AI is still utter sh*** in 2024 (at concept art). Is there hope?
Vložit
- čas přidán 24. 07. 2024
- AI, typically in the form of Midjourney has long been the looming Darth Vader to the Luke Skywalker that is the working artist. It threatens to literally destroy our lives every single day and every single day it grows more and more powerful.
I myself have currently lost my career due to the games industry layoffs and risk having to send my wife back to her native Netherlands if I can't earn enough to get her past the visa income requirements in a year to two year's time. And if in that time jobs came back but were replaced by Prompt Bros, who merely laugh at the idea of mass unemployment for artists, well... depression and despair would be hard to fight off.
IN THIS VIDEO: we'll talk about how sh*t Midjourney is, what it CAN do well and what it CAN'T do well and I'll be referencing a real world example of not one but four prompters failing at a real world concept art job. Also I'll be refercing Feng Zhu and his recent thoughts on AI and how sh*t it is at concept art.
=========================================
Follow me on...
Artstation: www.artstation.com/artist/and...
Instagram: / andywalshart
Twitter: / andywalshart
Linkedin: / andywalshart
------------------------------
Buy my prints...
Etsy: www.etsy.com/shop/nostalgiapunk
------------------------------
Buy my tutorials...
andywalsh.gumroad.com
www.artstation.com/andywalsh/...
------------------------------ - Věda a technologie
bro, I am not a scientist or anything but I am like 99 percent sure, people who don't think AI is utter sh** in 2024 and can do what you do have your job. If your results are sh** with AI, you said it yourself, AI is not doing the work, the operator is. Get it?
You can’t even write a coherent sentence. You’re jobless.
@@AndyWalshwhy did you lose your job?
pin of shame
I’m an illustrator and haven’t really experienced any dry spells since Covid but I’m sure there will be some clients turning to AI or cheap illustrators who use AI but it can’t do revisions to illustrations, stay on model or anything that needs to be authentic to a specific culture or time.
Its an very encouraging message I see from Feng FZD design, yourself and even musician youtubers that creatives simply must source from reality more often and leverage their reasoning capabilities, intentionality to keep ahead of the curve.
Yet everytime the ML engineer crowd on twitter hypes the elimination of another modality/craft in a tone deaf unawareness that aspiration would die with it. it does feel disconcerting.
To share your ideas online now appears one must try to defeat or escape the algorithmic beast of obsolence.
I was watching a video of Trent kaniuga. He's a popular concept artist and he says a lot of people don't know what concept art is and what they're actually hiring you to do. Concept art isn't like illustration. It's like blueprints. You're trying to solve a problem. Like the functionality of the thing you're drawing like props. showing how it works and what its function is.
Or like a blueprint of a temple on how the character would like get from point a of the temple to point b and how they would have to like traverse through a specific part and how they would get there to get to the top.
He says they don't really teach you that in school and people don't even know how to put together a portfolio. After like spending $200,000 on Art school. A lot of people are just submitting him pictures of illustration or like of their dog or something related to what they want to work on.
I think the reason they don't teach you that in art school is because the instructors are afraid of you taking their jobs in the future so they sabotage you. I was watching like proko Podcast and Marshall said a lot of people in art school. They don't teach you the things that you need to learn because they're scared that you'll take their job
Yes, totally True
that is some foul shit with the instructors wow
@@mudbutt234 yeah I had that experience in school. They weren't teaching you anything. Luckily it wasn't an expensive school. But yeah, Marshall from proko talked to a bunch of professors and a lot of them do sabotage your students because they're afraid they'll get replaced. And I saw this other Art Channel. He used to be like a professor for drawing and he seen that in his university where a lot of professors just don't give out the information that students need because they're so afraid to be replaced.
Concept art is not anything particular, it can be many things, it's literally just concepts.
In fact the best approach to concept art is a state of complete openness and a curiosity towards all things.
Trent's approach to concept art is highly technical and "intricate", it's very production oriented. That doesn't mean it's the only way to do things. In pre production the client literally doesn't know yet what they want things to look like, they definitely do NOT need blueprints or turnarounds or callouts etc., they need ideas first, execution is an afterthought and just a method to get ideas. Ideally you should be able to do both pre production and production, however the mindset required for each is almost mutually exclusive because production art requires you to be very specific, organized and less open to new ideas.
