassed. mickey mouse was used to train the ai, so it will always be there in some fashion. also, it is trained on data stolen from the people the companies want to replace without pay or consent. i prefer to have at least some control over my sources both for nuance and i dont want to use something that is stealing food off my friends (or my) table. the issue of copyright is also a massive anvil hanging over our heads right now. depending on how things go legally there could be massive lawsuits 2) the internet’s (and whatever company’s ai you are using’s) internalized baises are now yours. AI generators are trained on the internet, including all its racism, sexism and other horrifying things. a while back some journalists asked midjourney to create large numbers of images of doctors (no minorities or women), prisoners (all minorities and men). i did the same for my industry, every director/production designer/DP it generated was an old white man. every actor was white, every costume designer was a white woman. this is neither reality nor the world i want to live in As we have recently seen with google’s ai, companies attempts to correct this dont remove the problem, only shift it (they had a problem where it wouldnt generate white people in images) they are trying to force diversity by changing the prompts under the hood, not my looking at the actual problems and it is not going well so when using them you not only have to be aware of fighting what you unconsciously bring to the table, but the amplified internet unconscious as well. 3) the companies producing these ai want to replace people/force the remaining people to work for as little money as possible. this is an inherit problem with capitalism. in the film industry we already have a huge problem with underpaying and overworking vfx artists, PAs and a lot of other positions. this technology will only make that worse as it encourages a race to the bottom. to adams example, why pay an artist when a PA can do “the same thing” 4) ai images have a specific look. and the more some directors are trained on that look, the more they want it. I have friends working who have had to make their digitally painted, non-ai images look more “AI like” as a result.
It does feel like an inflection point. How can it not be? And the issues around eg copyright, stereotypes etc will take decades to solve. But. It's here: 'The creative arts' survived and prospered after photography let anyone create an image, sound recording made live music redundant, as earlier, literacy and print displaced so much of rhetoric and story telling. Computing now allows anyone to write and edit, the internet is their global publisher. And in the end, big cultural changes came from small groups of people, or even individuals, not through a priesthood of experts, screenwriters, designers etc. Of course I agree with your concerns, and worry about how big capital might take over (they'll always get their percentage) but I remain hopeful.
I am not remotely an artist or anything similar - but I certainly agree with your points. Especially point 4 - all the art has the same 'feel' or blandness (maybe it's the lighting or inking type stuff) but it just feels something is missing.
Thank you for sharing your points. My favorite irony of the AI "art" development has been the difficulty to generate the human hand. The very human tool for creating art is incomprehensible by billion dollar AI programs because of its complexity and dexterity. 206 bones in the human body and 54 are in the hands. Is there anything the average artist can do to support the union artists who are working toward resolving the issue of AI in media or some legal action to support in general?
@@ericpitts4979 People feared synthesizers and General Midi to steal jobs of "real" musicians. And for synthesizer engineers it took decades to make a realistic piano sound, while instruments with less internal polyphony (resonant strings etc.) were much easier to reproduce. But did it matter? Synths became used to create many different things far beyond imitating the traditional instruments. Nowadays also proper sounding digital pianos exist (may they be physical modelling or made of many GB samples), and there are still pianists and acoustic pianos despite midi.
This is also a rising issue in the medical field, there are EMR systems that are recording Dr visits to learn the flow and chart for them. This will eventually replace scribes and assistants. Also there are systems that are slowly replacing receptionists to make appointments and handle phone calls. The medical establishments state this is for patient convenience and increased productivity. Ever try calling your medical insurance company for anything? If you think it's a nightmare now just wait till the AI takes over and thinks in knows whats best for you. (Benefits the company) or your doctor's office. Soon it will be one doctor one nurse and the AI receptionist to handle a clinic load of patients. AI doctors...its coming sooner that you think..
AI generated imagery is not art, it’s decoration.
The lightweight leather used for the Raider's jackets makes perfect sense if you think about it. Filming used to be done using very hot lights and you don't want your actors visibly sweating when the scene doesn't call for it. And if I recall correctly the jacket Adam examined at the prop store was worn by a stuntman, who would definitely want maximum flexibility and range of motion. Indy the character wore a leather jacket for its toughness and durability, but for the actor or stuntman portraying Indy it only has to look like a tough leather jacket for the duration of the scene.
Well they were also filming in very hot climates. But yes, the jackets from the film were made out of a very thin lambskin... and not terribly durable. A real life jacket like that from the 1930s would have been made out of horse or goat, or possibly sometimes steer, and would have been unbearable to wear above like 60F.
The thing about AI-generated content though, is that it relies on what humans have created before it. When humans no longer create, AI will no longer be able to.
That is what I have been saying. The whole idea seems very short sided. Until it can learn to create on its own, I think AI is limited.
@@Drstrange3000 except that is not how it works. New models will be able to do internal iterations, that is what was presented at Nvidia announcement. People don't quite understand it and think it reuses data, it is more akin to learning vocabulary and using that to make new sentences. Saying that it steals art is like saying anyone who speaks English is stealing words from the "original English speakers" or classic painters are stealing colors from the pigment makers. You can come up with a complete original idea and it can visualize it, it wouldnt be able to do that if it only re-uses other peoples work.
@@ilhan1936 That's not right either. True AI that learns like we do does not exist yet and is unlikely to exist for a long time yet. Self iteration is nothing new either, just look up any KZhead video about AI teaching themselves how to walk or play video games. There's also already been issues with degradation of quality in AI like ChatGPT due to the amount of AI garbage content on the internet. They're just regurgitating stuff based on the statistics of their data sets weighted by the prompts you use. Using the language example, an LLM doesn't understand vocabulary, it knows that you put the word "is" after the word "it" because that's how the data it was fed used those words. And it knows both of those "words" because the letter I is followed by the letter S or T a certain percentage of the time in its data (or its data set specifies a list of pre-existing words to pull from). It doesn't understand the concept of language or words and vocabulary, just an input (your prompt) and an output (the sentence), like any other program. If you used copyrighted/trademarked phrases to train a model (like McDonald's "I'm Lovin' It!" or if Donald Trump had successfully copyrighted the phrase "You're fired", or you used company names, etc.), and then sold it for people to use, you would be in legal trouble for using copyrighted/trademarked content for profit without permission. Just like how you can't start a new restaurant and name it Burger King. An art generator doesn't understand two point perspective, or anatomy, or lighting and color palettes. It just knows that images in its data set put these colors next to each other in 65% of its data related to your prompt phrases, so it puts them together. It's basically a more advanced version of using the Gaussian Blur tool on an image in Photoshop. LLMs are inherently unable to understand the concepts behind the data it "learns" from, which is the difference between them generating something and a person creating something. It's closer to laying photos on top of each other to see where colors overlap, and then repeating that 100,000 times until you can use the power of statistics to make what is shaped like a person based on probabilities than it is to understanding how to paint a person. Doctors use AI to check MRIs for brain damage and diseases, but they can't hand the MRI to the AI and go "tell me if there's anything wrong and what it is." The AI can only compared the new image to the images in its data set and go "this area here doesn't match the averages of what my data codes as normal." If you gave it a picture of somebody's butt, it wouldn't be able to tell you that it isn't a brain, because it has no concept of "brain" and "not brain". Generating an image isn't breaching copyright, but the company that made the generator did when they sold the program to you. Because they used the art that makes up its data set for profit without permission, and art is legally copyrighted upon the date of its creation. To the point where if you create a piece of digital art, print it out and mail it to yourself in a sealed envelope, and somebody else uses your art in a way that violates copyright laws later, that printed piece is legal evidence that you are the copyright holder for the art because it's got a federally recognized date on the envelope that pre-dates the other person using your art.
@@Khotetsu So open-source models that are avalible for free and are intended for private, non-commercial use are okay? I'm just making anime characters and Pokemon hybrids for giggles.
@@Eisenbison IMO, yes. I know there are artists who will disagree, and I understand why (because unless the devs asked the artists or the artists explicitly say somewhere that their art is free to use for LLM, they're still using their art without permission), but at that point I would compare using the images in a data set for an LLM to reposting art. I personally think a set of FOSS standards should be implemented for LLM that requires people to post at least a list crediting the artists whose art they used for the data set, in the same way that you should post the source when you repost somebody's art, but using one for your own private purposes like that is no different to me than using a meme generator or something.
The distance from the art sounds like the distance from a music track. The notion of listening to a song from a different room, or playing a newly recorded track on a transistor radio.
this is a time honored tradition in the recording studio.
I totally agree about AI "art." Ultimately it's an artist-who-appreciates-art perspective, and I'm finding a lot of people just don't really appreciate art much, but it just makes me more grateful to be a person who knows the value of art and human creation and getting to see that individual point-of-view.
You are so right about admiring the mind and heart behind the thing. Even for video games made in the 90s I wish wish wish wish that we had documentaries on the group of people who created sonic the hedgehog or Mario 64. These people seem so super human to me.
here's hoping that maybe, just maybe, enough people will still care about these things in the future for organic art to thrive
@@rdzdoodles8592 most AI Companies are pushing to Make this the New Norm, its invaded art schools if what i heard is true. I still see the fight still going on Artist vs AI Bros but i unfortunately see in the next 10 years (Considering how Fast AI Is Developing at an alarming rate) AI is sadly going to Destroy Organic Art entirely, all that will be left is basically Typing words for a prompt. And its going to be a bleak reality for normal artist
As Alpay Ephay says here on KZhead "AI creates images, humans create art."
The thing is, that's fine. We're surrounded by images - eye-catching bits of print whose cultural value is virtually nil and whose only real purpose is the visual equivalent of a klaxon going off to get your attention. In a vacuum, I don't see any harm in this being automated. What pisses me off, though, is that right now, making images is someone's job. It's how they make money and care for their family. Taking that away because it's better for the company's bottom line is trash.
that's a very delusional mindset. Industrial Art is already extremely automated. The real saying is, "if it's an industry, it not art". There's no difference between a human churning out a media offering and an LLM. Humans do the same thing an LLM does. Samples what's available, and creates a new combination of it. So all these "artists" working in the Media Industry, were never really artists. And neither is an LLM.
Humans create AI
@@Kleion_RFB Also to note, "AI" doesn't create images, it amalgamates, and procedurally mixes images using a pre-defined averaging algorithm and a user defined search tool. All images used by AI are just other peoples creations used without permission and in a way that hurts so many artists that no one can claim their piece for copyright (Except that one artist whos signature was so consistent is showed up on AI art as an essential element on a commercial product). And because of lazy and cheap programming these "AI" eventually average themselves into oblivion, without constant new, well labelled sources they all fall apart quickly.
@@ZennExile Truly spoken like someone who has never experienced the creative process.
love the question about ai! keep this thread going in future videos, your feelings towards ai make a lot of sense
I think you and the other commenters here have already covered all the bases on AI discussion, but I just wanted to put this here. I graduated DigiPen, one of the foremost game schools, in 2022 with a Bachelors of Fine Art, in Digital Art and Animation. About a month after graduation, the mass industry layoff waves started, along with AI. Most previous graduating years at my school had about a 90% hiring rate. My year is about 15-20%. With all this going on, I've applied for hundreds of industry jobs with no success, and now I am stuck working part time at a local big box store, and it nearly brings me to tears. I can only hope that AI doesn't replace artists in the industry, and that game studios realize that what they're doing is harming artists and other passionate employees who love working on their games, movies, etc.