@@MrMadalien thanks, that's good to know. Trent is the only one that talks about like the gaming industry and concept art. Not a lot of people talk about it so people don't even know what to include in their portfolio if they never had a job and concept art before. I just thought he would know what he's talking about since he's been working in the industry for so long and rent his own gaming company
Looking forward to this video as always.
I think the future MAY hold a dividing line, where some companies will draw a line in the sand and like 'we are an A.I light company that loves artists' type thing.
Ultimately artists want to work with great people. I've worked in the games industry for 16 years at this point and if the company/art director suddenly went all in on A.I then ultimately I wouldn't want to work there.
It's worth remembering that all the people pushing A.I currently are the people/companies who have something to gain from it. There are some people saying A.I could be a bit of a bubble.
I've seen a few companies post on their websites directly 'we have zero interest in A.I and appreciate people and artists so that's good.
It does kinda of feel like its going to go one of two ways, either it will just be a bubble that passes, or it's going to fundamentally change how we perceive the value of human creativity.
I do think we're about to enter an era of 'firsts' though. Things like, the first 'music video' created solely with A.I.
The first 'Movie', 'songs' etc {These things have probably already happened, but I'm talking about things that hit big/mainstream}
Anyway, apologies for the brain dump.
yeah it's hard to call it. It might just be all new and shiny now, and it then makes us go: oh yeah, soon you'll just make AI do aaaall the things. And then you try and go: oh, it can hardly do anything right. And that would be the bursting of the bubble. We really could have just done with another ten years of not having it though. Such bad timing.
"Artist want to work with great people" is a weird take because you are implying that AI prompters can't be great people. Sure, some of them might be lazy for never even wanting to learn fundamentals but what if a disabled person wanted to work as an artist but never had the capability until AI came along? The funny thing is that you might have already answered that question yourself. We have to fundamentally change how we see human creativity cause I guarantee that we will experience something made with AI that is so moving and made with soul that even us AI haters will have to give it props (yes, i too don't like AI that much even though it sounds like I was defending it lol).
@@andresct6250 I'm sure they can be great people, but they can't be great artists. In the same way someone clicking a button to create A.I music isn't a great musician.
People cannot take ownership of A.I output and expect to be classed as an 'artist'. That's not to say A.I can't output compelling art {copyright issues and all} that's a different matter.
@@HONDOMACLEAN83 Yep! In case you were curious, I don't use AI.
@@andresct6250 Out of interest, would you be willing to link to any of your work? Artstation etc?
Another good discussion, I find it very useful!
Thanks!
I knew you would refer to feng zhu's video. And whatever he said , makes sense actually
While I do agree that AI will probably suck at high-level concept art, I think it will affect the job market in the sense that most or a lot of jobs that juniors/small artists do will be replaced by it, either in small/not very good studios, that will simply adopt AI because it cuts a lot of costs, and a lot of studios/art directors in these small places, sadly don't care about these incongruences, mistakes, that AI has. They are more focused in simply matching the basic client demands in a short period of time. The client a lot of time simply doesn't care or knows well enough and the jobs they get aren't that high profile, specially in the advertising world, illustration, comissioning market, etc. A lot of these small entry-level companies today barely stay afloat and sadly if they could go without another artist, they would. I think high-level concept artists won't be affected per say, but probably it will impact the market in an indirect way that entering the "job" market and the industry will be much harder. We already have an oversaturation of people trying to get into the market, and AI will reduce that market even so more. It will probably replace a lot of illustrators and a lot of these illustrators might try to migrate to concept art if AI can't touch that. But that's my opinion. We also don't know how well or fast AI will advance so I think trying to guess what AI can and can't do in such a time frame is like trying to draw a line in the sand in the middle of a sandstorm. But honestly I wouldn't advise anyone right now to pursue the concept art job sense since that can be highly unstable.
Relax there's a lot of stuff to do for games, and movies, Ai is good just at mood, mood boards.
Like Feng said.. good luck with asking Ai to design props that have to work and look in a speciffic way. good luck with having it tell a story with the image, ask it to design an environment for a game using a speciffic art style and design cues that fit with the rest of the game.. yeah sure. :)
Illustrators won't be replaced soon, look at League Of Legends style splash art, the amount of story telling there is in those.. it's insane.
Art success was and is based on merit, nothing new, like it's said in games: GIT GUD
@@lucian5096 I don't want to be pessimistic, but AI CAN do illustration, specially in the style of League/Riot. It is in fact very good at that, sadly.