I'm aware of a number of game studios that fired their whole art team, to replace with AI, and... 6 months later they've tried to re-hire artists because the output of the AI wasn't good enough quality to use. It may take some time before more people come around to this point of view, and the big soulless companies that just want to make assets as cheaply as possible might never come around to it, but there's still some hope for human artists I think. (and also, the managers that think an AI can do the job of a human probably would suck to work for anyway, as they have no concept of the value that a trained artist can bring to the work.)
I prefer the age of Mastertronic, where a game for C64 could be coded in 3 months by a single human. Big business game industry lost its soul anyway. If people use AI to not need 1000 people and 100 mio US$ to code a game, it can become only better.
The layoffs aren't about AI, to clarify. They're because of financial blowback after all the major acquisitions in the past few years.
@@revlarmilion9574 Mostly that, yeah, but the temptation of AI has definitely added to it, or at least emboldened them
Many years ago, when I was at Universal Studios in Florida, I saw the trimaran used in Water World rusting on a back water lot. It was much smaller than it looked in the movie.
i mean it was 60 ft if it was the screen used one with a 100 ft mast and 60ft wide and it would not rust as it was fiberglass but someone restored back to working order in the mid 20's i think.
Here in the UK, the Intellectual Property Office (the UK's patent office, responsible for intellectual property rights in the UK) has halted their efforts to create an AI Code of Conduct, due to disagreements among stakeholders. Those stakeholders include tech companies that have developed and host these generative AI platforms. They won’t accept that their use of images scrapped from the internet without permission (to train their AI) is a copyright infringement. Since when did we start asking permission of those who perpetrate organised crime? The IPO is impotent.
It's not copyright infringement if its transformative enough that you can't recognize the original from the generated result. Same reason why parodies aren't considered plagerism.
@@Eisenbison It's not a question of direct copying, but a question of how the works were used. Copyright law is more than just "does this look too much like that?"
Both legally and morally, what you said doesn't make any sense. See @BubbleBoar reply. But also parodies are parodies precisely because you can clearly see that they're recognisable from the original form.@@Eisenbison
I am on the same page with you about the commerce of AI, and I believe there is a very real problem of ethics in training a machine learning algorithm on other's work, but I think the problem with AI aesthetically is that presently many people are looking at it as if it were a solution, when reality it's a tool--like Photoshop, or drawing over a 3D model. The difference is that creating an image that "looks good" isn't the hard part, it's adding the point of view. It's doing the human part of the work. The reason it's a tool is because it's a new medium, with its own rules and values. You don't see sculptors scoffing at cartoonists that their work sucks because they're using significantly less work to depict the human form with little effort. One day, when we've resolved the economic and ethical effects, I think the medium will see some serious growth artistically.
Well it's been seeing growth since day 1, artistically speaking. Every new model people train adds another near infinite new possibilities someone can use to express themselves and every new extension or algorithm is another arm to control the models with. I don't think economic and ethical issues will be sorted out, this is just another one of those periods in history where some workers are made redundant to the economy. It might even be a lot or most of them but time will tell.
I'm so happy you make content. Thank you for all that you do, I have so many wonderful memories from watching you in MythBusters. (:
In a great film the champion(s) are great at communicating the vision to the others. I've seen so many movies where every aspect is great, but they are each pulling in a slightly different direction. With a great film, all of these great parts come together in a way that reinforces each other. It's that focus that pushes it over the edge. It's also fascinating how this happens. Some directors are ... intense ... about making sure that everything was detail perfect. Others get everybody worked up and inspired, then sit back and let the kids play, picking and expanding on the happy accidents that hit the right notes. And, as William Goldman said, "Nobody knows anything" about making movies.
10:39 "[kubrick] built a machine finding [drama]" aka we'll find it in the edit*. *as long as we have tons of varied takes and not just 1 or 2 and we say that to skip doing the work here and now.
As someone who’s worked in radio news for 30 years, I very much appreciate what you said about AI in that space. I’ll guarantee you if management can figure out a way to save a couple of dimes using AI, they will.
"All this machinery making modern music Can still be open-hearted Not so coldly charted It's really just a question of your honesty Yeah, your honesty" - Rush, Spirit of Radio
If it's generated by AI, it's Not Art
1) A painting close and far away reads totally differently. At least color wise. You need to pre decide what distance your work will be seen from. beforehand.
You have such a refreshing and logically sound perspective on things in life. I LOVE YOU ADAM!
Could not agree more about using the camera ( or distance) when assessing progress on certain art projects. I paint on canvas sometimes, and build scale models/dioramas . Weathering scale vehicles and using a camera as You described has been part of my process for last 4-5 years now. Can’t imagine doing my builds without that step now.
My issue with A.I. is that people will and already are using it to cut out actual artists to cut costs and save money. People will use it for good but sadly the majority are going to use it for negative reasons that will harm creators. Even more unfortunate is that it's gotta be one way or the other and A.I. needs to be restricted before it causes more harm across the board as any middle ground offered will just be abused in some manner. In a perfect world it'd be used along side actual creators to create, but in our world its mostly going to be used to cause harm or give big companies an excuse not to pay workers/artists/creators.
You know the funny thing that people seem to miss: the graphic artists are not the creators. They bring to life the creative vision of the person hiring them. Just as software developers write the code that brings to life a piece of software envisioned by someone else. The true creators are the ones hiring these people, and making their visions come to reality quicker and cheaper is a good thing, just as it was a good thing when the loom replaced the seamstress.
@@allanshpeley4284Your analogy falls flat - looms are for weaving, and sewing uses machines that still require human operators
@@allanshpeley4284 I work as a creative director, so I am one of those people who hire others to execute a vision. And what you're saying is explicitly wrong. When I hire a team of artists, a large part of the discussion is about their areas of creative freedom, their perspective on the vision, and their unique process that inevitably translates the vision into something beyond what any of us could have imagined on our own.
@@Jack-la I agree with you. I am working as graphic designer/illustrator. Here is one of the task that i was challenged with. I needed to draw blockchain interoperability and not like this boring web lines with dots or cubes. Client wanted something new and interesting. I spend a lot of time i tried AI and the results was terrible mess. On top of it I needed to make animation in vector of this illustration. Any one can try to promt blockchain interoperability illustration and will see that AI is not so good. And if the client will tell you oh i like this but can you change this and that what you will do? Dont forget that all that should be in vector in layers and ready for animation. It is just one example of my work.
@@allanshpeley4284 utterly delusional trolling , completely disconnected from reality or actual experience.
Every time the AI conversation comes up, I can't help but think about the John Scalzi short story, When the Yogurt Took Over.
I always think of a 2000ad story from 1986 "the art of Kenny Who?" its a great story and predicted the future perfectly from almost 40 years ago.
I remember seeing close-up photos of original Star Trek props, with glue seepage visible, and crooked knobs and buttons, and scuffs galore, on phasers and tricorders.
Hi Adam, I have a question about the Lethal Lawnmower / Glass Guillotine episode. During your monologue describing how the glass guide rails work there was a crash of metal in the background. You reacted, stopped, and kept with the take; that scene even made it to the final edit. Did you ever find out what happened in the background? I'm oddly curious. Noah
Love the shirt, Adam. 💜
I see people saying how AI will replace this or that profession. Something like "in two years we'll just press a button and AI will generate a custom movie tailored for our preferences", and I see few people mention how uninteresting that sounds. There will at least be a period of time when AI makes creating things more efficient, removing all the boring parts of editing and post-processing.
This is where AI actually excels, refining the process of creating something to take out the frustrating bits. Not skipping the process entirely. That defeats the whole purpose of art, because creating something is as much about processing it in your heart the way we process stuff through our dreams at night as it is the final piece.
Adam you've always been one of my favorites. I'm so glad you feel the same way about this bullcrap.
I really liked this take on AI, and would be really interested to hear at some point more long-form discussion of what you mean by an aesthetic or creative point of view. I think it's a great way to put it.
YES, I feel so relieved when I see artists fighting it, I always get attacked online when I sound my opinion. I WANT ai in science, not art!
I hear you, but there is no money in science. I'm always annoyed how it uses super computers to generate a google search for lazy people. It's an incredible waste of resources when it could be folding proteins and learning to detect cancer. But nobody really cares about that, they only care about themselves and the now. Let it develop through popular means until it can solve some real issues.
Yep..Ai doesn't belong in Art but it has its place in Science.
I hear you, but there is no money in science. I'm always annoyed how it uses super computers to generate a google search for lazy people. It's an incredible waste of resources when it could be folding proteins and learning to detect disease. But nobody really cares about that, they only care about themselves and the now. Let it develop through popular means until it can solve some real issues.
I'm always annoyed how it uses super computers to generate a simple google search. It's an incredible waste of resources when it could be folding proteins and learning to detect disease. But nobody really cares about that, they only care about themselves and the now. Let it develop through popular means until it can solve some real issues, the amount of money found in science is never going to drive development as quickly as funny images and convenience.
Test comment. Being blocked for some reason. Edit: I'm always annoyed how it uses super computers to generate a simple google search. It's an incredible waste of resources when it could be folding proteins and learning to detect disease. But nobody really cares about that, they only care about themselves and the now. Let it develop through popular means until it can solve some real issues, the amount of money found in science is never going to drive development as quickly as funny images and convenience.
There is one thing about AI that most people seem to forget: Their pool of training data. If the things that were made with AI feed into their training data, it will generate a very bad feedback loop. We can already see it in the text AIs. A lot of their training data is already written by text AIs and it leeds to problems that need to be filtered manually. I recently saw a music video with ethically done AI Art. An artist trained an AI exclusively with their own work to make the music video from still images. All the rights stay with the artist, it's obviously their own style, there is no outside influence beyond the influence the artist has and naturally the director for the music video with feedback from the band. That's how you do it.
AI eating its own data getting worse... That's like the old joke. A: "This sausage tastes bad. Feed it to your dog." B: "I can't. He's already in it."
What's the point of being an artist if you want a computer to make it for you? I really hope this is just a fad. Do people prioritize instant gratification to the extent of handing over their artistic expression to a machine??? It's so bleak. Art isn't a chore that you buy a robot slave to do it for you. The effort you put into your art is the ENTIRE POINT I swear this is actually giving me an existential crisis.
@@elijahbarnes6145 Yes, they do prioritize Instant gratification that much and its exactly as disturbing as it sounds like it is. And its going to get significantly worse before it gets any better, if it gets better.
How kool! Do you have a link to the video?
@@elijahbarnes6145I think the difference here, from what I gather, is that the original artist was an illustrator of still images, not an animator. So they used their own still images to generate an animated work. Animation student here, animation is hard 😅
Exactly Adam. Art, so far, has been in each case the incredible insight of one person, into some aspect of life or imagination. This is why it can grab you by your soul. A.I. art is a mish-mash, an agglomeration, of many different visions, not one of which come through clearly enough to astonish or affect the beholder. Great art is never by committee.