@@lucian5096 But what you described is high-level concept art. There's a ton of jobs below that will be hit. Think of all the mobile games that just need generic stylized art assets which has a huge existing body of work for AI to rehash from. Or those casino games that just need pretty colourful images. Or those NFT/Metaverse games. Heck, low-quality crap aimed at kids, who are not going to understand or care about the difference.
There's already AI that can sample art styles and reproduce them. Not perfectly, but enough to made a low-tier game aimed at casual players that won't know the difference.
LOL splash art is top of the industry, but there's so many cheap League knock-offs that don't have the same level of storytelling and worldbuilding.
I think the original commenter's point still stands.
@@ZephyrinSkies Yes.. there are always winners and losers, not just at art.. at everything. That's life, want to be a winner?.. GIT GUD :)
Or cheat.
If you go back to the industry charge them a lot more. Better to make new smaller independent studios. I'd be fine making less money but working on small teams that are full of cool capable people. Corporate will always do what corporate does, a d that never works out for creatives.
Great video! I wonder if major components of industries that use concept art/illustration might just totally change their approach to accommodate the limitations of Generative A.I.? So rather than not utilize A.I. to the fullest because it can't do what concept designers do, they will simply change the process and accept the "worse" product because it is more economic.
like, maybe function will start following form vs the other way around? maybe they'll just create unworkable arty things.
@19:10 This is what concept art is all about and people aren't teaching new artists about this. As a 3D modeller I need an artist to gather real life reference images for measurements/textures aswell as a full breakdown of how this object was constructed. Drawing 2D side view/front view is a million times more useful than a fancy illustration. Concept art isn't something you hang on the wall, it's solving issues for the other members on your team. The end goal is a product at the end, not an illustration. People might not like this idea since it's not flashy or pretty but that's what concept art is about and majority of new artists don't even know.
I think it would probably be better just to forget about concept art. It would probably be just as hard or easier to be a fine artist and sell your own work prints and things like that and submit your work to galleries and things like this. A lot of people start their own art business. I remember on the Proko podcast they said becoming successful as an artist or getting an art career is not easy. It's completely random. Some people just get lucky and start working right away. Other s spend years. Just trying to get some work. I think if you want to work as an artist it's just better to start your own business or get into teaching. It also doesn't seem like a good idea to rely on companies for your income because they only need you for a short while. Then you have to look for work somewhere else and it's rinse and repeat. You could be losing a lot of money. Just looking for working between. It's probably better to have your own business. If you can make it work then you don't have to rely on other companies for your livelihood who require you to learn a bunch of different programs randomly just for them to hire you.
ye true i think fibe art career is harder than concept art career - you have to be lucky and have connections in art world, of course you can be independent artist and find your niche it can work fine - iam thinking what to do next - should i stay on concept art road , or try more art (fine art ish ) way, i am and was creative and i dont want to consider other career so its also a mission 😊
@@veneficarius yeah, I'm the same way. I couldn't dream of anything else than working in an art career. I've tried other jobs. I've even tried going to university but I just didn't like anything else that was out there. That's not creative. I couldn't like see myself working as a doctor or a lawyer or something high paying like that. Even though like I would make more money, I wouldn't find it fun or rewarding. I'm stubborn. I need to be able to like what I'm doing. I'd rather do like a low-paying job that doesn't pay me as good then something that pays me good. So many people hate their job and are just doing it for the money and they are kind of miserable.
yeah I agree about not relying on companies etc. But I think one doesn't have to just become a fine artist as an alternative. One can build a following on social media creating something quite commercial and build that up. As long as you have an eye an sensibility for what will reach enough people to pay you a living.
The type of functional design Feng refers to isn't always needed on a project, that's the problem. What If your making skins for Fortnite or another ftp game? Cosmetic content? Sure, maybe once in a while your asked to design some elaborate gameplay driven prop or gun that needs to "transform" or articulate in some interesting very specific way. Again, that really depends on the project.
Its all concept art btw. concept art is really just about communication. Sometimes you need to communicate something very technical and robotic and gameplay driven, sometimes it's just some flashy costume for a Fortnite skin or a vis-dev painting of the world your creating.
True. I've never had to deeply design such things, but then, I don't have that kind of work in my portfolio. It certainly occupies about a third of the required design schedule I reckon. Or thereabouts. But hey, even a simply environment brief requires an artist to clean up and tweak.