I teally appreciate you soeaking on the behalf of labor. You have an interesting perspective that is incredibly indearing to hear from
Art is as much about the process, as it is about the viewing/listening, you can't have a machine "enjoy" painting with water colors, or blasting out power chords through a wall of amplifiers, the machine doesn't feel, smell, taste, nor does its thumbs get sore, ears ring, or fingers get blistered, which lead it down some other path where true inspiration lies
This!! You are one of the few people I have seen talk about the actual process of creating something. You discover things, make choices, changes, during the process! That's why having a process is so important. Art is not just have idea, input idea, output image/music/whatever. It doesn't work that way and you can tell these AI bros do not have a creative bone in their body because they just do not see the creative process beyond some sort of logical equation.
I enjoy the process using AI. I hand draw some ideas, feed it styles I know about thru AI. Draw again, tweak fix, edit in PS. Makes iteration super flexible and fun
I appreciate the way Adam split the AI question into two parts, the visual and the employment. Because this reminds me so much of the invention of the fabric loom (the Luddites) and the sewing machine - there's no question that IT'S FASTER. But that in and of itself doesn't immediately translate to "therefore thousands of people are going to lose their jobs." THAT decision is made by the higher-ups, who have no incentive, in this economy, to prioritize anything but profits. (Go vote) An individual with a vision - like in Adam's relayed example, an illustrator using AI to iterate on ideas hundreds of times per day - can USE AI to get further and faster. But just like the "graphic design is my passion" memes, if you aren't going IN THE RIGHT DIRECTION, you'll get farther, and further, AWAY FROM YOUR GOAL. And eventually, like a calendar that only has 360 days per year, the widening gap will become more and more obviously in conflict with reality and a shift MUST occur. With no way to STOP people from using AI for writing student papers or other "cheating" scenarios (aside from draconian surveillance and ableist limitations), acceptance and adjustment is the only possible outcome. Hats off to the unions who will have to work extra hard to MAKE that "adjustment" process work for workers over profits...
The calendar metaphor is genius. I think I might use it.
Nah in this case, the machine literally only exists because it stole the work of hundreds of millions of people even using their names as style guides in the machine and the owners themselves are on discord evidenced as having said that "you can just launder the data through a fine tuned codex" Launder is illegal. So they know what they are doing is illegal or SHOULD be illegal. Then they turn around and straight up said "Conveniently forget the training data used" Covering up evidence. "And that's all the legal issues solved forever." They are scumbags that know exactly what they did. Not only that but this affects writing, music, art, animation, games, propaganda, bullying, gaslighting, scamming, political framing, misinformation. This is not just, "here's a tool to help speed up some tedium." Nor is it "demand outreaches the supply, we need more efficiency" Art is not NEEDED in the same way that certain textiles or cars might be required and therefore automated. There is ZERO reason to automate the creative process beyond making more money as fast as possible. This is a technology which should not even exist and WOULD not exist without stealing data, which includes child porn in some cases and private medical data they should NOT have and then going, whoopsy, we made a fuckie wuckie. Not our fault we gobbled up EVERYTHING on the internet. This is unlike any other revolution in history because this is essentially the equivalent of giving every person the means to just create a ferrari in their home, thousands of them a day. It would make car manufacturers entirely obsolete. Devalue every car down to nothing and make it so that no one gives a shit about anything related to cars anymore except, in this case, it's far worse. It's devaluing art and human expression itself. Trying to optimise the human condition. Musk only just came out saying shit about "many books are way too long. AI could be used to shorten them and focus on the salient points." Again, wanting to literally optimise human creativity, spirit and soul. It's sick and dystopian. At this rate we're headed to Cyberpunk dystopia. Not the techno utopia all these pro AI meatheads lie about.
Awesome talk, Adam. Thank you!
I've never thought about it before, but I'd expect all coats and things to be really light, because you don't want to actually be wearing warm clothing under studio lights and stuff
My problem with it is, in its current iteration, you can't control it anywhere near as much as traditional media. So while you can make something that's seemingly decent to external viewers, it's not yet a tool that can make your vision a reality. As a maker, that's made it unfulfilling. And like 3d modeling, I miss the tactile experience of physical media when I use it.
I realize what you're saying...but it's current iteration is what, three or four weeks old? With say Chat GPT breaking out maybe a year ago? This stuff gets better faster than is comprehensible. How long ago were Your friends saying AI could never win an art contest?
This is why I see it used so much for brainstorming and concept art. Once they get something close, then it becomes the starting point and it gets refined from there
@@handquakeWait, how do you even "win" something as subjective as an art competition? Sounds like a publicity stunt for technology companies. And since AI is incapable of original creation, it'd be the original human creator's content "winning" for the AI anyway. As for the future of AI, I'll believe it when I hear it on the radio in my flying car!!!
@@TheJofurr Am with You Car Fly Guy. Stoopid "Art Competition". But a human only creates art based on everything it's experienced in it's life...and an AI only creates art based on everything it's experienced in it's "life". The only original creation a human would create would be one where it never interacted with the world at all...one where it has never experienced anything but outputs beauty....but that's never happened far as I can tell. If it has seen the world...if it has seen many things...then it outputs art...not if it has seen nothing. AI works the same way. It sees, it learns.
@@handquakeAbstract art isn't experiential.
My experience with Generative AI Art is that it's more quantity over quality. It's great for getting the big picture started but when you try to go deeper and more specific, that is where it starts to fumble. I still would rather have an actual human artist with a flexible creative output however, if I'm working on a personal project and have no budget for a pro artist, AI does help a bit.