Functional design is in everything though, even in character cosmetics. As a 3d character artist, if I need to model a soldier with a WW1 helmet, I need to do research to find out how the chin straps are riveted, and what kind of padding it has on the inside.
@@AndyWalsh there is a whole industry of churning out these type of "Concept Artists" here in Los Angeles (Art Center College of Design) where everyone is trained to basically go right into a sweatshop factory studio like Treyarch (COD) or Respawn (Apex Legends) where ALL THEY DO IS SKINS. Do u need high level of skill for these jobs? no not at all, yet its so hard to get into these jobs that ironically pay 6 figures for such a simple task. Why is it hard to get into one of these jobs? you have to be in specific cliques at Art Center College of Design and kiss their @ss.
Hello, do illustrators have big problems too, or are there glimmers of hope?
illustrators are f**ked mate. That's IF they're the kind of illustrator that just wants to get hired to draw any old thing for a book cover.
But if you want to be a successful illustrator, you MUST create a world, a product, a universe and build on it and sell it in every way possible all over social media. Be a business person first and an artist second.
As a Concept Artist coming from a Product Design background, and have worked on theme parks and game art, IMO design is very achievable to learn through self directed learning, especially for experienced artists who usually have already cultivated an above average eye for analysing shapes and proportions.
As long as you design around the framework that somebody has got to manufacture it, and somebody has to use it, most of it boils down to common sense/critical thinking about safety, and ease of use in relation to the human (or alien) body.
I'd definitely recommend for anyone reading interested in design, to start analysing everyday products around you, analyse where the form meets function, and where the designer got to make stylistic choices.
Most design doesn't require reinventing the wheel, so analysis of existing similar things usually is part of the design process for something new. You don't need to be an expert in every subject a client throws at you, just flexible and open enough to do in depth research before you start.
I wouldn't get too intimidated by those complex looking images, since the actual design side of those image can usually be distilled into very simple mechanisms that someone who can't even draw could do, then the idea gets sold via by fancy set dressing/detail.
I believe you are right. I have a background in ID with a major in Product Design myself. Though it's not easy, it's not rocket science. It's about redesigning everyday things 90% of the time with few opportunities to truly innovate. How it's made, the design of everyday things and books on processes are easier to get now than when I was in school.
How did you make the transition from product design to concept design? Are you in games, animation, or film?
@@dwintsterThe main thing that facilitated the transition for me was creating a specifically Concept art/Design portfolio, with the goal of making not only good designs, something fun or entertaining.
I made the pivot after I graduated and never pursued the product designer career job wise after that
To answer the other question, my experience is mostly theme park work, indie games, and 2 major board game shipped. All freelance with recurring clients throughout the years. I haven't worked on any films yet but I would still like to!
@AxiOll That's awesome. Thanks for the detailed response. Keep at it. I'm certain you'll get into film eventually as well. I'm currently working on making a concept art portfolio geared towards animation.
@@dwintsterNo problem, I wish you the best also!
I will say IMO now more than ever building personal connections will be important for a career in concept art. As much as people want to talk about A.I., people have to remember that the people hiring also have a job they want to find fulfilling on a day to day, and that usually involves people.
The few repeated opportunities I have had was often not because I was the best artist they had ever seen, but because I did the job well and they liked working with me.
That being said, despite all my accomplishments, I'd be lying if I could say that the freelance lifestyle I've chose is financially secure.
There is plenty for me to learn on the business end, which is why I'm watching refreshingly honest videos like this. Shout-out to Andy for the good timely video's.
@AxiOll that's true with all businesses, I have worked in family business since I was 8 on and off in Jamaica. I also ran my own business a couple of years ago along with freelance. Even with brick and mortar store there are fast periods when you are making good money and slow periods when you might be breaking even or even at a loss after all your expenses.
You do your best to keep positive relationships with your clients and customers. 90% of your business will be repeat business. Right now I'm also trying to get a 3d printing business started, you have the background (product design) for it like myself so it might be worth looking into if you are interested as a way to supplement your income. To paraphrase Maya Angelo, "People will not remember what you did, or what you say, but they will always remember how you made them feel."
I found a new video on Lightbox expo's CZcams page on Ai interesting. It seems the film industry is trying to put up some type of resistance to it or safeguards. The gaming and animation side of things needs to catch up.
That's good to know. A few game companies individually will likely state NO AI in their company guidelines. Funny story though, a recent client stipulated NO AI in their contract, but supplied me with AI reference lol.