Really enjoyed your point of view on A.I.-Generated art. Being a creative person since 5 years old I don't see a future where our unique gift(s) and creative minds will and neither should be replaced by A.I. I am not against using any A.I.-Tools when it serves my creativity and helps with the process. It does feel "cheap" that creative companies are looking at ways A.I. can help them decrease the size of their team(s) but this could also be positive for creative companies who struggle with revenue. If you've ever worked at a studio or creative environment you'll know that A.I. can't replace the amazing people behind the scenes. The amount of skill, iteration, fun, passion, tears, etc. that goes into projects every day - that's where the clay is formed into something. Be creative, practice your crafts with passion and integrity and with or without the assistance of A.I. continue to create and learn from every opportunity. 🖌
Came for a statement on A.I, stayed for the great insights into filmmaking from Adams point of view
These computers were taught with stolen art. My friends are feeling the affects of A.I. luckily it’s really pure garbage but corporate companies are just looking at the bottom line, not creating art. I unfortunately see a future of “Good enough “ movies and art coming. So sad.
This 1000%
These big bags, these wolves in sheeps clothing, the companies who wish to devour and take. They’re not looking at the substance, not the Creative Integrity of making, the soul behind the creative mind. They only wish to take… But creatives and makers, we dream of giving. Not taking.
But the double sword is creative people wh0 don’t have Hollywood money will be bending AI creatively to create at a high level and have a voice without the restraints of money. Without having to think of money folks get more freedom to explore ideas. Any niche can be expanded on.
@@themightyflog Money is not what stops people from drawing and writing. That is bullshit
stops them from making movies with special effects, lights and actors. We were talking in the context of movies.
I think it’s also very important to emphasize that a lot of AI is trained on material from artists who did not consent to their work being used. Some even outright saying “I am not ok with this” and then their art was still used.
indeed, many of my paintings are in the LAION5B database and I have no wayof removing them, even if I could the training has already been done. There is also CSAM material in the training databases.
Is this not the same as an artist going to see a gallery and looking at others works and learning from them.
@@xxsemb it is exactly NOT that, because artists consent to their art being in galleries You care about consent, right?
@@ClintonAllenAnderson Nonsense. When an artist creates a painting they release it to all of humanity to look at, enjoy and draw inspiration from. If they only released it to a small group of private people then these models would never have the ability to use for training. And let's not forget nothing is being copied by these models. The images they generate are entirely unique, but inspired by what they've been trained on. Just like every human artist is inspired by what they've seen and experienced.
@@allanshpeley4284 So I guess you've never heard of copyright..... You aughta look it up somewhere....
Sweet Moonwatch, when did you get it and how do you like it? Did it live up to your ideal of historic aerospace gear?
You've got to do a video that explores prop replication and weathering further. There are sooo many things to consider in my experience as a relatively new modder!
AI dont have a "point of view" because there is no human behind it, but not only that.. as a receiver/enjoyer of the art we add a point of view because of the fact there is a human behind it, an example would be a specific song that means alot to you, the lyrics can often be vague about its meaning so you interpret your own into it. I have heard artist say they refuse to explain what their song is about because they dont want to take away from whatever their fans feel its about on an individual level
Creative AI functions just like an imposter, because naturally it *can not* have own personal experience. So it is like a fake clever doctor who has thoroughly read 20 books about medicine and improvises the things not found in books. This may work 90% of the time but turns into imaginary nonsense answers when he does not know a detail, and he can not estimate when he is telling nonsense. AI is a tool, not an artist.
I remember reading that about Hotel California, that they either refused to give a proper explanation, or gave multiple conflicting ones. Bohemian Rhapsody, also. i know there are others.
The prompter is behind it. The last time I checked the prompter is human.
There is not only a human using the AI to create whatever, but every single bit of the training data has human intent or knowledge imbued into it. And AI can obviously create vague or even nonsensical lyrics that you can interpret how you want to.
@@DemWaifus But you wont interpretet "nonsensical ai lyrics" how you want to, and that is my point. Because why would you try to do that with something thats just an empty shell trying to seem human. And just because something is made out of human-made components doesnt mean its anything like a human. Not at all
"like everything, it's a little bit awesome and a little bit terrifying"
As soon as I clicked on this video, I was like “Adam is going to say it’s because it lacks point of view” 😊 spot on haha
On the point about a film champion, I notice this with open-source projects too. Whether it's a founder-leader or a maintainer, _somebody_ has to have an inkling of vision in order to shepherd the individual contributions into some cohesive form.
What’s a few extra fingers among friends?
Or teeth.
Regarding your safety tips on another video, I was waiting for you to say, at the end of the cut, hold your hand position until the blade stops turning. It seems like a lot of injuries happen after the cut.
That's definitely a thing. I have ADHD, and my mind is like a task list sometimes. If you think of the task as done when the cut is made, you start going on to the next step immediately. If the tool isn't part of the next step, it's irrelevant and you don't think about it. But if you think of the task as being done when the machine is reset to safe mode, then you stay in the flow state until that step is done and then reset your brain for the next thing.
Recently i had gotten into minature painting, and i decided to see what i can do by running the minature through blender to get a depth map and a normal map. I put in a but of work on the prompt and told it to generate me a full color scheme for the minature. Perfect details in how it did it. Then i pained it like that, its neat to see how it works, and i use it as a tool, im not sure where its going, buh i know companies are going all in. Ups is already working on fully automated locations for their handling.
"Art is the world seen through a personality." An algorithm is not a personality.
Would the personality not come from the person making the prompt? It's all about how you write the prompt, you can't just write a description of what you want with nothing else, and expect it to make exactly what you were envisioning EDIT: Remember that there is a lot of misinformation regarding how AI works, eh?