@@AndyWalsh Wow at the AI reference. I know nightshade has been released and Glaze 2.0 just came out a few days ago, so hopefully these 2 combined are protecting artists. Also Hive moderation is an AI detecting tool that is the best on the market that actually has less false positives than other detectors.
I think it's a good mental exercise to think about how timeproof your ideas and assumptions really are. IMO, most AI debates about what AI can and can not do are solved with a single word: "Yet".
Feng says AI can't do functional illustrations, it certainly can't do them YET. AI can't match interior to exterior? Not YET but wait till some tool 3D tool gets invented where you make a crude 3d block and AI fills in the rest. Simply put, don't understimate the amount of development that is going into the AI field. If you think I'm exaggerating, look at the development rate of AI video and then tell me that AI will NEVER be able to do whatever you're thinking of.
IMO, the best thing you could do if you already have a good understanding of fundamentals is to learn AI and whatever AI tools come out in the same way you learned how to use Zbrush or UE. As Andy said in the video, AI tools that mix real drawing with AI will be powerfull. I can see tools like Nvidia's Gaugan become a good middle ground between art and AI where everyone will be relatively happy. Artist get to draw, AI saves time and companies save money.
As pessimistic as my comment my sound I do hope that y'all have successful careers. Good luck😁
I hear you, and I used to say that until I really (and very recently) realised just HOW MUCH human intelligence is needed. It'll have to be artists using AI rather than AI bros using AI. I reckon that's how it ends.
I think there is going to be now a more evident difference between a Concept Artist and a Concept Designer
yes, for sure. There really needs to be two different terms used.
I’ve heard a couple of illustrators criticize AI art for lacking the emotional quality behind it due to its fundamental design as a program. Furthermore, they predict AI art is going to feel repetitive and unable to really connect on an emotional/empathetic level, leading the industry away from its use and back to actual artists. I’m not sure exactly how that’ll impact concept design since it is different from illustrations, but I think it’s a valid point against the use of AI. Also, in terms of the boat and casino you showed, I’ve heard that referred to as production design.
yeah well let's hope so. I'm sure a certain aspect of the market will forever be taken away though, as some people just need a quick illustration for their first book and can't afford an illustrator to do it, whereas once they may have had to find the money somehow. We'll see I guess!
Practicing architect here who specialises in 3d design and illustration. MJ / generative AI is almost a collaborative partner, Pinterest of steroids, it’s a useful tool but not a means to an end on its own. You still need to make / design / draw / work it all out yourself, but It can be useful. It shouldn’t be discounted.
yeah, to be honest, it can be useful for very specific reference if you just can't find it from the usual channels.
this is just pure copium but it is very much needed in these days, thanks for the dosage comrade
very interesting word!
That is not a concept art porfolio. Infact that is a Key frame porfolio. You are correct, you belong to the art station bros.
We gotta do something to get directors to look at our stuff. It is what it is.
Unfortunately, all my actual concept work is under NDA.
I was in the same class as the guy that designed that phone camera - he's an incredible concept artist and super nice guy! My portfolio has a couple of these pieces of work and i can explain the process to you if you want!
Ive heard horror stories from FZD school. Mind games, grueling work deadlines. Not to mention you gotta be rich ;-)
@@Auxius. Yeah, it's very intense! Was quite frustrating to learn but was super fun. I had saved my money for like 3/4 years for it as well.
@@RobG95 Glad you're still with us Rob.
Would be great to learn! that level of stuff is just outstanding.
People considering high level concept design shoul really ask themselves if doing cut aways, orthos, prop design, among many things is something they would like. Worst thing to do is get into a field out of desperation.
Feng has adviced students to assign themselves real world based projects and design for them. A key trait of a good designer is a curiosity driven excitement for the real world. A mindset that everything is, "cool." History is cool, engines are cool, animals are cool, architecture is cool, fashion is cool, plants are cool. The ability to move from one subject matter to the next.
Of course most of us known perspective and figure drawing, how many people have the obsessive, problem solving traits of a good designer?
Yes, that is true. I have a few subject areas that I really love but outside that there's a lot of subject areas that I just don't like and unfortunately, most client work falls into those areas. But those cut-aways and callouts are not exciting at all most of the time. Hence why we do these bit open landscapes and stick 'em on ArtStation and hope for the best.
I think that there are "styles" or "schools" of concept art - one is classic fzd /art center with this type of drawings and fundamentals - its maybe tje best? but some years ago many concept artists who wete not from this "style"/world like Darek Zabrock - he was painter and than become concept artist - worked in many ip's now he has a school - had different style and approach more artistic expressive, of course he can design , the same case was with Jama he was heavy in painting and did many movie concept art - he is one of biggest names - he too can deaign , but its not this style of drawing like FZD used, but he uses tools like 3d to show design , plus of course one pixel team - heavy 3d and kitbash "style" but they also create concept art - have clients maybe its all about tools , because i agree with Feng 100% that portfolios look like epic mountain with stick figure , and they lack the who and why ? questions, but again i think if designer/artist can solve this strictly in 3d its great , if he can do it in painterly style its great , or in drawing industrial design style - maybe concept art is broad term , where we have , painters, cinematic concept art -keyframe, than industrial design that we dont see much on artstation, many dudes like to paint after work so even pro will blast epic tower with dragon , after work on "landing gear and solving problems" in fact its good news that ai cant solve problems , that i dont have to make another more epic sunset and fight for likes on artstation with 10k other epuc sunsets, bad news is that i have to go back to fundamentals, learn design , solve problem, it would be cool to draw like Feng, but i think if you can sculpt in 3d and design its great too. Or maybe its time to rethink or try sthg else , maybe become solo artist , make yt and try to be personal brand - i like both ideas , i have freelance jobs both more concept art and more illustration but also with some degree of concepting, but i sometimes dont know - its ok but idea of being free artist who can live off patreon or selling art is cool - but i think every path is hard
the more you think like a designer and story teller of what your conveying the better. that takes imagination. thats where the new comes from. the new is what people are after. recomination is good but it isnt great. for great you need a squishy quantum computer with an energetic lifre form inside it thats eternal. the people most in promotion of ai are people who are in total ignorance of what a human being really is. lastly im gen x and you young-ins coming after....get your basic skills in..please. it will pay you massive dividends. no shortcuts....adopt that mentality and you will have no more insecurities regarding your skills. contrary to pop belief the basics in any field convey ability to adapt to anything. foundation...to leap from.
That's interesitng. That the creators who want to use AI vs humans are ignorant to what a human is and what a human needs. They just want eye candy.
You could probably learn this stuff if you go to brainstorm. I think it's called and there's other online schools. You can probably go to feng school if you wanted to. I'm guessing it would be cheaper than traditional Art school, but it's probably like $900 a class so you might have to spend like $20,000 or something
I saw a post talking about it being 25k for a year back in 2015. But I've also heard mixed reviews from an online brainstorm school. You need to avoid the situation of one teacher doing a one hour video call per week with ten students. Which is the main online model.
@@AndyWalshWhy should we avoid that? Genuine question
@@why.do.I.even.try. because you get almost no one-on-one time with the instructor. I knew a guy who spent thousands on such a course, had no one-on-one time, much of the lessons were pre-recorded so he just went through the lessons and yes, got paint-overs but didn't learn very much. Gotta get that personal time with a teacher.
@@AndyWalsh Okay that's understandable, thank you for answering. I was between choosing some classes or a one on one mentorship with an artist I really admire, who surprisingly is also cheaper than said classes and offers more one on one time than any other mentorship I've seen so far.
@@AndyWalsh yeah I heard that teaching is really stressful so a lot of teachers like hardly give you any one-on-one time because they just have so many students and they're teaching so many classes. But that's still no excuse. Like you're a teacher. You should know what you are getting yourself into. Your students are paying a lot of money. I think there's like mentorship programs where the teacher works one-on-one with you for around the same price. I've heard like regarding brainstorm. They said you should research the teacher before you even enroll to make sure you like their work and also ask around if they teach well because there's a lot of teachers that aren't that great.
I actually find all discussions about "AI versus Art" nonsensical. I mean, people seem to forget what art is. Art is a way for humans to express their emotions, concepts and views, plus such things develop too during the artist's work and career, which is the point of art (a growing feedback btw the artist and his/her work, something that transforms to the artists and develops his/her inner world), and the audience takes part of that mutual feedback and transformation too. AI is simply a "machine" that creates images, that's cool, but that has nothing to do with art or with what I mentioned. Thinking that AI will replace art is like thinking that human beings would stop talking and expressing their thoughts and emotions (when talking) just because there are some digital speakers that reproduce human voice. AI is a tool, not a mind, not a being.
well, it's mainly the jobs side of things that we're worried about. The income side of things. And actually, the online popularity side of things too. It could just be that in ten years, no one can tell the difference between an oil painter and a machine. So it'll all just get lost in a sea of art. So then people might be inclined to not bother or give up. Then after a generation or two, they'll laugh at the idea that people used to spend years learning art. That's the fear.