@@ironencepersonal9634 You can't convey personality throught text, just as you can't hear the tonality of a person throught the way he/she might write a text massage unless you hear them personally throught a phone call. Same goes with this, a person uses brush to make certain brush strokes and shapes that only is assosiated to that particular artist (personality) UPDATE: I'm absolutely stunned how many of you didn't quite grasped the meaning behind what i wrote. There is a massive difference between conveying personality throught poetry vs SMS massage! obv I meant the latter. cheers
@@ironencepersonal9634I don't think it would as the works the AI uses to make the images are all stolen or purchased from whoever originally made them and their perspective. The actual merging of those images to create a new one based on a prompt were never meant to embody that prompt in the first place from my perspective at least as an artist.
@@ironencepersonal9634 a prompt it just an idea stripped down to the most basic form. Art is the idea filtered and expanded on by the artist. one is reductive, the other is additive. Compare it to a film made by passionate filmmakers vs a soulless studio mandated cash grab that takes the original and sucks all the joy out of it. The competency with which both are made is irrelevant, you can tell when there is passion behind a project and when there is not.
@@ironencepersonal9634simply, no
the thing about AI is they still fails to draw connections between things that has not been connected before. This is the same for both scientific topics and art generation. It will try to draw every element i prompted but fails to blend them together in a artistic way.
It also has no concept of what anything *is* - if you tell it to draw a female superhero, she will always have the build of a supermodel, in a skintight catsuit, because that's 99.5% of the dataset. It can't "imagine" a female superhero with the build of a super heavyweight boxer, or emaciated, or fat, or with no legs. A human can imagine these things and create something different or new, because they know these things aren't part of the definition of "superhero" If you tell it to create a female superhero with a lasso, then you will find suddenly she has black hair, a bustier, and wristbands - because its data set that was "trained" on images with "female, superhero, and lasso" as tags only contains wonder woman. You can't get a female Zangief with an undercut and a spiked bolas.
The distance and mirror for the art is about seeing imperfections. Usually, bad proportion and shapes standout if you look from a distance or if you place the art upside down
Where do i get one of those sturdy phone camera goosenecks?
Artificial , yes. Intelligent, no. Its generative art theft full stop. The only people who benefit from this technology are those who wish to reduce their bottom line . There is this overwhelming desire to create faster and crank out as much “content” possible these days to make as much money possible. In my opinion this attitude is a cancer on the art of film making. Stepping off my soapbox now.
Most people don't even know this aspect of AI art. They just think it looks cool.
People act like this just happened
@@internetpointsbank Many yes. Ive seen this coming for sometime. Having an opinion of it doesnt automatically mean people are unaware. This has been a point of discussion for several years for some of us. Its only because its making huge leaps forward in recent months and is now in the world spotlight.
Ah, but cancer has unlimited growth, and isn't that what investors ultimately strive for?
Yes to all of this.
AI by it's nature generates an average result. It's never going to be great because it's not designed to produce great results.
Considering that is true, an I as well believe its true. I am also afraid that common consumer really doesn’t or will not even question how materials or images were made when pushed out in the public en masse, for the simple reason that not everybody is critical enough nor cares about it.
"Point of view" isn't the term I would use, but it is the point I have been making. We as artists, as makers, put ourselves into our art. We express ourselves, we show ourselves to the world. The fact that a person labored on it is the very thing that makes it fascinating to behold. No AI generated content can truly replicate that because it bypasses the very thing that brought us to the art in the first place. It can make an amazing facsimile, and it could even fool us. But fundamentally no one actually wants to see machine generated art, because it has lost its value because it has lost its effort.
It's true that you lose massive detail when you take a picture. Try standing in front of a Van Gogh 'Sunflowers' VS seeing one of them in a book. Utterly different experience. The first time I saw VG in person I was blown away. It's as much sculpture as it is painting.
Anthony Daniels didn't want to do Star Wars. 😮.It was Ralph Macquarie's artwork of two lonely robots in the 🏜️ desert that called to him,that made him take the work...A.I cant do that😂
AI could have done that if it existed back then.
I'm an engineer and find AI an interesting concept. It has a very specific good value. I believe AI is great for spamming high volume output. It is good on the very front end of projects. I don't care what industry we're talking. When you want to brainstorm ideas and throw a whole lot of things against the wall, AI can be good at spitting out a lot with very little prompting. You can cherry pick promising outputs, and then use those as templates for real work. I do not believe AI is good for real work. It simply doesn't understand anything. It does not comprehend what it output. The lack of understanding is very dangerous. Inexperienced humans have the same problem, honestly, but the human can learn understanding. AI can't. AI can only goal seek based on it's training, but it doesn't know why. It doesn't know if the output is good. It simply scored high in correlation to deem worthy of output while having no comprehension of that correlation.
Creative AI functions like an imposter, because naturally it can not have own personal experience. So it is like a fake clever doctor who has thoroughly read 20 books about medicine and improvises the things not found in books. This may work 90% of the time but turns into imaginary nonsense answers when he does not know a detail, and he can not estimate when he is telling nonsense.
Hey Adam! I was listening to the Team Deakins podcast, they had a costume person on, and she said for Pirates of the Caribbean she put the boots in a cement mixer with some bricks and dirt for a week or so to make them look like they have been worn for years. Also one of her pet peeves is seeing costumes not weathered enough or at all. It's Always more than you'd expect them to be.
You can still hire that illustrator to generate art and edit it to your liking. I've used Bing image generator a lot and it takes about 20-80 image to find one that's close enough to your idea that you can edit.
It feels like it has no soul.
Just like the Replicants of Blade Runner 🥲
don't be silly, it's still created by a human, until AI does stuff unasked, by itself, without a prompt, it is still human effort.