The counter argument here is that the business side can react to this by choosing not to care. I mean that this camera design is great, but in the context of a fictional 2D/3D/film setting (ie not designing an actual product which will physically exist and function), they don't need it. They can always shrug their shoulders and say, "Wait, I can just lower my standards and use what the AI spit out?" So we will always be competing with zero. That's the one point of pessimism I think you didn't mention in the video, but it certainly doesn't negate what you're saying. People like Feng Zhu will continue to have careers, but it's hard to say for everybody else.
I think that's the main issue with the entertainment market and AI. While other markets such as vehicle designs, industrial design, etc need everything to be functional, in the entertainment world and from a design standpoint, the majority of the public (and so, the interests of those companies that sell their products to that public) don't care if a design is truly functional or not. It has only to look barely functional in order for it to be credible for the average viewer. While of course artist truly have an eye for those things and can detect what is shitty design or not, the truth is that the average person only needs it to look "good enough".
yeah but still the problem exists that art directors have briefs, very specific briefs and AI can't take very specific direction and won't be able to for a very long time. Then there's round after round of minute corrections.
fzd comes from a industrial design background. That's a skillset so many of the modern concept designers don't seem to have. Big shapes, atmospheric perspective, color contrast, photobashing, 3d isn't even half the skill set you need to do that level of work. Dudes just got too lazy watching youtube tutorials. the great thing about that is that the ai is being trained on the basic shit. The way they label the data probably requires engineers and not a bunch of poor kids in pakistan.
my feeling is that a.i will just make it easier for smaller teams to do high levels of work. Won't need to be dneg levels, just good enough. There will be more work, but not as stable. The job will be just spitting out content as fast as possible and some of the technical details won't matter as much. Basically, the youtubeification/tiktok of high concept films. A kind of hustle culture. The higher quality works might have the largest most stable of followings and contain people like fzd on the team but the vast majority will be everyday artstation clone level of work.
I don’t think AI art can ever be compelling, not unless an AI can actually advance to the point of having a personality and point of view. All AI art are just dry transcriptions of the prompt
I don't think prompt ones will ever work but they'll definitely use AI to like get ideas and then redraw them. I forget which one it is but I saw a video on it. There's this AI where you're drawing something on the screen and then the AI completely makes them image based on a simple Mark that you made. It's making a drawing of your drawing and making like 10 different versions of it all in real time.
So even if AI doesn't replace artists, I think it will significantly reduce the number of artists like if projects like a movie or a game needs like 50 artists. It'll probably only need like five artists and only the most skilled ones. So it will be way harder to get into the industry because you'll need such a high skill level to even enter. He's talked about it in one of his other videos that they're only hiring artists that are like an 8 out of 10 before they were hiring artists that were like a six or a 5 out of 10. So every year you'll need to be better and better just to even break into the industry.
I tried that AI thing you mentioned. It was shit.
Find me this Danielle Sanfillippo, as far as I can tell, she's a makeup artist, also, we are not only coming for your jobs but for your entire industry. 4 of us promptbros are already working on the next blockbuster film and triple AAA remake of said film, at the same time. Sorry Charlie Walsh.
Where's your portfolio with briefs.
no you're not.
I really doubt it.😂
I've spoken to production artists who've hired prompters it didn't end well most of them have been fired or had their contracts terminated. It may be a brief glimmer of life for your kind but it won't be long lived. Nearly every single director I've spoken to hates the idea of ai gen and is doing everything they can to eradicate it. You may sneak throigh some cracks in the walls but it won't be wide spread. What I'm actually seeing happening is real artists getting trained on prompting then using it to bare discover then paint over.
Unfortunately when the law suits settle and the legality comes up you will loose. If you look at the current climate the EU has already put in place laws and regulation making promoting with stolen data sets illegal, not to mention the Ultraman copyright case in China where the artist won and now China has deemed "AI IS copyright infringement" setting a huge president in our favour. You won't win you'll be short lived and you will die out just like every other fad because you can't design you can't create you can't problem solve all you can do is push buttons
Also AAA is ranking for games not movies but I can see why you would get them confused