@@HarryNicNicholasWell said
@@HarryNicNicholas wow anyone can't enter a prompt though....
@@HarryNicNicholasThe thing that I would do and test it with I'm sure they are in the back room doing... Scares the willies out of me. 👀
Imagine all the extras in movies that don´t get the experience of being in the movie.
I think there's something we have to keep in mind with AI. Currently, it is new and it is evolving rapidly due to massive interest in this type of programming. However, given that there are innumerable amount of people who are against AI, I believe that many businesses and such may opt into having humans continue to work for them. There are many things a human can give that an AI cannot (at its current level), and I think in many industries, it would be better to have humans. Industries like customer support, restaurants, and more.
Ok. This is a great video but I have never thought about how by looking at a mirror you are doubling the distance... and it blew my mind for a hot second.
3:28 Radio DJ's are one of the thing that AI can replace seamlessly as even real DJ's mindlessly regurgitate catchphrases and announce upcoming news & weather all day.
.....not because they want to, though. Most DJs on commercial radio are so incredibly tightly controlled by ClearChannel that they almost can't have a point of view. A good DJ, with creative license to pick music (most can't) and to actually portray opinions (most can't), is irreplaceable.
@@meepmorprobotcaptain Indeed, similarly Artists doing advertising/concept art/background pieces must adapt to a world where they're simply out-priced and out-produced in the "forgettable" category. Upside is less bad work: Downside is less work to built portfolio, experience and name with.
i mean readio is dead lol
The AI discussion feels very similar to the CGI one. In the early days people said it was soulless, it could never replace practical effects and it would never look as good. Now look at it, CGI is so prevalent in all TV shows and movies that a lot of the time you have no idea it's even there. Early on it was very easy to spot, these days it's all but impossible for a lot of things. Give AI generated content 10 more years and you wont be able to tell is my prediction. Just look how insanely fast it has evolved in just the past couple of years, it's nuts.
CGI is still created by people though, it has a purpose and is made by humans. AI has none of that.
@@technosworld2 It has it to some degree in that you can/need to prompt it to go in the direction you want it to. Again, give it a few years and the ways in which we can influence how the images/videos are "drawn" will also evolve. Just as a crude example, today you can sketch out a rough scene on paper, you then tell the AI what you are trying to create and it will use the sketch as a guide and fill out the details.
I feed AI my sketches then use the results for inspiration. I love this process because it's an affordable to get closer to my vision in a back and forth way. Caveat: I do not sell any of the stuff I make to avoid stepping on toes
If I’m not wrong Practical Effects are still being used (I’m not sure if Practical effects only count to suits and props onset) and they coexist with CGI basically giving both parties something to work on in a project movie/or show. AI on the other hand seems to going to territories of replacing all of them instead of being a tool aid and coexist alongside both parties
I'm thinking about some of the first steps in computer generated images from other sources like when smoothing out merge cuts in video production. You give it the last frames of the previous cut and the first frames of the next cut and you specify how many intermediate additional frames you want between the cuts and whether you allow it to slightly modify the existing frames. The software calculates the new frames to make the scene look like there are no cuts at all. But for copyright, the owner of the modified or generated frames is the owner of the original frames. When considering AI generated art, the same concept should be applied. The owner(s) of the original material are the owners of the generated material. Though this is the most logical way to look at it, it's also impossible to enforce, because the way AI works makes it impossible to identify any specific source that was used and by how much it was used. If SCOTUS eventually says that generated art must be attributed to the original human authors of the specific images that the generative system used in constructing an image, then it would probably destroy the business models that currently exist for public AI generative art since the AI models cannot specifically identify which sources were used in any particular generated image, essentially making the AI illegal to use unless you train it only on public domain images or images you own or have the legal right to use.
Video to Video conversation is totally a thing already. I do it quittttttte often for clients wanting stylized videos of themselves. But beyond that you can mix a stylization model, style your model, then train your model on the stylization, which results in QUITE stable shots that are stylized to a specific person/character. I think as we move forwards its really going to just become more and more interesting personally.
The point of art is not for something to look pretty, it's about making it, spending time with your thoughts and interpreting your feelings and experiences through a medium. An algorithmic reproduction of a painting is meaningless, it's nothing more than a novelty.
if you have something that you find inspiring a wonderful piece of art ... would you change your mind about its inspirational or emotional value after finding out AI made it ?
Using AI to create art is a skill just like all other art. Its not going to spit out exactly what you need the first time. You have to be very descriptive and carefully craft your wording to get what you want.
@@QueenyCrowley yes, because the context of a piece of art is arguably the most important part of it, and I don't believe that generative art can actually capture someone's individual style and emotion, it can only emulate it. Again, the end product of a piece of art is almost incidental, and when the process of creating it is (nearly) eliminated, we are left only with the medium. Using generative art signals to the viewer that the idea was not worthy enough to invest in a physical medium, so any inspiration I may feel about it is gone, because I understand how the piece came to be, which in my opinion, is a rather lazy and indifferent one.
@@TomsBackyardWorkshop sure, this is true for other digital mediums, people felt this way with synthesizers and DAWs, but the difference is the amount of aggregate data that generative art is able to use, compared to layering single sounds at a time in a DAW and developing a style by carefully selecting what to include and exclude. I think generative art has potential be a great tool to help with ideation and references, but the dataset has already done the most important part of creating a work of art, which is interpreting the world visually, and therefore I do not care. It's like fast food. I do not think that what we need in our society right now is more quick, cheap and disposable ways to express ourselves or to consume. It simply doesn't mean anything to me that you had to change the syntax a little bit. Make it for real or don't bother
I disagree. The point of art is enjoyment. You can enjoy Ai art.